Board 8 > Tsunami's Post-Contest Analysis (should not need a second topic)

(+) Yeah! (+)

Probably should've tried to finish this up this past week.

---

Eh you can still do it. Probably.

---

Well, this coronavirus pandemic has given me some time off from work, so maybe I'll get back to this soon.

---

Match 146: Mario vs. Samus Aran

Mario 14343
Samus 14324

Board 8ers love them some status quo. There were many matches where the registered and unregistered voters were in disagreement, but just three where the registered vote bonus actually flipped the match: Phoenix vs. Ike, the first Mega Man-Pikachu match, and this one. This one was by far the most dramatic, though--both Ike and Mega Man would've won by fewer than 50 votes if all votes counted equally. In this match? Let's give a hand to the two lowest-scoring Oracle predictions, Samus 51.00% by Furious Fura and Samus 51.27% by paulg, because in a world where all votes counted equally, they'd have been a very close 1-2. That's right, 51.13% of all votes cast in this match were for Samus Aran...and yet because of the huge disparity between registered and unregistered voters, Mario still won by 19 votes. To be more precise, Mario had a 467-voter edge among registered voters, and Samus had a 915-voter edge among unregistered voters. I'm really disappointed that the registered voter bonus is back for GotD2; it isn't really going to do much to stop a genuine rally, but it can still flip a close result.

That said, this was an intense, back-and-forth match. Samus led for most of the early match, including building the match's peak lead by either side, 206 votes at 1:20 AM Eastern. That's just over a quarter of the way into the match. Things continued to be interesting throughout, featuring a dead tie at the 8:55 AM update, Mario winning an update by 2 votes at 4:20 PM and it constituting a lead change and Samus winning the next update by the same margin, and of course, the 50.00% on the penultimate update (Mario ahead by just two votes). It was a down to the wire match, setting a new record for closest wire-to-wire 24 hour match (only Shadow of the Colossus over Metal Gear Solid 2 in the first GotD was closer) and making the top 5 for closest final margin as well (trailing Liquid-Alucard in 2007, Frog-Chief in '04, the third-place match from Rivalry Rumble, and Charizard's win over Zelda in 2013). And in that respect, the registered voter bonus did make things more interesting, because like I said, in raw votes, Mario didn't even break 49%. Our update trackers can only show the listed totals, with us only getting the raw vote data for the end, but just looking at the final margin, it certainly wouldn't have come anywhere near the record for closest wire-to-wire match--even if Samus's biggest lead was right at the end it would've been more than double what it was. Wouldn't have even made the top 25 for closest final margins even among 1v1 matches alone. It probably would've still gotten lesser spots on the wire-to-wire lists, and it might've even scored fairly high on "wire-to-wire matches where the loser never takes the lead"--because with a maximum lead of 128 with the registered voter bonus, it's hard to picture that Mario would have ever had one without it. Well, maybe when he had the 104 vote lead a mere 40 minutes into the match. That would've already precluded it from the loser-never-takes-the-lead list, since Samus did lead at the freeze and one update later as well. Geez, you guys really hate it when the status quo is upset, don't you? Because make no mistake, this is "FREAKING MARIO" to the max. You always expect him to pull out any close match, because he's Mario. And right here, he did.

But this is definitely a reason to pay close attention to the types of games this site likes and doesn't like for GotD2. RPGs will reign supreme. So will remakes that somehow were allowed into the contest, because this site absolutely adores the 20th century. Keep that in mind, and you'll probably go far.

---

I'll finish this up tomorrow. I think.

---

Legends Bracket Final: Link vs. Zelda

Link 18399
Zelda 10547

There was a very real fear here that we were going to get an SFF blowout in what would be the final match if we still had a single-elimination format, resulting in some absolutely hilarious raw X-Stats. Then again, it's not like Link had been all that terribly challenged in earlier rounds either. There was a lot of debate as to which matches to use for the purpose of calculating X-Stats, the Legends Bracket (because that's what a traditional bracket would have produced) or the Losers Bracket (because those were the ones that caused eliminations and thus "meant more"), but let's face it, either way they were going to be inaccurate. Which pretty much is the same as every other year, because character battles have never been transient. The raw X-Stats are a joke because of SFF and bandwagoning and all sorts of weird factors that make up the matches, and the adjusted X-Stats are an even bigger joke because they represent guesswork and are subject to biases. In the early days, when a character was behind SFF, the tradition was to just assign them the previous contest's X-Stat value, which is how we ended up with 84% of Gurus picking Magus to win three matches in a year he won none. Frankly, this still comes off as more sensible than the current method of casually ignoring direct results that are inconvenient to the current narrative because there was obviously some sort of rally (example: Mario being ahead of Zelda in this year's adjusted X-Stats). I mean, yes, when there really was a rally, it massively obfuscates any sort of meaningful calculation, especially since it becomes harder and harder to figure out what the entrant's unrallied strength is because the rallies show up earlier and earlier. We all joked about Mass Effect 3 being #2 in the X-Stats in 2015, but it was the one that forced Undertale to make its largest comeback. Meanwhile, the "just assume that Link/Legend of Zelda is #1" logic wasn't even close to viable because the finals against Ocarina of Time was legitimately the only time that Undertale never trailed. So which of Undertale's victims was given the #1 spot in the adjusted X-Stats? Pokmon RBYG, which had been building a lead even as Undertale was trying to rally? Melee, which actually managed to wipe out a four-digit deficit against Undertale before eventually fading away? ...Yeah these are really the only two logical options even if technically Mass Effect 3 was closer. And the answer is...none of them, because the acceptability of a rally is based not only on the rallying entrant, but on the opponent--possibly more on the opponent. Well, no, it's on both. Zelda may not have been allowed to remain ahead of Mario in the X-Stats, but I doubt those people minded too much the previous match when Samus would've beaten if not for the registered voter bonus, even if the fact that it was the registered voters who allowed Mario to win suggests otherwise. Because Samus is a member of the Noble Nine, and one who the X-Stats frequently suggested was indirectly stronger than Mario anyway. Her finally beating him is good theater. Zelda being here was a "farce". Point is, the whole board was fine with RBYG and Melee counterrallying against Undertale, but Melee has to sit behind the last two foes it vanquished, Chrono Trigger and Final Fantasy VII. OoT actually did get the #2 spot, ahead of FFVII but behind CT. They saw CT put up over 81% in Round 2 and then get 65.52% on a game that had been in the Top 4 of each of the last two games contests and they were like "yeah CT's winning this as long as we can stop the Undertale monster". Now we couldn't, so it was a moot point, but let's be honest here: if Melee hadn't gone on rallies, CT absolutely still loses to FFVII. The stats people would've still put CT as the stronger game, because that's an SFF match, but FFVII doesn't lose to other Square games. We've seen those two games face each other before, and there's no good reason to think CT could turn it around (though there are plenty of bad ones, like backlash against the episodic content of the FF7 rerelease or the fact that other games are declining while CT, being so old already, can't really decline because everyone left here is part of that core fanbase. Except these contests have never been decided solely by the board's or even the site's tastes, and the fact that the registered voter bonus only flipped three matches when there were far more where the registered and unregistered voters disagreed only accentuates this.)

Anyway, it turned out to be a moot point because Zelda easily avoided the doubling. In fact, let's go back and look at the last time Link and Mario went head-to-head in a character battle.

https://gamefaqs.gamespot.com/poll/3844

And now Zelda's win over Mario:

https://gamefaqs.gamespot.com/poll/7379

Based on that, Zelda would be expected to get about 36.07% on Link. If you adjust for the registered vote bonus in the Zelda-Mario match, it'd be more like 36.77%.

And her percentage was...36.44%. Actual percentage 35.78%, because yes, Zelda actually had a better percentage with registered voters than unregistered. Still, that's not that far off, which makes it all the more ridiculous that the adjustments put Mario ahead of Zelda.

---

Match 148: Cloud Strife vs. Mario

Cloud 14667
Mario 14526

Well I guess we have our answer as to why Mario was ahead of Zelda in the adjusted X-Stats. The raw X-Stats may clearly use the Legends' Bracket numbers, but if the adjusted X-Stats were basing things off of the Losers' Bracket, Mario did put up a better performance against Cloud than Zelda would the following match. After Cloud held a narrow 250-247 lead at the freeze, Mario took the second update by 66 votes and proceeded to extend his lead past 300 within the first three hours of the match, peaking it at 335 about 7-1/2 hours in. Cloud would fight back over the remainder of the night, getting it down below 50 around 6 AM, but Mario rode the morning vote to a lead of over 200 before Cloud started fighting back again.

Cloud took his first lead of the match with 4 hours and 35 minutes remaining, and the next update only increased his lead by a single vote before Mario took the next update--only by 5 votes, but that was enough for another lead change. But Cloud won the next update by 21 to take the lead for good, and the two battled for a bit, Cloud only taking a double-digit lead for the first time in the final hour. It was the #8 closest 1v1 wire-to-wire match, #6 among 24-hour matches...and yet, only #5 for this contest! It meant that over half of the Top 10 came from this contest, except they already did because the match that it booted down to #11 was also from this contest. Seriously, the all-time #17 (again, for 1v1 matches only, since we keep the 3/4ways in a separate list) can't even make the contest's Top Ten. Maybe it's the registered voter bonus making things tighter than they should be, or maybe it's just the lowered vote totals, but this contest had a lot of really close matches, strictly speaking. Not all of them felt close. Here, let me give you a rundown of the overall top 25, except instead of listing the actual match, I'll just list what contest they're from: Asterisks are 12-hour matches.

1. GotD*
2. CBIX*
3. CBX
4. GotD*
5. CBX
6. CBX
7. CBIX*
8. BGE3
9. CBIX*
10. CBX
11. CBX (that'd be this match)
12. CBX
13. GotD*
14. CBIX*
15. BGE2
16. CBX
17. CBIX*
18. CBIII
19. CBX
20. CBVIII*
21. CBIX*
22. CBX
23. CBVII
24. CBIX*
25. GotD*

9 out of 25 from this contest. Even CBIX, which had similarly rancid vote totals whenever Draven was absent and had matches that ran only half as long, only had 7. Overall, 24-hour matches do now hold a 13-12 lead over 12-hour matches in the all-time top 25.

I wonder what the upcoming Game of the Decade will look like. Will it break the records set here? And if it does, will it be because of intensity, or because of complete and utter apathy? I look at Minecraft, a casual game to end all casual games, and yet I have to take it to win easily in Round 1 because it's up against a MOBA and we absolutely abhor those after 2013. I look at Yakuza 0 getting its characters absolutely slaughtered in this contest and take it to Round 2 because it's up against something called The Witness. Overwatch seems like it'd fall squarely in the category of "things Board 8 hates, and is possibly slightly scared of the rally potential of", but iirc Death Stranding was considered a massive disappointment by all of those eagerly awaiting Kojima's next masterpiece after Konami kicked him to the curb and so I have to side with the casualbait. Except then you hop on over to the latest extruded Blizzard cash-extracting device and you're like "no, fuck that, I'm going with 'niche things that Board 8 loves'." I mean yeah it's probably still the wrong pick but it's a 7-10 match so eh, it's only worth one point. Though frankly the biggest problem is that all of these new games have generic "edgy" names. I initially had Bloodborne-CoD:BO as one of those "why do I have to pick one of these to win a match; they're both going to be megafodder" matches. But Bloodborne was actually respectable in BGE3, going up against a Final Fantasy game no less (not that I really trust recent Final Fantasy games to not be fodder either. Square spent most of the decade pushing Final Fantasy XIII sequels on us, and we responded by not putting any of the FFXIII games, not even the original one, in the bracket. And yes it was eligible; we were going by NA release dates unless no such date existed and Lightning rather famously appeared in a Winter 2010 Contest match before her game had released in the US. Not surprisingly, it was the strongest she'd ever look). The game that I was thinking of that got eviscerated was called "Hearthstone". Meanwhile, that same theory of "wait, this has actually been in a contest before" got me second-guessing taking Stardew Valley over Destiny, except Destiny actually did get blown out by something GameFAQs hates.

---

Losers' Bracket Final - Zelda vs. Cloud Strife

Zelda 13517
Cloud 14101

Next up in the department of totally made-up "common knowledge": all those classic RPGs are weakening because we're the only ones who still like them, but they'll never completely turn into fodder because this site is run by exactly those people, the ones who are still stuck in the 90s.

Yeah, not happening. For the second match in a row, Cloud would lose with the registered voters but still win because of his lead with the unregistered voters. I mean, yeah, antivoting FFVII is nothing new here, but I guess it's because we expect it to win anyway? Or perhaps we're dealing with "outsiders" who had GameFAQs accounts back in the day and still can log into them, or Board 8's tastes aren't representative of GameFAQs as a whole, because supposedly the board should've been pro-Cloud in both of these matches, yet the registered voters alone wouldn't have given him either of these matches.

Zelda's peak lead wasn't all that far behind Mario's, but Cloud caught up much sooner--about halfway through the poll--and cruised to what for this contest probably qualifies as an easy enough victory, nearly winning by 600 votes and breaking 51%. I wanted to try to give an example of a match from our heyday where neither side ever broke 51%, but it turns out that a lot of our classic matches featured one side getting off to a sizable early lead and the other just not letting them go away easily. ...Actually that wouldn't have been the case for this one, either, even if Cloud hadn't gotten up over 51% at the end. I suppose that's a natural effect of every vote being a great portion of the percentage early on. As best as I can tell, it's never happened in a 24-hour match, not that that's a relevant fact since the disqualifying factor is usually that the match starts too uneven, one side getting a little bit of a percentage lead. Yet looking through the list of closest wire-to-wire matches, the only ones I can find where neither side cracked 51% past, um...wait, what should be the cutoff? Well, our list for "loser never took the lead" allows the loser to have the lead at any point before the end of hour one. So the two close matches from GotD definitely qualify--MGS2-SotC was inside of 51-49 at the freeze, an extreme rarity, cresting above it for a single update 10 minutes later, while Team Fortress 2 was at 53.15% on Fable at the freeze but that was the last time either side would even reach 50.6%. The closest 24-hour match I could find to pulling it off wasn't, in fact, Pikachu's victory over Mega Man (Mega Man didn't fall below 51% for good until the five-hour mark), but..uh...huh. I was about to say the Xenoblade Chronicles-DKC2 match, but then I was like "wait, did that ever go beyond 51-49?" because when I was scrolling to the beginning I only found one update listed at 51-49 exactly (and calculating it properly, it was still inside of 51-49, but I mean just barely, it's literally 50.999%-49.001%), but I could've sworn I'd seen it get bigger, and it turns out it was later. But even so, that might be the true closeness record for percentage because that match never, and when I say never I mean not only after the freeze but also at the pre-freeze snapshots for 4, 3, and even 2 minutes, got outside of 51.1%-48.9%. So I'm definitely optimistic that GotD2 will bring more thrilling matches.

---

Grand Final: Link vs. Cloud Strife

Link 16554
Cloud 11375

I mean, did we really expect otherwise? The final match was going to be a rematch no matter what. There was pretty much no way that the final match wouldn't end up being a rematch, because for it to feature someone who hadn't already faced Link, that someone would have to first beat someone who had beaten them earlier. Granted, Mega Man did exactly that this contest, and Samus nearly did, but it's a huge ask. I don't know, maybe if Samus had been able to pull out those few extra votes against Mario, she could've ridden a combo of the rallies that seemed to be popping up for every female character and her actual strength that got her to the finals of CBIX and won her the main bracket in '06 to upset Cloud, then SFF down Zelda to give us a fresh match. Which she'd lose, of course, but it'd still be better than having a rehash of a match we'd just had. That probably wouldn't have happened, though; beating Cloud without some mitigating factor is never an easy task if you're not Link. Snake's only victories over Cloud came when Cloud was being LFFed, first by Sephiroth and then by Crono (and maybe also Link, because swords? But mostly Crono). And his only other non-Link losses were...Mario in the infamous Planet Gamecube match, L-Block, and Squirtle in the Draven contest. And I guess the 3rd-place match of Rivalry Rumble but no one really cares about that. Actually when you consider she broke 47% on him in 2010 after being held just under 41% in 2003, maybe it's not so unreasonable to think that this would've been her year. Yeah let's go with this. Someone make up some adjusted X-Stats that say that Samus was actually the #2 character in 2018 and only lost because Mario continues to be a bad matchup for her. It honestly probably wouldn't be any more wrong than any of the other ones we have floating around. Hell, if you use the raw vote totals instead of the ones adjusted for the registered user bonus, Cloud got 51.37% on Mario, while Samus got 51.13% in her second match. That's still a slight advantage for Cloud, but it's really slight and character battles are not and have never been transient. Given the way trends were going at the time, this seems really probable. As for Zelda, the percentages would be more along the lines of their 2006 matchup (roughly 55-45) than their 2010 matchup (roughly 64-36), but it's hard to think that Zelda could keep up her momentum when faced with another powerful female. Zelda may have long since moved beyond her damsel in distress days, but Samus has always been the heroine of her games and her first game is just as old as Zelda's. She's the original feminist gaming icon, and not even Other M was able to ruin that.

Anyway, since Samus wasn't able to get that last push against Mario, we had Cloud in this match instead. He improved on the first match against Link, managing to break 40% this time. But this was never going to end in anything other than a comfortable Link victory. Honestly it's hard to imagine that Link will ever lose to anything other than a massive rally. Even Draven barely hung on at the end, relatively speaking--his raw margin of victory was higher than a lot of matches this contest, but we had 150 matches in this contest and if not for registered users' votes counting double, Shepard's vote total in the Link-Draven match would've been good enough to beat the winning total in 126 of them. Of the 24 that it wouldn't, 15 of these were in the first round, 5 in the second, 3 in the third, and then there was Link's SFF beatdown of Ganondorf. Even with the double-counting votes, Shepard's total was higher than Zelda's winning total against Mario, Crono's against Mega Man, Tifa's against Sephiroth, and the asbestos standard, Bowser's winning total in LB1 against Alucard, which failed to break 13000. The actual raw total of 8578 is almost certainly the lowest winning total in any character battle (though it's still better than the Years contest), which doesn't sound too bad until you remember that CBIX was three-ways with 12-hour matches. But even in a 12-hour match where third place had 30%, the lowest winning total in that contest was 8979. In 24 hours, in a 1v1, Bowser got fewer votes than that and won.

---

If votes do go up significantly in GotD2--which I kind of expect them to--we should just stop having character battles. Do people even really care about individual characters anymore? So many of the newer games have customizable characters anyway, which is probably why so few new games even get characters in these things. I mean, let's look at the characters who debuted in this contest. Aren't like half of them from indie games, which have a disproportionate tendency to be "retraux"? I'm not going to go over them all and count them, but it looks like most of them are either from indie games or from newer entries in established franchises (e.g. Noctis, Aqua, Captain Toad). And out of 150 matches in this contest, do you know how many were actually won by a character making their contest debut?

8. 8 matches out of 150. 5.33%. Except that figure is inflated, because there were 5 first-round matches where both characters were making their debut, leaving a grand total of 3 matches where a character making their debut beat a character who wasn't--all of them 1-seeds (Geralt twice and 2B once). It's not 3 out of 145, though, because there were also some first-round matches between characters with contest experience, and even fourpacks where all four characters had contest experience. (For a moment I thought Division 1 had an entire eightpack like that, because of generically named male syndrome; Victor Sullivan was the only newcomer in the bottom half of the division). It's 3 in 104. There were 104 matches out there that, at the start of the contest, could have been won by either a veteran (even by the loose definition of "1 previous contest" making you a veteran) or by a newcomer, 41 guaranteed to be won by a returning character (34 of these in Round 1 and 7 in Round 2), and 5 guaranteed to be won by a newcomer. Of those 104, new characters won 3 of them. 2.88%. I feel like that's low. Maybe it's not really that low, because the more character battles we have, the more returning characters we have, but it seems like it ought to be low. It's obviously lower than CBIX, because Draven alone won 4.13% of all non-bonus matches. CBVIII? Honest to god, there were more matches that were guaranteed to have a winner who'd been in a previous contest than not. 66-61, including a trio of R3 matches, and of course not a single match where both characters were making their contest debuts. Two of those three divisions where one of the R3 matches was guaranteed to have a returner win also had a third R2 match with such a guarantee, and a seventh R1 match on top of that. Literally only one character out of 16 in the division was making their debut.

One of those two characters that were the only ones in their entire fucking division to be making their debut was Charizard, who won said division. That alone accounts for 6.56% of qualifying matches, and 3.15% of total matches. So yeah, 2.88% is low. A single newcomer managed to singlehandedly exceed that percentage without discrediting matches where all possible winners were returners in each of the previous two contests. Same goes for all earlier ones, because there were only 63 matches in any of them which means it only takes 3 wins. I mean if you want to quibble about requiring a first place finish to count as the "winner" in the fourways, then no single character did it in 2008, but you had one first round fourpack where two newcomers beat two returners, and you did have three newcomers win their first round matches outright and reach Round 4. 2007, L-Block did it. 2006...okay, 2006 was probably lower because any female character with experience was naturally stronger than any of them that needed the split bracket to make their debut. But that percentage is horrible. Character Battle is dead; long live Best Game Ever.

Oh right this vote total thing distracted me from my original point, which was that while Link would've won no matter who he faced here, I honestly believe that if she'd just pulled out those few extra votes against Mario, Samus would have put up a better percentage than this. Not much better, probably only around 42%, but better. Unfortunately this undermines the point I just made because we'd need another Character Battle to test this theory.

---

Nice that you finished that before the new contest starts. And to tell the truth you are right. No more character contests. As time goes by the old characters will keep doing better and better because they just have more games. New characters can't do anything.

For example the main character in Red Dead Redemption 2 which has sold around 30 million games so far would lose to a character from Animal Crossing simply because Animal Crossing just has more games.

So yeah. Bring the games. I guess for next year we should have worst game ever, or stuff like that just to break the mold.

---

No, "Worst" contests are a bad idea.

And I'm not sure if that particular example works, because Animal Crossing isn't really about the other characters for the most part. It's another one of those games with a player avatar, and player avatars are, for the most part, not going to do well. It's not impossible--Commander Shepard is a customizable character and they were able to become a midcarder for a time, but even they seem to have faded. Maybe. Okay, now it's time to do a character analysis:

2008: Finished in 4th place in R1 against Sonic, Sandbag, and Magus. The margin looks bad, but a Noble Niner, a proven midcarder (albeit one with a reputation for choking), and a joke entrant are tough competition when you've only had one game so far. This isn't a horrible outing.
2010: Breaks 75% against fodder in R1 before a 39-61 loss against Pikachu. Mass Effect 2 was already out by this point, and Shepard is clearly a midcarder here--Pikachu was top ten in the raw X-Stats in 2010, and while his raw number is almost certainly inflated because he got a HGSS sprite vs. Solid Shit, it was pretty clear that, yes, Pikachu was already among the 10 strongest non-NNers by 2010. You could probably argue that Pikachu hasn't had a single bad result since returning from hiatus in 2007, and you certainly can't say he's had a bad match since 2008 with the possible exception of letting Mewtwo edge him out for fifth place in 2013.
2013: Cruises to a fairly easy round 1 victory, though their failure to double up second place shows that even fodder Nintendo is still never going to completely fold to something non-Nintendo. Pulls off an impressive win in Round 2 over Aerith, the first time that Shepard went beyond just "looking like a midcarder in a win over fodder or a loss to a higher midcarder" and actually beat a midcarder. Round 3, uh...Shepard was, as I mentioned above, the poor unfortunate soul to draw the spot of "also technically present in the Link-Draven duel."
2018: By raw results alone, Shepard's path went exactly as expected: beat K. Rool in R1 and lose to Ryu in R2. However, only winning with 51.23% against the former and narrowly avoiding the doubling against the latter is...a bit of a disappointment. This was Shepard's worst contest since 2008 at the very least, if not the worst ever.

---

Okay let's do this.

Match 1: The Legend of Zelda: Breath of the Wild vs. The Outer Worlds

Breath of the Wild 19352
The Outer Worlds 3409

Not much to say here since it's the prohibitive favorite destroying some random fodder that I've never heard of, but with arguably the most well-received Zelda game since Ocarina of Time itself making its contest debut...vote totals were still down from the previous contest, where our opening match was Dante vs. Cuphead. Okay, yes, with this four-matches-at-once format, technically we also had Legend of Zelda in an opening match in CBX, too, but here's the scary part: when you take out the registered user bonus, not only did BotW-TOW draw fewer voters than Dante-Cuphead, but Breath of the Wild drew fewer votes than Dante! The registered-unregistered ratio was more lopsided here, though; in Dante-Cuphead, registered voters only outnumbered unregistered voters 8819-8756, whereas here it was 8093-6583. Which, uh, is kind of scary given that BotW's winning margin among the registered voters was 6925-1168. More registered voters voted for BotW than unregistered voters voted period! That's ridiculous. I guess there was a lot to say after all.

Match 2: Halo: Reach vs. Life Is Strange

Halo: Reach 12677
Life is Strange 9899

It's hard to have faith in anything Halo here, but Life is Strange was pretty fodderiffic when we saw it before. I thought that maybe with B8's penchant for VNs, Life is Strange could pull off the upset, but I forgot that they hate the game. Either way it was a meaningless 1-point match--and there was at least some debate here, a 77.3-22.7 split on the Guru compared to some of the other awful picks I made.

Halo really does look pretty bad here, even in victory.

Match 3: Final Fantasy XV vs. What Remains of Edith Finch

Final Fantasy XV 16854
Edith Finch 5722

Noctis may have embarrassed himself in 2018, but this was never in doubt. Final Fantasy may be declining as a series, but I think the opinion on XV was "it's still not great but it's better than XIII" and besides, What The Hell Is Edith Finch. I've heard it described as a "walking simulator", which I think means it's some sort of casualbait.

This match followed the trend of nearly all of the early matches: despite the disparity between the number of registered and unregistered users, the loser's raw votes were very nearly a 50-50 split between registered and unregistered, while the winner's registered votes exceeded their unregistered votes by over 1000. The exact meaning of this, however, would not be clear until we saw a match that broke that trend.

Match 4: Hollow Knight vs. Tales of Berseria

Hollow Knight 14183
Tales of Berseria 8399

This was considered a very debatable match, because Hollow Knight was an indie game, which rarely fared well, while Tales of Berseria was, well, just look at the first two words of that title. The Tales series has a grand total of four wins in contests. Tales of Symphonia has three of those--once with Lloyd winning in Character Battle IV, and twice with the game itself advancing to Round 2, in both the fourway BGE2 in 2009 and in GotD in 2010, though it should be noted that its R2 loss in GotD was literally to one of the two games it beat out in R1 of BGE2. The most recent was in Character Battle IX, where some combination of DAT TOP OPTION (technically left option now, but still, higher seed in a fodder match) and TJF allowed Berseria's protagonist to reach R2. And while the fact that Berseria already had more wins than Tales of the Abyss might have been encouraging, we all knew full well that that was a fodder-vs.-fodder match, and Yoshi proved it by breaking 70% on Velvet in R2. The distrust of all things Tales narrowly won out--in order to avoid the procrastination that plagued my 2018 write-ups, to the point that the final rounds have references to this contest's bracket, I'm doing reactions during the Contest this time, so Guru submissions aren't technically closed yet, but at this writing it looks like about 54.6% of Gurus had Hollow Knight. And Hollow Knight won handily, with the casuals beating out the Gurus by 20% on the prediction percentage.

---

Oh you do this contest too? I say you make a new topic like me though.

---

Match 5: Monster Hunter: World vs. Bravely Default: Flying Fairy

MH:W 13203
BD:FF 9581

Now this was an interesting match to talk about, if not to watch. It was in the write-up of the Suikoden II-Monster Hunter 4 Ultimate match that "The Flowchart" was introduced:

Nintendo > Squaresoft > Japanese RPG > RPG > Japanese > old > new > contest experience > top option > everything else.

However, while this did correctly lead to the answer that Suikoden would win because it was older, had contest experience, and was the top option, it was not, as stated in that write-up, a JRPG vs. a mere RPG. Monster Hunter is in fact a Capcom franchise, so that was JRPG vs. JRPG.

Here, however, it was a bit murkier. Bravely Default was the older game, 2014 vs. 2017, but that probably matters less in Game of the Decade because the gaps are smaller. Monster Hunter: World had the huge seeding advantage, and was released on consoles while Bravely Default was a handheld game. But Bravely Default was Squaresoft--Square Enix, technically, but while it was well post-merger it was a throwback to their older games, in the same vein as their earlier spin-off "Final Fantasy: The 4 Heroes of Light". Hell, look at those abbreviations again: Bravely Default even had a subtitle with the initials "FF"!--while Monster Hunter: World was, as stated before, Capcom, which stunk up the joint like crazy in 2018. Yeah, yeah, "Characters =/= Games", but still, about the only Capcom character to do anything impressive was B8 mascot Phoenix Wright winning one of the three matches that was outright flipped by the registered user bonus, and one of the other two resulted in a Mega Man loss so there you go, Capcom sucks. Well okay Chun-Li also seemed to impress in 2018 but nearly every female character in the bracket seemed to be a step up from before. And with unregistered voters, it was indeed pretty close; 3543-3173. But with the registered voters, MH:W was above 60%, so even without the registered voter bonus, this wouldn't have been close.

Monster Hunter: World had an 87.66% prediction percentage overall, so this was just a clear case of me overthinking things. Then again, we'd seen enough 15-2 "upsets" that the seedings alone didn't seem like an indicator that MH:W had this on lockdown.

Match 6: Sekiro: Shadows Die Twice vs. Ni No Kuni: Wrath of the White Witch

Sekiro 13347
Ni No Kuni 9442

And here's where The Flowchart, and everything else I thought I knew about GameFAQs, went completely down the toilet. I wasn't that familiar with Ni No Kuni and even less so with Sekiro, but remembering how Yakuza bombed all over the 2018 Contest and being pretty sure that Sekiro was an action game and not an RPG, I figured this was a lock. Ni No Kuni was fodder, sure, but Japanese RPG > Japanese, not to mention I'd probably take a FPS over an Action Game on this site. Just look at how Devil May Cry regularly struggles to even get games into the Games Contests despite Dante being a very good midcarder. It's clear that both he and Kratos are far stronger than their games. And if Yakuza couldn't even do well in a Character Battle, how was an action game going to do shit in this contest? And my Oracle was only spared from being in an even bigger world of hurt by my misconception that this was Japanese RPG > Japanese. It's not. Unlike the series I keep comparing it to, Yakuza, which is legitimately Japanese in origin, Sekiro is set in Japan but it's made by Activision, best known in this site's glory days for their sports games like the Tony Hawk series, and more recently for Call of Duty. When I learned that, I said I was glad I didn't know that because I probably would've taken Ni No Kuni to double Sekiro if I had. Researching them for this write-up, however, I learned that Activision also managed to acquire some old platformer icons like Crash and Spyro, not that those characters will ever be theirs, and they...are now known as Activision Blizzard?!?!?!?! Goddamnit, if I'd known that I don't think I'd have taken Sekiro to avoid a tripling by much! You've got a company that this site hates, in a genre that this site hates...and it's not only not challenging for the title of "lowest X-Stat value by something that isn't actively being anti-voted", but comfortably advancing to Round 2? And just like Bravely Default in the previous match, over 49.5% of Ni No Kuni's raw votes came from unregistered users; in fact, the splits are extremely similar with both voter groups-- 4833-3206 vs. 4877-3165 with the registered voters, 3543-3173 vs. 3599-3116 with the unregistered (in both cases, this match was the slightly more lopsided one). This was, plain and simple, the site's tastes being far different than what I...oh. I, uh, may not have looked quite close enough. Activision merely published the game internationally; it was actually developed by a Japanese developer, FromSoftware, best known for...the Souls series?! Well, geez, I feel silly now. The Souls series is one of the few IPs to debut during the contest era that has shown some strength in Games Contests. Dark Souls made Round 3 in 2015, where it nearly broke 30% on A Link to the Past, and managed to double its Round 1 opponent and beat an established franchise in Round 2, while Demon's Souls, the game that started it all...okay it's 0-2, but it looked good in both its losses and both were to quality competition. Probably. They've only had one appearance in a Character Battle and the character was put into a bad situation immediately, but if we ever go back to Character Battles, the series might have enough clout to not completely suck.

Like my bracket.

---

Match 7: The Walking Dead: A Telltale Games Series vs. Bastion

The Walking Dead 10880
Bastion 11907

While Day 2 of the contest was pretty bad for my bracket, it could've been worse. I took three upsets, and one of them paid off! Sadly it was the one that was facing the obvious winner of the eightpack in R2, so I was still screwed out of points in one R2 match no matter who won, but still, it softened the blow of being so far off base on Sekiro.

The Walking Dead was definitely the "safe" choice, because it was both the top option and had contest experience, sort of, while Bastion had none. I say "sort of" because this was both games' Games Contest debut, but The Walking Dead had managed to get two characters into the 2013 Character Battle. Thing is, they both sucked. Spyro vs. Clementine vs. Reyn had the closest Guru spread between the top two characters of any R1 match (though not the lowest prediction percentage for a favorite because there was one match where all three choices were above 20%), and then Spyro just absolutely dunked on Clem, winding up with close to 57% of the vote in a threeway. Lee Everett, meanwhile, was the Guru favorite over Mr. Game & Watch and Meat Boy, because it was assumed that G&W being basically just another avatar for Smash wouldn't be enough and that Meat Boy's ridiculous design would cancel out G&W's silly design and they'd "joke SFF" even though Meat Boy is literally a main protagonist whose name is in the title of his game. Yes it's still ridiculous, but it's not like the "joke characters" who are largely side characters with memes. Game & Watch won anyway, because even foddery Nintendo is still Nintendo and should never be underestimated. I got both of those right then and I got this one right now. The Walking Dead is not the type of game to be trusted to be worth anything in contests, ever. Kind of makes me wonder why I picked Life is Strange but you can chalk that up to Halo being another thing that should never be trusted in contests.

But here's where things get interesting. Whereas in the other seven matches over the first two days, the loser got nearly half of their raw votes from the unregistered users (only Life is Strange managed to even have a 300-vote disparity between registered and unregistered) and the winner received more than 1000 more votes from the registered than from the unregistered, in this match, that trend was reversed. Bastion's lead with the registered voters wasn't even 100 votes, at 4054-3987, but with unregistered users it was 3803-2912! It was largely the loser that benefited from the extra power of the registered bonus, making it closer than it should have been. Interestingly, this was also the only one of the first eight matches that was an upset, with just under 1 in 3 brackets getting it right. Hmmm...the number of registered votes for the first two days was just barely in the low 8000s. The number of brackets filled out, based on the prediction percentages, appears to be...8703! Holy crap, there are more brackets than there are registered voters! That wasn't the case in 2018; though the decrease in vote total is largely on the unregistered side, there were more registered voters in most of the 2018 matches than there have been so far in 2020, while the total number of brackets was just 7221. That's scary, because the general "line" on this site is that despite being largely millennials, we have the "boomer" mentality and are largely stuck in the 1990s. A contest entirely comprised of games from 2010 or later, it wouldn't at all be surprising if almost every voter had at least a few matches where they really didn't know much about the contestants and cared even less. Add in Allen forcing voters to vote in all four matches if they want to vote in any of them, and you get people voting with their brackets. I may have to go back over my Oracle and change every pick where I'm backing a Guru underdog, even though that would've burned me in this particular match, because it really feels like we might be in a situation where predictions and results have a higher correlation than in years past.

Oh, and I guess there's something else funny to be talked about, too. With 4 and a half hours left in the poll, The Walking Dead went on somewhat of a run, winning six straight updates before losing one by a mere two votes, then the two games alternated update wins for 40 minutes except one of Bastion's wins was actually a tie. Since Bastion had only been up by 927 when the run started, it looked as though we were going to see the match go the full 24 hours with the maximum lead under 1000. It would've been just the 47th 24-hour battle to go wire-to-wire within 1000 votes and just the 11th in which there were no lead changes past the first hour (and just the 17th with no lead changes past the first hour even after throwing the 12-hour matches into the mix). But then Bastion started winning updates again, and it looked inevitable that its lead would reach 4 digits. With an hour and 20 minutes left in the match, Bastion led by 990 votes. The next update rolled in, and Bastion won it...by one vote. Next update, same thing, 1-vote win for Bastion. Then came a 2-vote win for The Walking Dead, wiping out the previous two updates. The next two updates were dead stalls, and then TWD won one by 6. 30 of the 80 minutes stalled out, and TWD had refused to let it happen! Bastion spiked an 11-vote win on the next update, however, and cracked 1000 with 35 minutes to go, en route to a final margin of 1027. In the Golden Age of Contests, I'm sure the stats topic would've been losing their shit about "barriers" and TWD's valiant fight to keep the margin in triple digits when its fate was sealed, but it's hard to muster up any sort of enthusiasm when those dead splits were 12-12 and 14-14 and the 11-vote "spike" was 21-10. If two entrants in a 1v1 each managed to get 2/3 of an update within half an hour of each other in the old days, it'd be a sure sign that one of them was cheating. Now, it's just expected statistical variance because the updates tend to be on par with what the fodderiest of fodder was pulling in in the contest's glory days. Remember how lopsided some of the early matches in 2009 were? Wold you believe that the final update that managed to get Crystalis up to 4% even was only two votes fewer than The Walking Dead got on its final update? Or that in the final update of that ridiculous match where SMB3 pulled down over 75% of the total vote, the second-highest last-update total was exactly equal to Bastion's last update? These votes are awful and this contest is awful.

Match 8: Dragon Quest XI: Echoes of an Elusive Age vs. The Legend of Heroes: Trails in the Sky - Second Chapter

DQXI:

Match 9: Mass Effect 2 vs. Resident Evil 7: biohazard

Mass Effect 2 14507
Resident Evil 7 8543

Finally, on Day 3 of the contest we had a match with relatively even proportions between registered and unregistered voters. ME2 scored 63.11% with registered voters and 62.39% with unregistered voters. Which is not all that impressive for a 1-seed! The stats topic was discussing whether disappointing later entries in the Mass Effect series soured the voters on 2, but this seems like the exact opposite of what normally happens. Maybe with casual voters who aren't familiar with either game, the series name plays--in which case Resident Evil 7 was the perfect entrant to take advantage of this, as RE1-4 were all fairly highly regarded--but usually the more entries a series has, the higher rated its oldest entries get. Pokmon is the best example of this, where obviously you have the people who say that everything after Gen 1 (or sometimes Gen 2) is trash, and then you have a backlash where there are others that say that the series didn't even start getting good until Gen 3 and the first two generations are unplayable, but generally speaking the later generations grow in stature each time a new generation is released after it. I was really surprised that the Gen 5 games didn't make it in; I figured that with Gen 8 being out already, they could be really strong, since usually the rule of thumb is that the two most recent generations are "trash" and everything older is at least passable (though I'm not sure if Gen 8 has given Gen 6 this boost yet). Though the only game to get in is actually Gen 4, so yeah. (Good choice, too; I'm personally of the opinion that the second half of Gen 4 was when Game Freak decided to finally bother making their plotlines actually engaging instead of just a throwaway to hang the monster collection on. I guess they sort of put some of the lore into Diamond/Pearl, too, but they just did it so much better in Platinum!)

So this result had everyone calling for ME2's demise as early as next round. Certainly possible, but it seemed a tad premature.

Match 10: Fallout 4 vs. VVVVVV

Fallout 4 16532
VVVVVV 6518

It wasn't just Mass Effect 2's own performance that had people calling for its head, however. While it was busy going roughly 5-3 on RE7, Fallout 4 was going 5-2 on VVVVVV. After the first two days, there was a perception that we'd suddenly turned into IndieFAQs, and this helped prove that we had to look at things on a case-by-case basis. VVVVVV was thought to be another strong indie game, but maybe we should've looked at its protagonist's CBIX showing and second-guessed it. Yes, Characters =/= Games, and VVVVVV is the type of game that I'd expect to be stronger than its protagonist, but Viridian nearly got sextupled by Big Boss. Managed to finish in second place anyway, 53.81% on Peacock, but still. Even if you want to chalk it up to the fact that MGSV was just coming out in 2013 so Big Boss didn't have an old man pic, it's not a great look. And on the topic of releases near contest time, Fallout 4 came out during the 2015 Games Contest and Fallout 3 was already a GotD semifinalist, so it was supposed to be a real threat again in 2015. If only it could have delivered!

---

Match 11: Borderlands 2 vs. Bloodstained: Ritual of the Night

Borderlands 2 13624
Bloodstained: RotN 9427

Borderlands 2 was our GotY in 2012, but it's frequently considered one of our weakest GotY winners ever. And we once had a Call of Duty game win GotY, which in the world at large would certainly be stronger but Call of Duty hasn't been worth anything in contests since, well, around the same time that it won a GotY. Anyone who knows me should know that I'm a huge fan of both music and statistics, so it shouldn't surprise you too much that I frequent WhatNotToSing.com, and the good folks over there have a perfect explanation:

The American entertainment industry is all about the here and now. All major industry awards, be it Oscar or Tony or Emmy or Grammy, tend to be as nearsighted as Mr. Magoo. In Hollywood and New York and Nashville, one year is about as far back as anyone seems to be able to remember. As such, every winter these various acadamies manage to embarrass themselves by bestowing a major award on something or someone that will be thoroughly forgotten five years hence.

That applies to the Internet and its court of public opinion, too. CoD 4 did very well in the 2009 Games Contest, and then Modern Warfare won 2009 GotY the following January. And then...we just kind of forgot about it. It was a fad. Also, how did I never notice that typo in the editorial I just quoted?

But as for Borderlands 2, we saw its weakness. 0-3 in CBIX, with only one of the characters even getting second place. Got upset in Round 1 of BGE3. No characters even made CBX. Can we go back to CBIX again, where one of them managed to allow a non-Symphonia Tales character into second place? Okay yes that was the resident "joke entrant", but still. And while it had the greater name recognition here, it's not like its opponent didn't have the pedigree. Bloodstained is a spiritual successor to the Castlevania series, produced by the series' long-time producer, Koji Igarashi. The similarity of its subtitle to that of the most popular Castlevania game is no coincidence. And remember how much this site loves Japanese things. This performance was very encouraging for Bloodstained, and if we ever get another Character Battle, I wouldn't be surprised if its protagonist was able to do okay. Not great, of course, because very few characters to debut during the Contest era have been worth much (Amaterasu, Altair even though he missed the last one entirely, Sora if you want to count by NA release dates but his game was already out in Japan by the start of Character Battle I). But decent enough for a newcomer. (And no, Geralt doesn't count; Witcher 3 may have been the game to boost him to prominence, but even just as a game character he dates back to the 90s, and across all media he's as old as Link.)

Match 12: Horizon Zero Dawn vs. Fortnite

Horizon Zero Dawn 18655
Fortnite 4397

This match followed neither of the "trends". Whereas in all 8 of the matches on the first two days, the "extra voters" on the registered side largely congregated towards one entrant, and in the other three matches on Day 3, the proportions were very similar, in this match, the raw margins of victory among registered and unregistered voters were very similar: 4772 among the registered and 4714 among the unregistered. Since there are more registered voters than unregistered, that means that this match wasn't even as close as it looked! With the double-counted registered voters, it still came out to a little under 81%; in reality, it was a tad over 82%. Which doesn't sound like a huge difference but you have to remember that the difference in percentage between 1.5x and 2x is larger than the difference in percentage between 3x and 4x. In this case, it's the difference between a little under a 4.25x and comfortably over a 4.5x. Apparently Fortnite is just that hated; I knew it'd lose but I figured its popularity would at least let it not get embarrassed.

Match 13: Resident Evil 2 vs. Danganronpa 2: Goodbye Despair

RE2 19972
DR2 4054

So yeah, remakes were allowed into this contest, though at times it could be hard to tell because there were also a couple of games with the same name as their franchise's debut game, like Doom and God of War. I'm curious as to whether Bacon showed any sort of discretion; if it was just generalized "yeah all remakes are eligible", it's kind of tough to imagine that RE2 got a 1-seed and OoT3D didn't even make the field. Or did the lack of recency and the fact that handheld games are largely viewed as second-class citizens here make people forget entirely that it was eligible? No, that doesn't make sense; HeartGold/SoulSilver made it in and that's even older. (Then again, it's also a much bigger improvement on its original game!)

At a glance, it appears that this was another case where almost all of the "extra" votes went in one direction. Then you realize that in a blowout this big, the proportion isn't all that far off! The official percentage is 83.13%; the raw votes have it at 83% even. Either way, it's close to a quintupling, which either means that RE2 is an absolute monster or that even by the standards of VNs, Danganronpa is piss-weak. In 2015, four VNs made it in, two each from two series. The original Phoenix Wright, a 16-seed just like DR2 here, narrowly avoided a doubling from Skyrim, while the third PW game even more narrowly avoided a tripling from Symphony of the Night in a 2-15 matchup. In a pair of other 2-15 matchups, 999 got 22.2% on Persona 4, so roughly halfway between a tripling and a quadrupling, while its sequel, VLR, got a little under 21% against Mass Effect 2. Now, we know that Phoenix has pretty much gone mainstream, but I'd expect the Zero Escape series to be roughly on par with Danganronpa. The original, not-remade RE2 was also in 2015's contest, where it lost in somewhat respectable fashion to Final Fantasy Tactics. In fact, let's pull up the X-Stat calculator to see what the expectation would be:

Resident Evil 2 (2015g) VS Virtues Last Reward (2015g)

Resident Evil 2 has a strength of 23.93.
Virtues Last Reward has a strength of 10.70.

Resident Evil 2 wins with 77.64% of the vote!
A win of 33,656 with 60,875 total votes cast.

77.64%. Not even what Persona 4 got on the stronger 999. Either the RE2 Remake was just that good (not entirely inconceivable since you're dealing with updating a game from the late 90s to the modern HD graphics and controls), or Danganronpa isn't even remotely in the same conversation as Zero Escape. Probably some combination of both.

---

Danganronpa is definitely stronger than ZE, though their fanbases overlap a lot.

---

Match 14: DOOM vs. INSIDE

DOOM 20674
INSIDE 3349

THIS MATCH IS IN ALL CAPS AND IT STILL HAS FEWER CAPITAL LETTERS THAN THE DQXI MATCH. WHAT IT DOESN'T HAVE IS VOTERS FOR INSIDE. EVEN BOTW DIDN'T SCORE THIS BIG A BLOWOUT.

...Yeah that's going to get annoying real fast. What's there to say, though? INSIDE narrowly took away the title of "fewest registered votes received" from The Outer Worlds, though it at least managed more unregistered votes than Fortnite. DOOM did have the most unregistered votes received, though, with RE2 second and Bloodborne fourth, behind Horizon Zero Dawn but ahead of Breath of the Wild. So it's clear that vote totals are starting to climb. On both sides; The Outer Worlds only beat INSIDE among registered voters by 5, but DOOM beat BotW with them by 449. I guess the repeated delays messed people up, so day 1 got bad votals? Literally 3 of today's 4 winners had a higher vote total than BotW did, both raw and adjusted.

Match 15: Fire Emblem Awakening vs. Deus Ex: Human Revolution

FE:A 15079
Deus Ex: HR 8948

This can only be considered a disappointing performance by FE:A. This was the game that really brought Fire Emblem into the mainstream, much to the chagrin of longtime fans who thought that IS had taken all the challenge out of it. They didn't really...sort of. Yes, they made it more accessible than it was in the Super Famicom era, but then again just about anything would be more accessible than Thracia 776. Awakening was meant to be a celebration of the series' history, and the mechanic that actually got the most flak came from Genealogy of the Holy War!

Well, no, that's not true. Second Seals really were absolutely broken, and combined with DLC that allowed for infinite grinding, Awakening really shouldn't give you much challenge unless you want it to. But the elitists were running challenge runs on the older games, so that's really nothing new. The "anyone can be paired with anyone (with exceptions)" mechanic, along with the children characters, is indeed borrowed from GotHW, and I'd say that Awakening pulls it off just as well as Genealogy did! Fates, not so much, but Fates isn't in this contest, because Fates was IS seeing dollars signs and trying to cater to their vast new fanbase of shippers, while also trying to placate their core fans by making the infinite grinding DLC less broken (and in the process, very nearly defeating the purpose of having it--making the DLC scale in level by your story progress just means that if you want to get all the children characters to their max potential, you basically have to kill all your momentum by grinding everyone up after every level.) Fates was a cash grab, pure and simple. I'm glad it's not in this contest. I mean, Awakening was also, but you basically needed the earlier DLCs to be able to beat Apotheosis. I've heard rumors of people handling it without Limit Breaker, but why would you want to?

But yeah, Deus Ex: Human Revolution was never going to be worth much of anything. The first Deus Ex has done very little in every contest it or its main character has been in, with its only win coming against Rock Band 2 and by an extremely slim margin, and Human Revolution's protagonist made the 2013 contest and...barely managed to come within 10000 votes of Vivi while failing to double a Touhou character (and that's after she had fraudulent votes removed; if they'd been allowed to stay, Adam barely even gets second place).

Match 16: Bloodborne vs. Call of Duty: Black Ops

Bloodborne 19353
Call of Duty: Blops 4679

So remember back in the Borderlands 2 write-up, how I said that Call of Duty was popular here for a very short while and then it just kind of fell off a cliff? Yeah, it really tanked. Of course, even in 2015, the X-Stats say that Bloodborne was good for 63.25% on CoD4.

2009 CoD4 is allegedly good for 69.79% on 2015 CoD4, though, so there you go, that's a huge decline. I keep expecting Bloodborne to be weaker than it is because all of these new games have titles that sound like something an emo teenager would dream up. I can hardly keep track. Bloodborne's another Hidetaka Miyazaki production, though, so I'm not surprised that it's decent. Is FromSoftware going to become this generation's Squaresoft? It seems odd to think that way given that they've been around since the mid-90s themselves, but they broke out in 2009 with the release of Demon's Souls and they've had one hell of a good decade. Given how fed-up gamers are with so many of the big names of the past--Capcom, Konami, and even Squeenix themselves--there's certainly an opening!

---

Match 17: Super Smash Bros. Ultimate vs. Tekken 7

Ultimate 19843
Tekken 7 4071

The expectation here was that Ultimate would stomp all over Tekken 7, which had never been higher than the #4 fighting game series on this site--#5 if you count Smash itself--and possibly even lower, though it would be a bit hard to gauge. Street Fighter and Mortal Kombat were obviously 1-2, and Tekken's sister series Soul Calibur was likely #3 as evidenced by Nightmare's repeated decent showings in character battles. The only Tekken character to ever win a match was Yoshimitsu...who is also a Soul Calibur character. But in the early contests, there were a lot more fighting game characters, and some of them managed to make Round 2, most famously when Gordon Freeman barely broke 40% on Tina Armstrong. And, well, in a vacuum this was a stomping--nearly a quintupling. But this match came one day after both Resident Evil 2 and DOOM managed higher percentages, with the latter breaking 86%. So it seems strange to say that a game scoring almost 83% of the vote disappointed, but that was the perception, and the board eagerly awaited Super Mario Odyssey's performance against Mortal Kombat 11 to gauge their expectations for a potential Divisional Final.

Match 18: The Legend of Zelda: A Link Between Worlds vs. Bayonetta

ALBW 15181
Bayonetta 8727

If there was any more need of evidence of a paradigm shift on this site, this was it. Bayonetta may have been an XBox 360/PS3 game when it first came out, but the character was largely considered pseudo-Nintendo after the sequel was a Wii U exclusive and the original was ported to Wii U at the same time, and then Bayonetta was added to Smash 4 back when third-party characters were still relatively rare in Smash. Given the general opinion of the Wii U, you might even consider the possibility that 3DS would trump Wii U despite the longtime mental block this site has had about handhelds (see: Link's Awakening as a slight Guru underdog to Mega Man X to advance out of R1 behind FFVI in 2009), but there was never any doubt that A Link Between Worlds would win. It's a Zelda game, after all, and one that got GotY in 2013, though it was far from dominating in doing so, only narrowly squeaking past Pokmon X/Y for 3DS game of the year and then getting LFF'd in the poll of all winners not only by the Wii U winner, Super Mario 3D World, but also by the fact that the winner of the "Download-Only" category was a fellow 3DS game, Phoenix Wright: Ace Attorney: Dual Destinies. This led to it narrowly finishing behind GTA V for third place in the poll of all winners, comfortably behind The Last of Us, before powering ahead in the final poll of the top three finishers to take the top spot. Because while it's clearly its own game, A Link Between Worlds will feel very familiar to anyone who's played A Link to the Past, and given how nostalgia-driven this site has been in the past, especially for Super Nintendo games, that should've been a very good thing.

And while A Link Between Worlds did win comfortably, it wasn't even a doubling, which given the way the contest had gone so far was a very weak win indeed. Over half the matches through five days had ended with a doubling or worse! And this wasn't even an action/RPG hybrid like the games that had been doing well on previous days. This was a full-fledged hack-and-slash that was avoiding the doubling by, essentially, A Link to the Past 2. While it was obvious by its absence from 2015 that ALBW was categorically not held in the same regard as its spiritual precursor, this was an embarrassing performance that, of course, would be completely meaningless because it was expected to be hopelessly outmatched in Round 2 anyway.

Of course, if anyone had been thinking quickly enough, they would've realized that this was a major hint as to the outcome of the most debated match of R1, which was due up the very next day. But who reacts that quickly, especially when the actual match outcomes were all so obvious that it was fairly safe to ignore the match once you'd voted in it?

Match 19: Marvel's Spider-Man vs. Dead Space 2

Spider-Man 16717
Dead Space 2 7190

There is one debate that rages almost constantly on Board 8, and that is how characters from other media would fare in Character Battles if we allowed them in. The general consensus is that the most well-known characters would be very strong indeed, though there would of course be plenty of fodder. Perhaps the biggest issue is that by opening it up too far, you'd have such a large list of potential entrants that it would be absolute chaos. My solution, of course, would be to limit it to characters who have been in a "good" video game, since merely forcing them to have been in a video game doesn't really limit it much. Nearly everything has had licensed games these days. So how would we determine what's "good"? Why, how about "things that have been in Games Contests"?! Of course, we'd probably have to have Games Contests a bit more frequently, but as long as Allen's still gatekeeping to make sure only quality games get in, that'll do fine. Right off the bat you get a bunch of Disney characters via Kingdom Hearts, and a bunch of Marvel characters via MvC2. DC Comics doesn't quite get as good a deal as Marvel does, but the Arkham series has been a mainstay in these contests. The silly nomination system for the first BGE also allowed the Simpsons arcade game to get in, so that's a few more characters with potential.

But there are people who figure that since this is a gaming site, anything non-gaming would get eviscerated by even a weak video game character. These people are wrong, but occasionally there are results that continue to fuel their belief. The Walking Dead getting upset by Bastion was one of them. Another one that was considered a major upset ironically came in a Character Battle, where people thought that having "Darth" in his name and being the top option would allow Revan (a KotOR-original character, therefore eligible for contests) to beat Terra Branford. Yeah, I don't understand why that's relevant either. The Star Wars Extended Universe is massive, including those video game-original characters. The fact that the Star Wars name alone doesn't make you strong doesn't mean that Han Solo or Darth Vader wouldn't be strong. We've seen that even in our top video game franchises, when we rallied Groose into the 2013 contest and he fell flat.

Spider-Man fell just short of 70% against Dead Space 2, while GTA: San Andreas was only able to get 65.45% on the original Dead Space in the previous GotD. Some food for thought. Gamers are not so single-minded that they ignore anything that isn't related to video games.

---

Match 20: Minecraft vs. DotA 2

Minecraft 18100
DotA 2 5806

This was just...why?

The outcome of this match was never going to be in doubt. I'm fairly certain that the Guest picker on the Crew only picked DotA 2 just to make sure there was no Crew Curse, because he, like all but one other Guru, had Minecraft in this match. But I'm going to quote the beginning of it anyway, because before it gets into the crazy logic, it makes a good point.

Neither of these genres are particularly liked by GameFAQs. Its strange though, because both of these games are popular beyond belief.

True, although I wouldn't necessarily call it strange. While in the previous write-up, I defended the overall gamer community as not being a bunch of elitists who don't care about anything non-video game, here I'm going to have to turn around and say that old-school gamers, like the majority of GameFAQs membership, are proud of the fact that they loved gaming before it was "cool" and hate anything with mainstream popularity. In fact I already talked about this exact same phenomenon on a single-fanbase level in my Fire Emblem: Awakening write-up. So while it was obvious that DotA 2 was going to get smashed just for being a MOBA, even if it wasn't the one that spawned the hated 2013 champion, it remained to be seen how high a percentage Minecraft of all things would be able to get. And the answer is...not too high, but pretty high. DotA 2 got a better percentage than Fortnite did, at any rate.

Match 21: Undertale vs. Octopath Traveler

Undertale 12087
Octopath 13790

Oh hey speaking of hated champions.

The vote total for this match was the highest of the contest so far, but not by an overwhelming amount. And I have to say, this was not an outcome I expected! I figured that there was a chance that Undertale got anti-voted into oblivion the way Draven did in 2018 (and Sans himself to a lesser extent), but that if that didn't happen, Undertale probably had the natural strength to win its fourpack. Instead...this happened. The match actually got off to a slightly delayed start, for no apparent reason, and when it did, Undertale got off to a quick start, before fading back to closeness. By the freeze, Undertale was up by 5 votes, and on the next update, it raised that lead to 7. And then, bam, 41 vote swing, Octopath up by 34, and it never looked back.

And it wasn't the registered voters killing it, either, although they were likewise motivated to come out in higher numbers. With the registered voters, Octopath won 4779-4436. On any previous day, 4436 would've easily been a majority of registered voters--hell, Bastion was at 4054! It was the unregistered voters who buried it, 4245-3229. Also the most unregistered voters of the contest so far, but it was enough of a blowout that 3229 would not, in fact, constitute half of any day's unregistered voters. (It would've come close on Day 1, where unregistered voters numbered in the 6500s.) We did get confirmation that all the usual spots from 2015 weren't allowing rally topics, which begs the question, where did these extra votes come from?

Also, the mythical "casuals" didn't see this coming either. 21.29% prediction percentage in Round 1 was the fifth-lowest of all time, trailing only a pair of lolChief results (Yuna in 2013 at #1 and Sub-Zero in 2010 at #4) and a pair of 2009 matches (in)famous for very different reasons: the Oblivion/Symphonia/God of War/GTA:SA fatal fourway at #2, which had the smallest spread between first and fourth of any fourway and is remembered fondly for it, and the Super Mario Bros 3 match that had the largest spread between first and second of any fourway and is laughed about as a result at #3.

Match 22: Shovel Knight vs. Dragon's Dogma

Shovel Knight 16664
Dragon's Dogma 9211

While you'd never know it by my picks this year, I've actually gotten off to some pretty good starts in Guru before, even holding sole possession of first place in the Guru after correctly calling every match in Division 5 (two outright upsets and two others where the favorite had less than 60% of the brackets). That ended with Shovel Knight failing to be the indie darling I'd expected it to be. It's commonly accepted that for a rallied entrant to get moving, they need to have weak early competition to get past, because something too strong will kill it before it starts. L-Block didn't have to get past anyone stronger than Laharl in Round 1 and got a favorable LFF situation in Round 2, and Draven only needed to get past Jak and Chie Satonaka. Hell, L-Block and Draven were both Guru favorites to reach Round 2! But only that far. And Undertale got the polarizing Mass Effect 3, just weak enough for its rallies to work. Shovel Knight, in Round 1 of the 2015 Contest, was set to face Grand Theft Auto: San Andreas. Betting against GTA is never terribly risky on GameFAQs, so this was certainly a possible upset. And waiting in Round 2? The casuals would no doubt think it to be the massively popular World of Warcraft, as evidenced by prediction percentages in previous contests, but of course we all know that WoW is trash here, and so the Guru consensus (if you can call a roughly 5-3 ratio "consensus") was that it would be Chrono Cross, which is much maligned for being an alleged sequel to Chrono Trigger and not being even remotely as good a game. Either way, seems like a perfect target to start a bandwagon. If I'd really been thinking about it, I probably would've considered that enough time for the bandwagon to be fully rolling and picked more upsets based on the rallyFEAR effect, but mercifully I didn't, because while Shovel Knight did respectably, it never came close to beating GTA:SA. Come 2018, however, it seems that the casuals don't learn, because they must've seen 38.72% on GTA:SA and said "this game's really strong!" Shovel Knight the character was, astonishingly, comfortably favored in Round 1 against Captain Toad; the Gurus weren't so easily fooled.

So when Shovel Knight came out strong against Dragon's Dogma, people were suddenly amazed. It was favored, of course, but it was still considered a strong performance, and the masses who picked Undertale for two rounds were feeling even worse because Shovel Knight to R3 wasn't nearly as unpopular a pick as it would be if Octopath actually got there.

---

Match 23: Devil May Cry 5 vs. Donkey Kong: Tropical Freeze

DMC5 14856
DKTF 11020

Oy. Just when Donkey Kong had finally shed the label of being a choker, this happens. Though in fairness, he was never really in this match, unlike so many of his famous choke jobs where he threw away a win. It took about 2.5 hours for him to win his first update, and the match was more than a third of the way over before he won one by more than 5 votes. And while it doesn't make the list of most embarrassing results for the Gurus, it's pretty embarrassing. The Gurus had Tropical Freeze as a slight favorite, while the casuals pegged this match correctly at nearly a 70% rate. The reasoning was simple enough: DKC games had performed decently enough in contests when given half a chance (which is to say, in 2015 alone, because the voting system for 2004 meant there was no way one would make the field over fellow SNES classics like ALttP, SMW, Super Metroid, FFVI, and CT, and 2009 both DKC games just got fed into bad SFF situations--and one of them still managed to reach R2!), while Devil May Cry...uh...yeah. Devil May Cry still has yet to get a game into an "open" Games Contest. No representation in 2004, none in 2009, and none in 2015. DMC1 made the field in GotD out of the vote-ins, but unlike in the Character Battle earlier that year, the vote-in winners performed like their low seeds suggested they should. DMC3 was a bit higher up on the seeding list, and it...failed to double Mother 3 in Round 1 before a wholly expected R2 loss to Metroid Prime. Now, let's put things into perspective here: Mother 3 has never been released outside of Japan. The theme for Round 1 pics was "box art", which is perfectly fair and balanced, but it inconveniences Mother 3 because its box is just the name of the game on a solid red background. So, huge pic advantage, huge playrate advantage, only 64.86%. Wretched! Its main character has done fairly well in contests, but mostly in that he's put up good numbers in losses and wins over fodder. The strongest character he's actually beaten in a 1v1, based on strength at the time, is probably Ryu Hayabusa in 2006, though based on current strength, it's his 2004 win over Tails. In multi-ways it's another story; he made the semifinals of both of the fourway contests, and, well, his 2013 wasn't exactly inspiring but at least in his exit he managed to beat out a decent midcarder for second place. But 2007 and 2008 being by far your two best contests is far from encouraging.

So what changed? The site's opinion of action games, obviously. Devil May Cry's creator went on to create Bayonetta and the two series have very similar combat systems, so after seeing Bayonetta overperform against A Link Between Worlds the previous day, this shouldn't have come as a huge surprise. Apparently this is GameFAQs' type of game now. Go figure.

Match 24: Super Mario Odyssey vs. Mortal Kombat 11

Odyssey 22245
MK11 3631

While I'd advise against putting too much stock into this result, as the Undertale match raised the total vote count beyond what it had been in previous days, Mortal Kombat has always been a stronger series than Tekken, so for Odyssey to win this decisively was reason to question whether Smash Ultimate was really a lock against Odyssey in Round 4. I've always been of the opinion that there's little to be gleaned from the size of a blowout, however...at least, not for how foddery the loser is. If you get 8% or 12%, there's really not that much difference; it's just a measure of how strong your opponent is. But in that context, from the perspective of who wins those blowouts, it does say something, so Odyssey is probably the favorite right now!

---

Dante got last in the quarterfinals in '08.

---

LinkMarioSamus posted...

Dante got last in the quarterfinals in '08.

Oh right. Got the rounds off by one. Eh, I'll fix it on the wiki.

---

Normally the board moves faster during contests, but this place is so dead that I have my doubts.

---

Oof. I'd been trying to stay within a couple of days but I've been kind of busy lately. I'll try to catch it up tomorrow.

---

Or maybe I'll get a little in tonight.

Match 25: Grand Theft Auto V vs. Baba Is You

GTAV 19151
Baba Is You 5008

Baba is WHO? Seriously, this is exactly what a 1-16 match is supposed to look like, more or less. It's not a massive blowout by this contest's standards but that's probably largely GTA V anti-votes. There's absolutely nothing in this performance that says one way or the other whether GTA makes it to Round 3 or not.

Match 26: Cuphead vs. XCOM 2

Cuphead 16493
XCOM 2 7662

Wait, "XCOM 2" is really the full name? No subtitle? No extended name that XCOM is short for? Geez, no wonder it sucks. This result was not really that surprising. Cuphead's only previous outing was in a Character Battle, but the titular character only just squeaked into CBX with a 16-seed, and the one placed against the presumed #1 overall seed based on match placement. Indie darling though he may be, that was always going to be a losing battle. GTA V is probably the favorite for Round 2 based on playrate alone, but then again, GameFAQs has always tended to reject the mainstream in favor of niche stuff, so a game that's so retro that its aesthetic pre-dates video games by multiple decades might be right up their alley.

---

Match 27: Persona 4 Golden vs. FTL: Faster Than Light

P4G 18785
FTL 5371

Well, this just kind of proves the maxim that it's better to be an old game than a new one. Granted, FTL has always been extremely niche, but the fact of the matter is that arguably the most impressive win of the day came by the remake of an older game. At least I think it was the most impressive? Maybe FTL's so niche that it actually would lose to Ghost Trick?

Either way, the upcoming P4 Golden-RDR match was looking like a 6-point match. It was highly unlikely that merely getting similar numbers to the two of them in Round 1 was enough for GTA V to be on their level, given how bad Baba Is You was expected to have been.

Match 28: Red Dead Redemption vs. Ghost Trick: Phantom Detective

RDR 2 19051
Ghost Trick 5105

Ghost Trick is basically like a spiritual...cousin, I guess you'd say, to Ace Attorney. Same creator, but it's a stand-alone game that doesn't fit into their continuity. It is therefore expected to be weaker than them, not that it really matters because no Ace Attorney games actually made this bracket.

This is roughly a good percentage for Red Dead Redemption. It's tough to tell who should be favored in Round 2 due to the weak opponents everyone had.

Match 29: Xenoblade Chronicles vs. Splatoon 2

XBC 19671
Splatoon 2 5411

I've heard a lot of speculation that Splatoon 2 would be stronger than Splatoon because it was for the beloved Switch while Splatoon was for the hated Wii U. Was Splatoon really that recent or has the Switch really been around that long? *checks* Bit of both, honestly. Actually I guess that makes sense because I thought the Wii U had a pretty short shelf life. I never ended up getting one because by the time there were actually any games I really cared about, the "NX" had already been announced and I made the mistake of thinking that there'd be backwards compatibility like the Wii U had to the Wii and the Wii to the Gamecube. Luckily the one Wii U game I was really disappointed about missing out on just got ported to the Switch! (That'd be the Fire Emblem/SMT crossover; I love both those series so I was really excited to hear about it.) Then again, was XBC ever really going to be strong? It was a very late-stage Wii game and one that had...availability issues. Well, since Splatoon was in BGE3, let's go to the X-Stat calculator!

Xenoblade Chronicles (2015g) VS Splatoon (2015g)

Xenoblade Chronicles has a strength of 23.73.
Splatoon has a strength of 13.60.

Xenoblade Chronicles wins with 71.34% of the vote!
A win of 25,987 with 60,875 total votes cast.

Just ignore the hilariously outdated projections on votes cast, where even with registered votes counting double, this match had fewer total votes than the projection's margin of victory. Now, granted, this contest has felt like BlowoutFAQs so far, but Splatoon 2 got 21.57% in this match. Splatoon's number for 2015 came from getting 20.93% on Metal Gear Solid. So...probably yet another case of the original being stronger. Being for the Switch might've actually hurt Splatoon 2 because it gets overshadowed by so many other games that the Wii U just didn't have! And given the usual fall-off for sequels...I'd say this feels like a constant enough result for XBC. Also, let me just check something...Okay. I honestly couldn't remember if Shulk was in 2018. Given that he'd had multiple Smashes since 2013 (well, okay, Ultimate came out during CBX, but they'd already announced that everyone was back), him missing the field entirely would've been a huge indictment. But he did make it. Didn't perform all that well, but he was there. Good enough to get by, at least given the competition for Round 2.

Match 30: Overwatch vs. Death Stranding

Overwatch 14022
Death Stranding 11061

Hoo boy. This was going to be the battle of the bad. On the one hand, GameFAQs hates the eSports scene. MOBAs get the brunt of their animosity because of what Draven did in 2013, but all of those types of games with teams and stuff are really not their thing. So the idea of Overwatch being worth anything was laughable. But its opponent was Death Stranding, a highly anticipated, recent game that had disappointed immensely. There was a topic on the board during this match titled "I thought it was popular....", that started thusly (emphasis added by me):

Why, why must you lose Death Stranding. I havent played it, but before it came out people wouldnt shut up about it. Well apparently it sucks.

Yes, yes it does. Prerelease hype is possibly one of the worst predictors of contest strength ever. I hate to go back to the same examples over and over again but Lightning managed 36.14% on Sonic the Hedgehog while FFXIII was only out in Japan, then proceeded to get literally last place in a vote-in for Rivalry Rumble, failed to take advantage of Nintendo SFF to dispatch famed choker Donkey Kong in 2013, and only managed 34.38% on Dante in 2018 (which is admittedly probably more impressive than the loss to an LFFed Donkey Kong, but still looks bad compared to the Sonic match). Meanwhile FFXIII and its many sequels and spinoffs have yet to make a Games Contest, including this one. The fact that FFXIV, an MMORPG, is in this contest and FFXIII isn't is perhaps the biggest indictment of the game imaginable. So even though the idea of Overwatch having strength here was ridiculous, it was still expected to win this match, and it did, with an ease largely facilitated by the registered voters no less. Or maybe they were the ones rallying? This day clearly had a vote spike from the previous one. There hadn't been any talk of rallies, though, so it clearly wasn't evident enough to set off any red flags with the hypersensitive board.

Surprisingly, it was this match, not the big upset that also occurred on this day, that knocked off the last remaining perfect bracket. We know this with 100% certainty because the last remaining perfect was a Guru.

---

Red Dead Redemption 1 was against Ghost Trick. RDR2's match hasn't taken place yet.

---

Match 31: Dragon Age: Inquisition vs. Ori and the Blind Forest

DA:I 12065
OatBF 13024

I knew that Ori was at least somewhat popular among indies, but it was never going to stand a chance against a popular RPG series like Dragon Age...right? Well, let's take a look at the results:

In the first GotD, Dragon Age: Origins piled up 69.21% on F-Zero GX in Round One (Oracle expectation: not quite 61%, with a few rogues picking GX to pull the upset outright) before pulling 31.13% in a loss to Twilight Princess. Now, if taken at face value, the 37.73% that Super Mario Galaxy pulled in on TP the following round suggests that DA:O would be able to break 45% on it, but of course SMG-TP is an SFF match so the adjustments suggest it's more like 41.25%. Actual result when the two games met in Round 1 in 2015: DA:O was held just under 36%. Origins also managed 30.32% in the final of the 2009 GotY, while Inquisition, the game in this match, also made the final vote in its GotY five years later and actually took second place (the 2009 final was close enough that 30.32% was actually third place out of three.) So by all indications, it's a strong series here. Probably. As such, just 17.61% of Gurus took Ori to pull the upset.

"Oops."

The casuals didn't exactly embarrass us on this one, but they clearly weren't as enamored of DA:I as we were; 37.68% of brackets had this match right. DA:I never led in this match, with the max lead for Ori coming about 3 hours from the end. But max leads are fairly meaningless in this contest, because with how low the votals are, there's barely any movement late in matches.

Match 32: Fire Emblem: Three Houses vs. South Park: The Stick of Truth

FE3H 16721
South Park 8364

As if we needed more evidence that non-video game origins are no object for GameFAQs voters.

This was a horrible result for Three Houses, which was seemingly a lot more well-loved in the fandom than Awakening or Fates even though it's probably even easier than those games. Seriously, this game still has a "Casual Mode" despite the fact that Divine Pulse makes it so that you can take the risks in Classic Mode that you wouldn't dream of taking in anything other than Casual Mode in any other game. I think I actually had to start tightening up my play at one point in my first playthrough because I was running low on (and eventually ran out of) charges, but I never had to reset to save a unit, and on NG+, you'll never even have to worry about running out of charges at any time. Which is fine for me, because I'm more of a story guy and replay for the completion of filling up the support logs (a fruitless task in the later games, though one I'd managed on both 7 and 8). In Awakening and Fates, that meant one playthrough on Classic Mode to prove myself capable and then all replays on Casual. In Three Houses, it's Classic Mode every time, because there's still pretty much no risk to doing so. And against South Park: The Stick of Truth (which is apparently also an RPG?), it...came up just seven votes shy of the doubling. Yeah, we'll just call it "a doubling", but that it was even close, let alone technically coming up just short, is embarrassing. And in the raw votes, it's not even close; FE3H polled at 67.64% among registered voters and 64.19% among unregistered.

And yet, would that really affect anything? Three Houses was considered a massive favorite over Dragon Age: Inquisition, and it was even harder to imagine Ori and the Blind Forest pulling the upset. Awakening was expected to get 48.23% on Xenoblade based on their matches with DKC2 in 2015, so Three Houses barely has to be stronger than Awakening at all to win that (which it still should be just from being on the Switch, given this site's apparent anti-handheld bias). So, what, we're panicking about its chances against the P4G-RDR winner in Round 4? Were we favoring it there to begin with? *checks*

...Apparently not only is FE3H the Guru favorite there, but GTAV is the favorite over the P4G-RDR winner in Round 3, and there's enough of a split on that R2 match itself that Xenoblade is actually the #3 choice to win this division, with P4G #4 and RDR #5. So either I'm going to be made a fool of, or I'm still in prime position in the Guru after all.

(Unless it's Red Dead Redemption rather than Persona 4 Golden that knocks GTAV off, in which case I'm really screwed.)

You know what, let's settle this once and for all. Back to the X-Stat calculator!

*looks at outcome of hypothetical GTA5-P4 match in 2015*

...Da fuq? Is this some type of weird trickery based on GTA5 being in the SMRPG quarter of the bracket? ...No, P4's considered the slight favorite over RDR, and RDR was in that quarter as well, albeit in Division 8 to GTA5's Division 7. Well now I feel silly, but come on, it's GTA. When has that ever been worth anything in Games Contests?

Match 33: The Witcher 3: Wild Hunt vs. Assassin's Creed Odyssey

Witcher 3 20059
ACO 3801

Way back when Assassin's Creed was a new series, it was a legitimate contender for being the rare new property that GameFAQs actually embraced. In 2008, Altair rode the Guru nomination to a highly advantageous bracket placement (the fourways didn't have proper seeds, but we're assuming he was like a 2-seed or something) and won his opening fourpack outright and advanced in second place out of Round 2. Come 2010, Assassin's Creed already had one sequel, and both Altair and Ezio got 2-seeds and made Round 2, where they ran into some very strong characters. Both those games made it into Game of the Decade, though the original needed to make it through a vote-in and got knocked out by MGS4 in Round 1; II got a 6-seed and reached R2 before losing to Wind Waker. Ezio also managed to have enough strength to get into Rivalry Rumble in 2011, somehow. He and Rodrigo Borgia were a 6-seed, which was the lowest seed possible without going through vote-ins; I have a sneaking suspicion that they wouldn't have fared so well in vote-ins. Come 2013, with the sequels piling up...Ezio still had enough popularity to get a 2-seed and had no trouble dispatching the 17 and 26 to reach R2, where he finished last behind L-Block and Auron. Altair, meanwhile, fell all the way to a 25-seed...which was not a bad place to be given that the 3-seed was reserved for "hyped newcomers who will inevitably flop". He took care of Shulk, the 3-seed in question, and Ratchet, then finished ahead of Lara Croft, former 1-seed, for second place behind, uh, Kefka. Yeah no one expected Kefka to be there but we all expected the winner of his R1 match to win in R2. Only in 2018, when Altair missed the field entirely and Ezio got fed to Zelda in Round

Match 34: Mass Effect 3 vs. StarCraft II: Wings of Liberty

ME3 14466
SC2 9392

Oh, hey, sequel fight.

https://board8.fandom.com/wiki/(7)Starcraft_vs_(2)Mass_Effect_2_2015

That was a hotly contested battle, and the general consensus was that neither of these games were as popular as those games, even if the adjusted X-Stats say ME3 > ME2--I have no idea how they came up with that, though, because I think the match against ME3 was the only one where we got to see what Undertale's unrallied strength looked like? Obviously the raw X-Stats have ME3 behind only Undertale, but that'd just be silly.

...Hm. Well that's disturbing. It turns out I remembered wrong, and we did get to see unrallied Undertale for quite a bit of the match against Fallout 3 in Round 2 (in fact, Fallout 3's peak lead came more than 13 hours in, which is especially scary given that it fell more than 2500 short of Undertale's winning margin). And of course, it's fun to note how even though Undertale was continuing to win updates by large raw amounts right up until the end, the total votes in the late updates aren't really any higher than in a lot of the early updates where FO3 was winning (though given how the updates usually get quite a bit smaller, the fact that they're of similar size is still the rallies' fault). Kind of lends some credence to what Allen said about the Link-Draven match, that people could always come back and vote later. But what's disturbing is that there, uh...was kind of a blip in that poll, too. Nothing prolonged like what happened in Link-Draven, but if you look at the updater, there's a single update where Fallout 3 won 4-0. For a second I thought that the updater had accidentally taken a double, since the seconds on the two updates are different, but it's actually the later update that has the earlier seconds reading--they're 4:59 apart, not 5:01.

...Or maybe not. It took a second update a mere minute later and that showed a more normal amount of votes. But then why did it show any change in votes at all? Strange. Must've just been a hiccup in the counting software.

But back on topic. Presumably based on the pre-rallying ME3-UT and FO3-UT values, ME3 was in fact supposed to get 51.92% on ME2 based on 2015. ME2, in turn, got 51.12% on StarCraft in a match where both sides were rallying. So this was a fairly obvious result and I feel silly for thinking it could be close even though I got it right.

Match 35: Stardew Valley vs. Destiny

Stardew Valley 16979
Destiny 6879

Another match that everyone seemed to know the winner, but no one really felt comfortable about it. Stardew Valley was yet another indie game, and even though it is technically an RPG, the game series it's most commonly compared to is Harvest Moon. Actual Harvest Moon, not its RPG spinoff Rune Factory. Destiny, however, is neither Japanese nor an RPG. It is in fact made by Bungie, best known for notorious contest disappointment Halo. Destiny had been in a contest before, in 2015, where it lost in Round 1 getting 25.51% on...StarCraft. (Oracle expectation: 39.85%) Well, that certainly provides a good comparison for Round 3! Based on the 2015 X-Stats, however they were calculated, ME3 would be expected to get 76.02% on Destiny. Stardew Valley got 71.17%. So that comes out to...Stardew Valley being worth 48.35% on ME3? That seems awfully strange, but okay!

(Far more likely that Destiny has fallen off a lot more than ME3 has; I wouldn't even really be surprised if ME3 had been vindicated by time somewhat, either because those who didn't get it right away only know the revised version or because the even more hated Mass Effect: Andromeda has lessened the hatred towards 3.)

Match 36: Super Mario Galaxy 2 vs. Return of the Obra Dinn

SMG2 20816
RotOD 3043

Just when you thought the vote totals couldn't get any worse...

I'm fairly certain that that's a record low for votes in a 1v1 Games or Characters contest, even with the registered voter bonus. The Years Contest had a lot of first-round blowouts where even the raw figures for Return of the Obra Dinn were higher than the loser's vote total, but it's certainly worse than anything else this year or from 2018, including Draven getting hatestomped. Though at least the unregistered voters alone gave it more votes than Chester got in 2013. Honestly once you add in the registered vote bonus this isn't all that far off Chester's total for the end-of-round-1 "Worst of the Worst" bonus match.

It's impossible not to look at the raw totals for this match and not think of the Years Contest, though, because Return of the Obra Dinn literally got 1998 votes before adding in the registered voter bonus.

So now we were left to determine whether 87.25% on Return of the Obra Dinn was as impressive as 84.07% on Assassin's Creed: Odyssey. It doesn't seem like it should be, even if it was the biggest blowout of the contest so far by percentage. The biggest by total votes, though, was Super Mario Odyssey, so Galaxy 2 doing this as well...well, it was easy to think that Mario was just going to be a beast this contest. Hopefully Round 2 would give us a better handle on what to expect for the big SMG2-W3 matchup.

Or, wait, is that even right? That was the foregone conclusion, but I can't help feeling that there's something in the way... After all, SMG2 narrowly lost to Mass Effect 1 in 2015. If The Witcher 3-Super Mario Galaxy 2 is supposed to be a close match, shouldn't Witcher 3 have trouble with ME3?

...Huh. The Witcher 3 did better in defeat than I remembered. Still, it does seem odd that The Witcher 3 is a unanimous Guru pick to reach R3 when the 2015 X-Stats say that it's only expected to beat Mass Effect 3 with 51.74%. What's SMG2's expected percentage against ME3?

...51.58%. Holy crap. The X-Stats have The Witcher 3 winning with just 50.16% of the vote against Super Mario Galaxy 2. There are so many things that can change in five years. And yet Witcher 3 is more than a 5-to-1 favorite on the Guru (along with a sole Guru taking Stardew Valley to upset both of them--one of only three to even take Stardew Valley to upset SMG2.)

Bring it on.

---

Match 37: Fallout: New Vegas vs. The Stanley Parable

Fallout: NV 18217
Stanley 4726

Having 4 matches per day is great, because it allows the boring Round 1 matches to be cleared out of the way quickly. And this contest had no shortage of those. Through ten days of the contest, there had been a grand total of two lead changes after the freeze. Undertale managed to make it to the freeze with a 5 vote lead and increased it to 7 at the next update before falling behind for good, and Walking Dead came out so on fire that its max recorded lead was before the freeze (64, compared to the 55 it had at the freeze) and managed slight gains over the next two updates before Bastion took the lead for good at the 40-minute mark. That's it. Even Ori and the Blind Forest, which had a smaller maximum lead than Bastion, never trailed. Honestly, in the absence of rallies, we're probably going to see this in every match.

Fallout: New Vegas was considered by many to be a disappointment, yet it was still better than a 3 to 1 favorite among Gurus for Round 2. Given what we'd seen all contest, this was probably cause for concern.

Match 38: Dark Souls III vs. Dishonored

Dark Souls III 16041
Dishonored 6897

We'd seen Sekiro do well, and Bayonetta overperform on ALBW, and DMC5 "upset" DKC:TF. And now, we had an actual Dark Souls game. And it did...quite well! It doesn't look that way based on percentages, but with votals so low, the difference between winning with 69.93% and 79.40% is only a little more than 2000 votes. And to me, almost 70% on Dishonored does in fact feel more impressive than 79.4% on The Stanley Parable. The Stanley Parable is described as an nteractive drama and walking simulator". In other words, it should be utter fodder. Failing to break 80% on it is not great, though coming that close isn't bad, either. FO:NV is a game that was in 2015, where it got almost 56% on BioShock Infinite in R1 and almost 42% on RE4 in R2. So, not bad.

But let's look at Dishonored. We don't have any Games Contest results to work from, but in GotY for 2012, when Allen decided to switch from polling by system to polling by genre, it came in second place in "Action/Adventure Game", still well behind poll winner Assassin's Creed III, but well ahead of third place and in fact exactly doubling up 4th place (in a 10-option poll). We also do have a Character Battle result to look at! Corvo Attano, the main character of Dishonored, made CBIX as a 15-seed. He was blown out by 4-seed Pokmon Trainer Red, but nevertheless managed to just barely double 24-seed Sissel, the protagonist of another game in this bracket, Ghost Trick: Phantom Detective. Now, there's no good way to figure out the relationship between character battle strength and game strength, but a doubling is a doubling even if it's, to round off, about a 70-20-10 threeway. This is a good performance by Dark Souls III, and it wouldn't surprise me at all to see it pull the upset next round.

Match 39: Binding of Isaac: Rebirth vs. Nioh

BoI:R 9972
Nioh 12968

Nioh was a contest newcomer, an action RPG (good thing for this contest) made by Team Ninja (maybe not so good? This is a very Nintendo-centric board and a lot of Nintendo fans are probably still mad about Metroid: Other M) with a sequel (actually prequel, but its name is "Nioh 2") that was released not even two weeks before this match (always a good thing, but imagine how high it could've gone if the start of the contest hadn't been delayed!). Binding of Isaac: Rebirth had made the 2015 Contest as a 14-seed and proceeded to come up against one of the few things against which the "older = better" would've actually worked in its favor, which along with the fact that it was from a series I'd actually heard of, I initially entertained it as a possibility for our requisite 14-3 upset (3-seeds always tend to be hyped newcomers that inevitably flop; Allen even made a joke out of it in CBIX and of course a 3-seed won it all). Mercifully I was successfully talked out of that; The Witcher 3 nearly tripled it. But "familiar" usually trumps "unfamiliar", so even though the only contest experience BoI had was being weak (the protagonist also made CBIX as an 18-seed, but that meant an immediate date with a Noble Niner), the fact that it had any experience at all made it a favorite on Board 8, at a little better than 2-1 on the Guru and 5-1 on Oracle predictions, including at least one person who said straight-out in the Oracle Discussion Topic that they had Nioh in their bracket but their bracket was trash anyway so they were hoping that it was their Oracle pick of BoI:R that was correct. The casuals? Well, they were upset, too, but not nearly as badly as the Gurus. 43.53% of brackets had this match correct, so while they too were ultimately wrong more than they were right, they correctly saw this as a debatable match.

And let's face it, winning with 56.53% qualifies as "close" in this contest. I'm honestly not even sure why I'm still making Oracle predictions under 55%; I suspect it's to mitigate disaster when I know I'm picking against the majority.

Match 40: God of War vs. The Talos Principle

God of War 19480
The Talos Principle 3460

I really don't want to talk about this match.

No, seriously, I really don't want to talk about this.

I got this right in Guru, of course. Everyone did. And at 94.81%, it was second only to Breath of the Wild in prediction percentage among the casuals, too. Was I under some delusion that God of War was no longer popular here? Did I accidentally mistake the Talos Principle for something that might have anything remotely resembling strength? Or was I just not paying attention?! Because honestly, there is no good excuse for being more than 10 points behind the second-worst Oracle prediction when you're not predicting an upset or even a threat of one. This match has single-handedly convinced me to start keeping tabs on what I put into Oracle for every match and adjusting them if I think I messed up. Predicting 61% for a match where the actual winning percentage was nearly 85% can do that to a guy.

Even without the benefit of being an established franchise, God of War's genre is much more of an asset this contest than it seems like it would've been in years past. That was just such a huge blunder on my part that I actually went to check my Guru to see if I'd screwed anything up in later rounds. Then I remembered what its next opponent was/would have been. Yes, it turns out, I am in the majority that picked God of War to make Round 4 and then lose. I just had a major stupid when it came to

Match 41: Shin Megami Tensei: Persona 5 vs. Heavy Rain

Persona 5 19455
Heavy Rain 5395

So of course, as I so frequently do, I immediately followed up my colossal blunder with an outright win. Only went over by 0.02%, and unlike some of my previous wins where I just had the high/low pick on a match we all messed up, there was some tight competition. In fact, the Oracle consensus was only 0.08% off from what actually happened--the top scores were the ones closest to the median!

Also, despite having literally filled out a bracket, it was at the very least when I was making Oracle predictions and possibly even on the match day itself when I went "wait, how is Heavy Rain in this contest? Didn't that game come out in like 2008 or something?" Nope, but I understand my confusion. Exact years for things in the past often elude me, so I probably remembered seeing ads for it on TV in my college dorm room and figured "yeah, that's late 2000s"; it actually came out in February of 2010 for the Playstation 3. It was then ported to the PS4 in 2016...and to Windows in 2019. So, yeah, recent rerelease, makes sense that it somehow managed to squeak out a bid. Got crushed, of course, because Persona is mainstream now even if SMT as a whole still isn't.

Match 42: Assassin's Creed IV: Black Flag vs. Sid Meier's Civilization V

AC IV 14026
Civ V 10824

We were just two days removed from seeing Assassin's Creed Odyssey get quintupled by Witcher 3, and the original Assassin's Creed had needed the vote-ins just to make the first GotD while Assassin's Creed II was the series' lone representative in 2015 and it was a 16-seed. All indications were that the series was toast. And yet, we still believed in it here, for the simple reason that Civilization has always been absolute garbage here. Civ V itself had appeared in 2015 and got doubled by Dark Souls in round 1; Civ IV made it into the first GotD and lost a close match with Persona 3 where it was actually favored roughly 57-43 by both Gurus and casuals; the series as a whole managed to just narrowly clear 10% in the opening match of the 2006 series contest against The Legend of Zelda; and in the series' lone failure to come in dead last in a contest poll, the first game in the series beat out Castlevania III for third place by almost 1000 votes while still coming nearly 28000 votes shy of second place Street Fighter II (which was itself nearly 10000 votes out of first). The Persona 3-Civ IV match might've been sticking in Gurus heads, because honestly, neither 47.65% on P3 nor 33.3% on Dark Souls feel like awful results. But the fact that it lost as a favorite, even though with hindsight it doesn't feel like it should have been a favorite, leaves you with a "won't get fooled again" mentality. This is exactly why Kefka's been burning us over and over with his nonsensical newfound strength. But with AC also declining, this was far from consensus; over 25% of Gurus had Civ V, and about 28% of Oracle predictions were for it. The casuals were similarly conflicted but clearly leaning in the correct direction; 63.47% of brackets had this right, which feels about right for a match that had a clear favorite but was somewhat debatable.

Match 43: Sonic Mania vs. Ys VIII: Lacrimosa of DANA

Sonic Mania 15592
Ys VIII 9258

Sonic's woes have been talked about at length for many contests, so a blowout win felt unlikely even though Mania was the return to the series' roots that the fans had long been clamoring for and Ys has always been a fairly niche RPG series. Then again, it's still an RPG series, and a Japanese one at that, so, you know, Flowchart.

Or maybe not. In fact, out of 57 Oracle predictions, a grand total of one predicted Sonic Mania to win with a percentage lower than it actually did, though it did so by a wide enough margin to not even get a top 5 out of it. Was it Sonic antivotes or just RPGFAQs up to its usual shenanigans? We'd hopefully find out next round, since its next opponent would be an indie game regardless of who won the next match!

Match 44: Super Meat Boy vs. Slay the Spire

Super Meat Boy 17138
Slay the Spire 7713

Super Meat Boy was the Guru nomination, which it rode to a 4-seed. Meat Boy had been in the contest before, once solo and once in *sigh* Rivalry Rumble. The 2013 solo outing is nothing to speak of--he came in third to Mr. Game & Watch and Lee Everett from The Walking Dead--but the ability to not get doubled by a Final Fantasy hero/main villain pair, even if it was only Zidane/Kuja, is actually kind of impressive. Though it could just be the silly character design at work. Giving it a fellow indie game to beat was a nice gesture by Allen, since so many Guru noms just fall flat immediately. But would it be beloved enough to get past an established icon like Sonic the Hedgehog?

...My bracket certainly hopes so! (seriously what was I thinking)

---

Match 45: Portal 2 vs. Tomb Raider

Portal 2 15978
Tomb Raider 8493

There was a lot of doubt as to whether Portal 2 would be terribly strong. The series seemed to be something that came and went--the first game did well in the first GotD because it was still fresh in everyone's minds, but it had a short shelf life. And this result...seems to suggest that the concern was warranted. Tomb Raider hasn't been worth a damn here in ages, if it ever was; Lara was at least worth a little bit in the early Character Battles but then again she was one of the pioneers of TJF in video game character design.

Of course, people had concerns about its next opponent as well so who knows what would happen next.

Match 46: Kingdom Hearts III vs. Disco Elysium

KHIII 16654
Disco 7816

The good news is that Kingdom Hearts III got the doubling. The bad news is, it only barely got 68% against a PC-exclusive (with ports to other consoles planned for 2020, but still, currently a PC exclusive) indie game that is nominally an RPG, but contains no combat and instead progresses through skill checks and dialogue trees. Which kind of sounds more like a visual novel to me! So, VN/RPG hybrid, that's fine I guess, but it doesn't sound like the type of thing that should have the strength to get almost 32% on a big name Square Enix series. They can't do anything right these days, can they? Of course, Kingdom Hearts isn't just Square Enix, and its other parent company is known for being one of the most well-oiled machines in the world. Maybe it'd be better for us all if they just bought up the video game/manga giant (yes, Square does more than just video games).

Apparently the title "Disco Elysium" was chosen for multiple meanings, including "disco" being Latin for "I learn". I am certainly learning as a result of this project.

Match 47: Animal Crossing: New Leaf vs. Street Fighter V

AC:NL 16274
SFV 8194

Animal Crossing has been nothing short of terrible in past polls. Or has it? Tom Nook did manage to get not-last place finishes in R1 matches in multiways, both in 2007 and 2013, with the latter being against someone who actually has...a not-last-place finish in 2008. But that was against someone with genuine contest wins, and in bizarre rSFF fashion, no less! Nook and his rival Crazy Redd also nearly avoided a doubling against Amaterasu and Orochi in Rivalry Rumble. The original Animal Crossing (or at least, the first one to be released outside of Japan) also managed to put up 41.45% on World of Warcraft, one of the few times that an entrant selected to win by over 90% of brackets failed to double its opponent and the only one to fail to even break 60%. But its most interesting poll was this very game appearing in the 2015 contest, where it proceeded to look especially like trash in a Nintendo SFF battle with Super Mario World. SMW got the lead to 1000 in only 15 minutes, to 2000 in another 20, and over 3000 by the end of the first hour, an 85-15 type of beatdown. And remember, 2015 didn't have great votals either, so those numbers are a lot bigger than they would be in our prime. So what made the poll so interesting? Well, like this contest, there were four matches per day. Which means that all of your potential opponents for the next two rounds are also active whenever you are.

https://gamefaqs.gamespot.com/poll/6152

"Oh."

SMW had already been slowly bleeding percentage throughout the match; it's only natural when you've got such a big lead and the vote totals are so low, because even a few votes for your opponent can cut your percentage (in fact, hour 14 ended with SMW winning an update 14-4, which is a percentage cut). But once Undertale started rallying, the percentage bleed sped up considerably, and New Leaf even started occasionally winning updates outright. In the end, an entrant that was under 14% near the end of the first hour and at 15.86% when Mass Effect 3's lead over Undertale peaked over two-thirds of the way through the match wound up at 26.73% when it was all said and done. It was a clear sign that if anything was stopping Undertale, it certainly wouldn't be Super Mario World.

Flash forward to 2020. We've all grown a bit older, and many of us have kids. I fully suspect that's as much a reason why Nintendo always does so well here; we've moved past the "M-Rated video games so cool!" phase and want E-rated stuff we can share with our families. And then the pandemic struck, and we're all stuck in our homes, and what should release but a new Animal Crossing game for the Switch! And the thing about Animal Crossing games is, they're not terribly different from one another. If you love one, you probably love them all. The weakening of fighting games made this round a mere formality, not to mention Street Fighter V was hated even by Street Fighter fans. (I wish I'd known that; I figured this was a no-brainer based on Animal Crossing's past results.) The following round, however, would be far more interesting.

Match 48: Red Dead Redemption 2 vs. Dying Light

RDR2 19917
Dying Light 4551

Since this was an obvious blowout (over 94% prediction percentage), it's fun facts time! The first Red Dead Redemption was originally considered a sequel, later redubbed "spiritual successor", to a game called Red Dead Revolver, but when Redemption proved far more popular, they renamed that as part of the series instead of just being the "Red Dead" series. Red Dead Revolver was itself considered a spiritual successor to an NES game called "Gun Smoke". Correction: an arcade game that was ported to the NES. If you're wondering how a Rockstar Games production could date back to the 80s, the original Gun Smoke was a Capcom game, but Red Dead Revolver kind of just sat in development hell for years until Rockstar took it off Capcom's hands.

But things with pedigrees going back to the 80s tend to be very strong here, and even if most of the voters didn't know it, RDR qualifies.

---

Match 49: Dark Souls vs. Hotline Miami

Dark Souls 20080
Hotline Miami 4117

Dark Souls became the fourth game this contest to break 20000 votes, though it was the first of those to give up at least 4000. We knew Dark Souls was going to be strong, so if anything, this is a good performance by Hotline Miami. Nothing to worry about, though; the next match was going to produce a weak opponent for it no matter what happened.

Match 50: Rocket League vs. Dragon Ball FighterZ

Rocket League 12498
FighterZ 11694

This match had more total lead changes than the entire contest to date, including the first one to occur outside the opening hour. Rocket League led at the freeze, but FighterZ spiked a big update at the 15 minute mark to go up by 5. Rocket League retook the lead on the next update, but FighterZ was right back in the lead on the next update and was still ahead an hour into the match. It was the last time all match that it led, but still, progress! A match was actually won by the entrant that trailed after an hour! Okay yeah, "vehicular soccer" game vs. licensed fighting game was the definition of a fodder match, but you have to take the little victories. And this was the tightest Guru spread of the first round, 73-69 in favor of Rocket League, so we knew this could be a tough one to call.

But all of this was overshadowed by a hilariously inept troll. About 40 minutes before the start of the match, a topic was posted stating that "Rocket League must win; I'll alert the Discord with 200,000 followers". And the opening post had a picture to prove the follower count, along with the statement "Someone has to stop BotW". He was, of course, laughed at, and when told it wouldn't work, he said "probably not but someone has to try, we don't need a Wii U game winning the whole contest". When it was pointed out that Breath of the Wild is a Switch game, he instead started defending himself against non-existent accusations that he doesn't even play Rocket League with more pictures showing his accomplishments, or someone's accomplishments at any rate since the username in the pics didn't remotely match the one he was using here. When told that it was fine for him to dislike BotW without taking it personally, he responded that he does like BotW, but as you can see he does play Rocket League.

By the way, this was the second straight day that votes declined. Nice rally, bro.

Match 51: Yakuza 0 vs. The Witness

Yakuza 0 18727
The Witness 5470

Two Yakuza 0 characters made the 2018 contest, one of them getting a good seed. Neither put up any sort of resistance against characters who were either proven fodder or frequently anti-voted. And yet, we all knew that Yakuza 0 would win this easily, because of the quality of its opponent.

Though based on what we've seen this contest, action games are actually quite strong here, even if their characters rarely are. Or, wait, no, that's the opposite of the way things used to be! Guys like Dante and Kratos always far outperformed their games. Something's shifted.

Match 52: Metal Gear Solid V: The Phantom Pain vs. Dead Cells

MGSV 17542
Dead Cells 6651

While the days of SFF matches out the wazoo seems to finally be behind us, Allen can't help but get cute with the bracketmaking sometimes. Dead Cells going up against a Metal Gear game is funny enough, but we're also looking at a probable 2015 rematch in Round 3. It wasn't even a particularly close match in 2015, though it was one with what seemed like a fairly even split in prediction percentage.

MGSV was expected to have suffered from backlash. It feels like nearly every established franchise has been facing backlash towards its most recent games. No wonder indies are doing so well; any franchise not owned by Nintendo is essentially getting punished for how good earlier games in the series were, or are perceived to have been. We're in the midst of maximum hype about a Final Fantasy VII remake, and seeing the ads, it certainly does look beautiful. Modern graphics with, presumably, the original story. But I can't help but wonder, if it weren't a remake, would it be nearly so beloved? I haven't gotten that far into FFVII yet, but so far I'm less than impressed, and pretty new graphics won't fix that.

Though I was happy that the game took mercy on me after I failed the stealth sequence in Shinra Tower too many times. Funny, I thought that the problem with modern games was that they held your hand too much? It wouldn't surprise me at all if some gamers held that mercy against the remake, if it's still there.

---

Hey nice.

---

Match 53: Batman: Arkham City vs. Xenoblade Chronicles 2

Batman 13688
XBC2 11445

Remember everything I said about action games being the in thing now, and non-video game origins not being a detriment? Yeah, this match basically decided to throw that all out. 93% of Gurus got this right, and 71.32% of casuals, which means that it was by and large not a debatable match--yet look at how close it was! This was an absolute embarrassment, barring the possibility that XBC2 was almost as strong as the first--based on 2015 X-Stats, Arkham City is worth 52.86% on XBC, yet it could only get 54.46% on the sequel. And in the next round, it was set to take on a game that finished ahead of XBC in the 2015 X-Stats.

Match 54: BioShock Infinite vs. Terraria

BSI 17005
Terraria 8123

If there was any good news for Arkham City, the presumptive favorite in Round 2, it was that its opponent, BioShock Infinite, was also disappointing in Round 1. Terraria is a sandbox game similar to Minecraft, hardly the type of game that GameFAQs goes crazy for. Yet here it was pulling down 32.33% against BioShock Infinite. Admittedly, Infinite hasn't had nearly the success in these contests, be it in games contests or Character Battles, that BioShock or BioShock 2 have had. But it's still an established franchise, which means it should be worth more than this.

Match 55: Diablo III vs. Zero Escape: Virtue's Last Reward

Diablo III 17340
VLR 7788

This feels like it was actually a pretty good performance for Diablo III. No, VLR isn't strong, but Diablo III was absolutely excoriated by the fandom when it first came out. Though I guess nowadays, how good a game you are when you first come out is completely irrelevant because everything's so wired that they can just patch things? And D3 is apparently a better game than it used to be. We'd see that again soon.

Still, Diablo II beats RE2 outright in the 2015 X-Stats, 56-44. If we peg Diablo III at even half Diablo II's strength, VLR's projected to get 26.73% on RE2, which is higher than what VLR is actually pegged to get on RE2 via 2015 X-Stats. Even adjusting for the expectation that this RE2 remake is actually stronger than the original, which flies in the face of everything this site has stood for for the past two decades...there's still no X-Stat voodoo I could come up with to suggest that VLR is weaker than DR2.

Also, did people forget that 999 came out in 2010? It seems really weird that VLR made the bracket and 999 didn't.

Match 56: The Last of Us vs. Crusader Kings II

TLoU 20566
CKII 4560

No meaningful conclusions could be drawn from this match, because Crusader Kings II was always going to be weak. But still, is it really stronger than Hotline Miami? Because it got a better percentage on TLoU than Hotline Miami got on Dark Souls, and the 2015 X-Stats say that TLoU should be stronger than Dark Souls. (Technically speaking, they say that The Last of Us should actually get a higher percentage on Dark Souls in the division final than on Arkham City in the division semis, but Arkham City was in the top half of Division 1 which means its X-Stat number is horribly distorted by the collective delusion that a. Chrono Trigger would've won in the absence of rallies and b. that matters. Because it doesn't matter. Rallies have always been a part of the contest. Many of our greatest contest matches were great because both sides were rallying. And I guess they were fine with that because even if they shifted the outcome, they probably didn't shift the percentage that drastically. But when one side manages to tap into a source of votes that the other can't match, suddenly we have to correct for what "should be" instead of what is. And once a rally has knocked someone off, there's no longer any need to account for logic in projecting what "should have" happened. I'll say it again: Yes, Chrono Trigger was indirectly stronger than Final Fantasy VII in 2015. You can see it in the actual percentages they got against Melee and you can see it in the percentages that Allen was kind enough to provide for "registered voters only" for all of both Melee's and Undertale's matches. Wouldn't have mattered. Final Fantasy VII won't lose to a fellow Square game. There was a pair of PotD's shortly before the contest, one for "What is your favorite Final Fantasy game" and one between the top two vote-getters of that poll for "which of these two do you like more". Both times, FFVII lost to FFVI, and then FFVII still narrowly broke 60% when they faced each other in the Division 2 Finals.

Also, let's take a look at those "registered user only" numbers, and pretend they're the actual numbers. Note that this isn't the same as subjecting the 2015 contest to 2018/2020 rules because we don't know what the registered/unregistered splits are and either way, 2-1 isn't the same as 1-0. Looking at Melee's percentages against Undertale and Chrono Trigger, Undertale would be projected to get 22.33% on Chrono Trigger.

Undertale vs. Mass Effect 3 - 29.49% 70.51%
Fallout 3 vs. Undertale - 75.67% 24.33%
Undertale vs. Super Mario World - 19.28% 80.72%
Pokemon Red/Blue vs. Undertale - 76.30% 23.70%
Undertale vs. Super Mario 64 - 26.21% 73.79%
Smash Bros. Melee vs. Undertale - 69.59% 30.41%
Undertale vs. Zelda: Ocarina of Time - 33.44% 66.56%

Yes, Super Mario World, the game that started hemorrhaging percentage to Animal Crossing: New Leaf when Undertale started rallying against ME3, is projected to beat Chrono Trigger by this metric. It's probably misleading because it's only looking at "logged in", not the age of the account--while it's possible that Undertale legitimately picked up a bandwagon here on GameFAQs, it's far more likely that Undertale's percentage rises each round after the SMW match because more of the people rallied from off-site had accounts, possibly because they wanted to taunt the "butthurt" people (though I'm pretty sure that BracketEntry was a false flag operation by someone from Undertale's camp).

So, yeah, that happened then, and Allen conceded to give longtime GameFAQs users (well, anyone who had an account at any point before the contest began and still remembered how to log into it) a greater say, and it...flipped a grand total of three matches last contest, probably only one of which actually made B8 happy, and through pretty much all of Round 1 there hasn't been a single match whose outcome would've changed no matter how the votes were counted because registered and unregistered voters, while their percentages have varied, have always chosen the same g

"Also, did people forget that 999 came out in 2010?"

I did or I would've nominated it. <_<

---

Match 57: The Elder Scrolls V: Skyrim vs. Subnautica

Skyrim 20647
Subnautica 3933

84.00%. A nice round number, the type of match that perfect Oracle predictions are made of. And I threw one away! Among the many games that this board plays to liven up the contest even more is the "Spread Betting" topic. Allegedly the spreads are set by the Oracle picks, but having just been told by the Analysis Crew (and having done my own research and agreeing) that Subnautica would be a contender for weakest game in the bracket, I figured that Skyrim covering the spread was the smart pick, then checked my Oracle prediction to make sure that it was high enough to cover the spread.

The Oracle consensus, even after me raising my prediction from 84% even to 86.67%, was only 84.77%. The average of the lowest and highest predictions was roughly 85.5%. Where the hell did they get -72 from?!

So, yeah, Subnautica's weak, but getting 16% on Skyrim easily takes it out of contention for weakest game in the bracket. Far weaker games than Skyrim got higher percentages in R1.

Match 58: Fate/Grand Order vs. Journey

FGO 8307
Journey 16271

Journey was piss-weak in 2015, so I figured the Fate name had enough clout to pull this one off. Then the Crew reminded me that even the people who play gacha games hate them, possibly even more so, because they make their money by making all the good stuff really hard to get. My Oracle still turned out pretty bad, though, because I still didn't have enough faith in Journey to not be awful too! And the apathy votes should've still leaned towards FGO because, again, recognizable name. It could very well fare better with people who haven't heard of it and don't know it's a gacha game than those who have!

Also, in a contest full of blowouts, Journey failed to double a gacha game. On GameFAQs, which hates mobile games and other "casualbait" (even though FGO is less casualbait and more otakubait.)

Match 59: Final Fantasy XIV: A Realm Reborn vs. Crypt of the NecroDancer

FFXIV 17583
Necrodancer 6997

This was one of those matches that I barely even considered debatable when making my bracket and it turned out it wasn't debatable, but in the other direction. I mean, let's look at the facts here: this site has never been big on MMOs, as evidenced by the way WoW constantly gets embarrassed, frequently losing as the favorite and failing to win convincingly even when it does win. Again: only entrant, game or character, to win with less than 60% with over 90% bracket support. And this site hates most of Square's recent output, so even if this were a regular Final Fantasy and not an MMO, it wouldn't be that strong, though it'd still be strong enough to win. And most importantly, Nintendo has been famously reticent to allow anyone else access to their IPs, ever since the Philips CD-i debacle. Even though Cadence of Hyrule isn't the game in this poll, its mere existence is a vote of confidence from people whose opinions matter a lot to GameFAQs.

Turns out, FFXIV is actually really well respected now. That "A Realm Reborn" in the title isn't merely a fancy subtitle; it's to differentiate it from the original release of FFXIV, which was such a disaster that Square basically scrapped the whole thing and then went meta with the rerelease, setting it in a post-apocalyptic world that's basically the same cataclysm that they dropped on the original release's world when they killed it off. People love it, especially because it finally gave WoW some much-needed competition.

So after being informed that I'd horribly misjudged two matches in the same day again, I went into the Oracle topic, came up with what I thought was a good percentage for FFXIV > FGO, then went back before posting because I'd already forgotten what the other game that I'd wrongly had losing was...and realized, oops, FFXIV was facing Crypt of the NecroDancer! So I had to redo it, and it was killing me because I still believed that if the opponents were reversed, I'd only have to change one outcome; that Crypt of the NecroDancer could still win.

After this? I no longer think Crypt of the NecroDancer could beat Journey. Unless FFXIV is just that beastly, getting only 28.47% is just not a good performance, and I really overestimated the psychological effect of "Nintendo let these guys make a game with one of their most valuable IPs".

The actual Cadence of Hyrule probably would've been stronger. Too bad it's a 2019 game or else we could find out in ten years.

Match 60: Uncharted 4: A Thief's End vs. Mario Kart 8

UC4 10114
MK8 14467

This match was debated for all sorts of silly reasons, most of them stemming as to whether voters would see it strictly as "Mario Kart 8", the Wii U game, or if they'd see it as both Mario Kart 8 and its update Switch version, Mario Kart 8 Deluxe. Because of course, voters will anti-vote anything associated with the Wii U. </s> Uncharted has never been that strong here; Nathan Drake managed a single first-round win in 2013 against Pac-Man and Steve from Minecraft, but is better known for coming in 4th place in his debut despite having CATS in his fourpack and then later becoming the first 2-seed to lose in round 1 in a format with 16-character divisions (though we'd had 8-1 upsets in the 8x8 era, so...) As for the games themselves...er, the game itself, Uncharted 2, went two rounds in GotD, beating The Legend of Dragoon and Street Fighter IV, before losing to Halo, then beat out The Walking Dead in round 1 in 2015 before losing to Sonic the Hedgehog 2. Neither Uncharted nor Uncharted 3 have ever been in a contest. No, Uncharted wasn't even in the vote-ins for the first GotD. (Uncharted 3 would've been in this one, not that one.) Honestly, this feels like a really bad result for MK8. I honestly wasn't that concerned about having blown two matches in the same day because unlike the Bravely Default/Ni No Kuni disaster, at least my wrong calls would both be R2 losers anyway, but after seeing MK8 held under 59% by UC4 while FFXIV was putting up over 71% on NecroDancer, I'm not so sure anymore! Obviously MK8 had the stronger opponent, but was it by enough to account for that big a difference?

Though it should be noted that at 55.86% correct, this was the most debated match of Round 1 with the casuals, narrowly edging out Nioh's upset of Binding of Isaac: Rebirth which 56.01% of brackets had incorrect.

---

Match 61: Pokmon HeartGold/SoulSilver vs. A Night in the Woods

HGSS 19898
Night 4379

Long before the contest was even announced, I theorized that due to the apparent "two generations" rule, where two generations have to pass before people can start getting nostalgic for a Pokmon game, that Black/White could be a beast in the Game of the Decade because it'd have the benefit of three generations having passed. Then Allen decided to allow remakes into the contest, possibly fueled by RE2Make winning GotY 2019 (wait, what?!), and even Gen V, which I'm pretty sure is remembered fondly because its metagame was very much still evolving when the Gen VI games came out (blame this on Gen VI being when Game Freak shifted to worldwide releases; Gen V didn't just feel shorter than previous generations; it was shorter), was completely missing from the bracket, let alone the newer games, but Gen IV got a representative with a 3-seed. Because as we all know, being essentially from the previous decade is a winning strategy for Game of the Decade. Just ask "Ocarina of Time 2". (I'm joking, of course; Majora's Mask proved it was legit in 2015.) Though of course, no GameFAQser would ever forget that HGSS was released in 2010 and not 2009. You could say its release date is...burned into their minds.

https://tenor.com/view/horatio-csi-miami-yeah-meme-gif-16346536

Other than that, there's not much to say about this match. No entrant in Round 1 received more raw votes from unregistered voters than registered, but Night in the Woods came closest. There were 1501 more registered voters than unregistered, and of these 1501 votes beyond the unregistered output, a grand total of 2 went to Night in the Woods. GameFAQs loves its Pokmon.

Match 62: Celeste vs. Rayman Legends

Celeste 12027
Rayman Legends 12244

FINALLY! FINALLY! FINALLY!

Yes, it really needs that many finally's. A lot of trends were broken in this match, saving Round 1 from being complete and utter monotony.

Celeste jumped out of the gate on fire; the earliest update on our vote tracker had it receiving 16 of the first 18 votes. Of course, there are a bunch of registered users who make a contest of getting the first vote, so it's probably fewer votes than that. It was a bit wobbly after that until Celeste spiked back-to-back update wins of 52 and 56 near the end of the first hour. From there, though neither side could put together more than 5 straight update wins (and Celeste had managed that feat before it hit the two big ones, all of which amounted to a smaller lead than either of those two updates alone), Celeste slowly built up a lead, peaking at 333 at 1:10 AM, about five hours in.

At 3:30 AM, however, it was Rayman Legends that broke that trend, winning its sixth consecutive update and pushing the deficit under 200. Celeste promptly won the next update, but then Rayman rattled off a string of seven, then a string of five shortly thereafter. By 5:10, the lead had dwindled all the way to 65 votes--but then Celeste reversed the trend. At 6:20, Celeste was coming off its sixth straight update win and had climbed back to 123 votes ahead (hey, we're still talking about this contest's crappy vote totals; large swings in updates late in matches just don't happen, except sometimes in the big blowouts where the leader can still put up decent net positives). The next update was a stall...and then Rayman started chipping away again. At 7:15, however, chipping was no longer enough; Rayman spiked a 29-vote swing to cut its deficit to 66. At 8:20, Rayman took its first lead of the match.

It was short-lived. Celeste retook the lead a mere 10 minutes later as part of winning 6 of the next 7 updates after falling behind.

At 9:35, the poll read 50%-50%, with Celeste clinging to a 1-vote lead. And again at 9:45, after it responded to Rayman's +9 with a +9 of its own, constituting two more lead changes. It then got a bunch more little wins to pad its lead all the way up to...18. At 10:25, however, Rayman spiked another +22, flipping a double-digit deficit into a double-digit lead. And again Celeste fought back, bringing it to a dead tie at 11:00 AM and winning the next update by 11. So of course Rayman Legends wins the next update by 12. Lead changes 6 and 7 in rapid succession.

And that was all the lead changes we had, though it took until the final four hours or so for Rayman to really start breaking away. In the end, it was a 217-vote margin and a percentage of 50.45%-49.55%. The Gurus were really tripped up by this, with a correct percentage of 27.46%, but at 37.67% the casuals weren't much better. And in addition to this being our first match that was honestly intriguing for quite some time, it marked the first match where the registered and unregistered voters disagreed. Celeste polled at 50.37% with registered voters and 47.51% with unregistered voters. Oh, hey, that means the unregistered voters actually flipped a match! Bravo for them, I guess. It's a 1-point match anyway as neither of these games stood a chance against HGSS.

I honestly thought this was the first appearance by anything Rayman in any of the contests. Courtesy of the wiki, however, I learned that Rayman the character has in fact been in two contests, though one of them was the bloated 2013 contest and his seeding was below the line where he'd have made a 128-character bracket (yes, even when you account for the 8-seeds being taken up by Board 8 rallies without regard to where they ought to have been seeded.) He didn't make it to Round 2 in either, though he at least managed to beat out Commander Video in 2013. Honestly I'm not surprised; even among forgotten platformer stars, Rayman feels irrelevant. What an amazing match though! I normally try to do the entirety of a day's matches at the same time, but my hands are already getting sore (admittedly, I'd already spent quite a bit of time on the computer today before writing these) so I'd probably end up half-assing the next two matches' write-ups if I tried to do them now. Let that be a testament to how good, and how necessary, this match was! We were just one day away from the highlight of Round 1 being a troll claiming he was going to start a rally before a match even started and votals for the day in question being down from the previous two while the game in question actually trailed after an hour but still won. Most matches could be called right at the freeze; only a couple of others even had any lead changes at all but that had been the only one where it took over an hour for the winner to take the lead for good. A match that stayed inside of 51-49 for the final

Match 63: Divinity: Original Sin II vs. Hearthstone: Heroes of Warcraft

D:OS2 16804
Hearthstone 7471

Hearthstone was easily the weakest game in the X-Stats in 2015, certainly unadjusted and probably adjusted too. (The unadjusted X-Stats are mostly worthless because Undertale scored its largest blowout in the final against OoT. I feel like that definitely says something about Undertale, given that Link very nearly managed to stop Draven anyway, but I'm not sure what. But they can still be useful for comparing games that were on the same sides of the bracket.) Divinity was therefore expected to win easily, and it did just that. Some people were a bit surprised at how easily it did so, however, given its weak showing in the only other poll it had been in, but we all know that non-contest PotD results, even from GotY, are often suspect when used to predict contest matches. Nevertheless, this was the match where it was theorized that 2017 had actually been a stealth monster of a year and that there had been a number of games, Divinity: Original Sin II among them, that had flown under the radar because their only prior appearance in a poll was a poor showing in that year's GotY. People started talking about how 2017 would've done in the Years Contest. The thing is, it probably still wouldn't have done that well. There was a running joke about "just pick the closest year to 1998", except it wasn't a joke at all; if you actually made your bracket that way and picked correctly on all the "coin flips", you'd have a grand total of two wrong picks, though one of them would have been on a semifinal match so you'd still be far from perfect.

Match 64: Nier: Automata vs. Bayonetta 2

Nier: Automata 16604
Bayonetta 2 7668

Both of these games were clearly going to be far weaker than their TJF-packing protagonists, although there was reason to think Automata might not be as weak as the original Nier presumably would be. I say presumably because we've never seen it in a Games Contest! It's actually recent enough that it would've been in this GotD, not the last one, and normally the instinct is that the first game in the series is always the strongest but Automata was far more successful commercially so it probably is stronger. Nier the character was weak in his only outing in the Character Battle but he was being SFFed by Riku. Then again, that kind of says everything, doesn't it? In fact, if you tried to use past results without understanding what they actually mean, there's a direct route you could've taken to getting this match horribly wrong: Riku beat Nier in 2013; Bayonetta beat Riku in 2017; sequels are always weaker than the originals; therefore there's no way that a Nier game could beat a Bayonetta game.

The only question left is just how much weaker than the multi-platform original the Wii U-exclusive Bayonetta 2 is. Because this is GameFAQs, where Link Always Wins, and yet Nier: Automata won more convincingly here than 2013 Game of the Year The Legend of Zelda: A Link Between Worlds did against the first Bayonetta. Even if being a handheld game is always a negative here unless you're Pokmon, I find it hard to believe that Nier: Automata would challenge ALBW. (Granted, it wasn't that much more convincingly, so the two Bayonetta games would have to be pretty close in strength for Nier: Automata to actually win.) Either that, or the "2017 factor" is stronger than we thought.

Match 65: The Legend of Zelda: Breath of the Wild vs. Halo: Reach

BotW 22268
HaloR 3972

Breath of the Wild set a new contest high for votes received in a match, which is honestly kind of amazing given how lopsided so many Round 1 matches were. Then again, one of the truisms of the board is "there's rallying in every close match". This does indeed appear to be the first day to have over 26000 total votes, so there probably was some rallying going on, and since voters have to vote in all four matches (even though the totals never come out quite equal, because there is some way around that if you know what you're doing), it meant that Breath of the Wild got to pile on Halo: Reach even more. No one was complaining.

Match 66: Final Fantasy XV vs. Hollow Knight

Final Fantasy XV 13229
Hollow Knight 13008

We'd nearly gone all of Round 1 without a single barnburner, and on day 1 of Round 2, it looked like the floodgates had opened after the final day of Round 1 broke the dam. Which I think is a mixed metaphor, maybe. They're both related to a lot of water all over the place. Final Fantasy XV won the board vote, which is hardly a very Final Fantasy thing to do, but it took only 15 minutes for Hollow Knight to take a lead. It held onto that for an hour and 15 minutes before FFXV took the lead for one update, and then the next update was a straight-up tie. Hollow Knight did swing out in front on the update after the tie, but FFXV was right back in the lead the following update. 15 minutes later, back to Hollow Knight; another 15 minutes, back to FFXV. At this point, we were 2 hours and 15 minutes into the match and had had six lead changes. And...we wouldn't get a seventh, though it was far from smooth sailing for FFXV. It briefly got its lead above 200 at 10 minutes to 1 AM, and a mere 45 minutes later, it was down to 118. 2 hours after that, it again got past the 200 mark, and again Hollow Knight immediately started cutting.

And FFXV started fighting back; at one point, HK shaved off more than 50 votes in just 10 minutes and it took only 15 more minutes for FFXV to get its lead back to higher than it had been before those 10 minutes. At 5:05 AM, FFXV's lead breached the 200 mark again, at 215 and this time it stuck. Really stuck; for the entire hour after that, the lead never dropped below 213 nor exceeded 225. It crested at 232 at 6:20, then Hollow Knight managed to briefly get the lead under 200 at 7:30. As was the match's trend, this meant it was time for FFXV to gain momentum; by 7:45, it had a new largest lead, 238, though this was merely the beginning; by the time Hollow Knight next took an update, it was to reduce the lead from 298, and FFXV breached the 300 mark not long thereafter. Another prolonged stall began until FFXV spiked a win of 23 at 9:40 to push the new high to 340, only for HK to spike a 27 ten minutes later. FFXV went on a push at 11 AM, winning an update by 32, and at 11:35 AM, its lead stood at 400 votes exactly. And that's where it peaked; HK started slowly battling back, but far too slowly; it wasn't until 6:30 PM, just 90 minuted from the end of the match, that the lead dipped below 300 again. HK would continue to make a late push to make the match look closer than it really was, but the mat

Been letting this slip

---

Eh if you work it i shard to keep updating things. Also since new matches keep happening we are bound to have new records.

---

Match 67: Monster Hunter: World vs. Sekiro: Shadows Die Twice

MHW 14644
SSDT 11599

The "Soulsborne" boost evidently doesn't extend to Sekiro, because this match was expected to be close and it wasn't close. Remember, this is the same series whose only previous outing involved being upset by Suikoden II. (Yeah, I know, not the same game.) More importantly, this is Capcom, whose last win in a debatable match was probably Mega Man > Charizard in 2013. (Oh, right, the DMC5-DKCTF match in round 1. There was that.) The idea of MH:W actually being strong seems ludicrous, but here it was, reaching R3 and still being the favorite to do so. Well, at least it'd be facing something strong next round.

Match 68: Bastion vs. Dragon Quest XI: Echoes of an Elusive Age

Bastion 7370
DQXI 18869

Still nothing to see here. DQ11's path was pretty much set in stone from the start: Win three matches, then get beat pretty hard by BotW. Nothing seemed to indicate that this won't still be the case.

Match 69: Mass Effect 2 vs. Fallout 4

ME2 18568
FO4 9197

Well obviously Mass Effect 2's going to win; a weaker Mass Effect game than 2 did better against Undertale than a stronger Fallout game than 4.

But seriously, this was a good showing for Mass Effect 2, the type of performance it needs if it wants to stand a chance at breaking 40% against Breath of the Wild in Round 5. (It probably won't do it, but it looks like it can avoid the doubling.)

Round 4 opponent? No, there's no cause for concern there, don't be silly.

Match 70: Borderlands 2 vs. Horizon Zero Dawn

Borderlands 2 12471
Horizon Zero Dawn 15292

When HZD scored a massive blowout of Fortnite, we all just chalked it up to Fortnite being antivoted to kingdom come. After this win, on the other hand, it looks legit. The Gurus were fairly split on this one, though they ultimately favored Borderlands 2; by the time Oracle came around, however, they were leaning towards HZD, likely as a result of BL2's Round 1 performance. With the casuals? Your guess is as good as mine who the favorite was. Technically speaking, more brackets picked HZD to win in this round than to lose in it, but the group that picked HZD to lose last round had the heavy plurality, 43.13% to 28.65% to 28.22%. Which means the favorite was probably Borderlands 2, which had an 80.77% prediction percentage in Round 1, but it's quite likely that "Horizon Zero Dawn-Fortnite winner" was favored over Borderlands 2 and certainly at least a chance that Fortnite could even be the favorite despite not being the favorite last round if enough of its R1 pickers were expecting it to be RallyFEAR and picked it to go deep. At any rate, this was a solid win for Horizon Zero Dawn. Its goal right now? Try to put up a better percentage on ME2 than its Round 4 opponent will. Prior to this round, I'd have called that impossible, but it just put up over 55% on presumably a stronger opponent than either of them faced, and neither of them broke 58%.

Borderlands 2 (2015g) VS Fire Emblem: Awakening (2015g)

Borderlands 2 has a strength of 21.33.
Fire Emblem: Awakening has a strength of 22.89.

Fire Emblem: Awakening wins with 53.41% of the vote!
A win of 4,149 with 60,875 total votes cast.

...Oh. Never mind then.

---

Match 71: Resident Evil 2 vs. DOOM

Resident Evil 2 15984
DOOM 11780

Huh, maybe DOOM's blowout of INSIDE wasn't just INSIDE being piss-weak? Not to say that it isn't, but apparently DOOM has some real strength. Or RE2 isn't as strong as it seems. Still, 57.57% is not a great score for Resident Evil 2; according to the 2009 X-Stats, RE2 should get 66.28%.

That may or may not have been an excuse to put "according to the 2009 X-Stats" into a write-up about a contest that only covers games released between 2010 and 2019. I was planning on using the 2004 X-Stats, but apparently RE2 didn't make that contest. I guess first entry syndrome won out there; I always thought RE2 was better regarded than RE1, but RE1's the only one that made it in '04. Granted, '04's nomination process was awful and resulted in a bunch of games from obscure systems getting in.

Also, unlike RE2, the "DOOM" in this contest is not a straight-up remake of the first game in the series. It's a brand new game, and was actually called "Doom 4" when it was first announced. So that makes the stat especially meaningless...except this is OldFAQs, where "pick the year closest to 1998" is a valid strategy. Resident Evil 2, the original one, is a 1998 video game. Wouldn't you expect it to get even higher on 2016's "Doom 4" than 1993's "Doom"?

Also, holy crap that is legitimately 21 years between original and remake. That's long enough to be "excusable". It's really been that long since the fifth generation, huh? And yet we're still technically considered to be in the eighth generation, even though we all know that since the Wii U is an eighth generation console, the Switch kind of has to be a ninth generation console by default. It'll probably be reclassified as a 9th gen console once the PS5 and XBSX come out later this year.

The Switch has officially succeeded the 3DS, also an eighth-generation system, as well as the Wii U. Even more evidence that the 9th generation is already well underway.

Match 72: Fire Emblem: Awakening vs. Bloodborne

FE:A 11688
Bloodborne 16076

Oh, hey, something that the lol X-Stats wouldn't have predicted! Two write-ups ago, I said that Borderlands 2 was presumed to be a stronger opponent than either RE2 or Bloodborne faced, then looked up the X-Stats for Borderlands 2-FE:A and found that according to 2015, that wasn't the case. Only problem is, the 2015 X-Stats not only have FE:A ahead of Bloodborne, but they also have Borderlands 2 ahead of Bloodborne! Narrowly, of course; something like winning with 50.21%. Apparently FE:A's a lot weaker than it was in 2015, which kind of makes sense because the backlash against FE:A from the hardcore series fans only got worse when the next game was everything wrong with Awakening turned up to 11. It's funny; Awakening was meant to be a big series of shoutouts to the entire series thus far, and then Fates ended up just being one long ode to Awakening, except it was far worse than Awakening ever was because the mechanics that made sense in Awakening were just kind of bolted onto Fates for the sake of keeping them. About the most encouraging thing about the series is that Echoes was still fairly loyal to Gaiden, because Genealogy is next up to get a remake and even though that's the game that Awakening was referencing with the child mechanic, it could get really ugly if they try to make it too much like Awakening/Fates.

Oh, right, Bloodborne. There's absolutely no good way to determine the strength of any of these games; it seems that the Undertale Effect was a lot harder to untangle than we thought. Still, Bloodborne put up a better percentage on FE:A than RE2 did on Doom. If we take that underperformance as a sign that contrary to popular belief, the RE2 remake isn't as strong as the original, well, RE2 isn't even projected to break 56% on Bloodborne by the 2015 X-Stats. The turnaround it made on FE:A, it's by enough that it should be able to get past RE2 as well. Remind me again where all these games actually were in 2015's bracket?

Okay, RE2 was on the opposite side from the rest of them, so its numbers are filtered through two Undertale matches including the final. And Bloodborne is stuck behind a bunch of SFF matches--it was in the division that had a ALTTP-MM final for the right to lose to OoT the next round. Yeah, these numbers are useless.

Match 73: Super Smash Bros. Ultimate vs. The Legend of Zelda: A Link Between Worlds

SSBU 17671
ALBW 9912

Pretty standard SFF match here. ALBW did well to avoid the doubling. It was pointed out that SSBU has more Links in it than ALBW has, so yeah, not even the LAW was going to help it here. It was a shame that two GotY winners had to face each other so early in the contest, but at least the 2013 winner could take solace in the fact that the 2012 and 2014 winners were out of the contest, too. In fact, the 2014 winner had never been in it to begin with. That was the year that Smash 4 came out, and being a Smash game is basically an instant win button in GotY oh hey Melee what are you doing there.

Most of the individual consoles' polls in 2001 only got their winner into the final poll, but for some reason the top three finishers in the PS2 poll all made the final poll, and finished 1-2-4. Melee lost to freaking GTA3. Weren't we supposed to be NintendoFAQs, not SonyFAQs? And yet Brawl, 4, and Ultimate all won their years. 4 did it so convincingly that they stopped at just the one final poll!

Match 74: Marvel's Spider-Man vs. Minecraft

Spider-Man 15903
Minecraft 11681

So about that assertion that Minecraft is weak...ah, wait, we don't have any gauge on Spider-Man. Let's try that again.

So about that assertion that licensed games are weak... 57.65% may not seem terribly impressive for this tournament, but it is a round 2 match, and more importantly...Minecraft's not weak. Back in 2015, it got 43.28% on Halo. Not a Halo sequel; the original Halo. Yes, Halo is known for being a disappointment in contests, but it is still a big-name franchise and this is the first game in the series we're talking about. And Spider-Man did better against Minecraft than Halo did.

Yeah, this was a match in which both games looked good. Minecraft was able to easily cruise past the percentage predicted for it by the Oracles, but Spider-Man still gets to say it beat a game harder than Halo did. Win-win! Except for the casuals; it's pretty clear by the prediction percentage that they thought Minecraft would win this outright.

---

Match 75: Octopath Traveler vs. Shovel Knight

Octopath 13696
Shovel 13895

If you thought the end of Round 1 and beginning of Round 2 were exciting...you ain't seen nothing yet.

It didn't look like it would start out that way. Octopath managed to take a lead 15 minutes in after trailing at the freeze, but Shovel Knight easily flipped it on the next update and got the lead into triple digits a mere 20 minutes after that. By the two-hour mark, the lead was over 200, and Shovel Knight followed the 35 vote gain that it made to end that hour with a 38 vote gain to start hour three. OT made some nice cuts as we reached 10:30, but again SK easily countered, tacking on 31 and 46 votes in consecutive updates. Every time Octopath looked like it wanted to cut, Shovel Knight countered.

At midnight, four hours in, Shovel Knight led by 340 votes. It would gain one vote on the next update, then spend the rest of the hour futilely trying to increase its lead. But then the next hour hit and bam! +20, +24, new max lead at 1:15 AM. Shortly after 2 AM, SK's lead peaked at 365, but then at 2:15, Octopath made a sizeable cut and followed it up with another one two updates later. SK recovered, forcing Octopath to wait until nearly 3:30 to get within 300, but from there, the comeback was slow but inevitable. Then, suddenly, it wasn't so slow. Octopath got within 200 at 10 minutes to 6, then within 100 at 7:05, and a 25-vote cut at 7:20 left Shovel Knight's lead at a mere 45 votes. From there, however, it went into a stall; an hour and a half later, Shovel Knight's lead was at 43, having not dipped below 35 nor risen above 56 during that time. Shovel Knight won the next update to increase its lead to 61 at 5 minutes to 9, only to be met with an Octopath run. 61 became 46, then 24, then 9, then 3, then 3 again, then 1...and with its lead down to a single vote, Shovel Knight finally won another update, then two more, then a net positive over the next two, and then another one. At 9:55 AM, Shovel Knight had restored its lead to 46 votes.

At 10 AM? Deadlock. At 10:05? Octopath Traveler leads by 50.

Wait, seriously? Did we have an honest to God rally going on? Because that is an insane swing for this deep into a match. Octopath slowed down after that, netting only +17 over the next half hour before picking it up again to gain its first triple-digit lead at 10:50. It hovered for a bit, got up to 151 at 11:40, let SK spend about half an hour gaining back about 30 votes, then quickly smacked it down in only 10 minutes. At 12:45 PM, Octopath led by 193 and seemed primed to win easily. Shovel Knight took a chunk back, but it could never keep much of its gains; at 2:15, Octopath still led by 171.

And then the comeback was on. Highlighted by a +21 at 2:25 and a +27 at 3:20, Shovel Knight got the deficit down to 36 before Octopath started putting up any resistance. And resist it did; it took until 5:55 for Shovel Knight to regain the lead, a +17 giving it an immediate double-digit advantage. Which Octopath wiped out in two updates to take a 1-vote lead at 6:05. Shovel Knight was back in front at the next update, but only by 6 votes, and the next update only grew it to 8...then back to 6...then 6 again...

And at 6:30 PM, Octopath spiked another of its huge updates. +36, lead of 30. They traded update wins for the next 20 minutes, with a net of +8 for Octopath.

And then, suddenly, it was Shovel Knight that had the huge rally power. A 38-vote deficit became a 1-vote lead, then 37, then 50, 76, 99, 137... Shovel Knight won 13 of the final 14 updates to end up winning by 199, the largest lead by either side since 5:50 AM. A title previously held by its penultimate-update lead of 194, one vote more than the largest Octopath's lead ever got. Shovel Knight started out looking like it would win easily, then spent much of the second half of the match trailing, and in the end, still won by a fairly comfortable final margin. The registered voter bonus carried much of that; Shovel Knight's raw vote lead was just 57 votes, courtesy of a 136-voter advantage with registered users offsetting a 79-voter deficit with unregistered.

The favorite in this match? Obviously it was Undertale. At 18.86%, Shovel Knight > Octopath Traveler had the lowest prediction percentage of the tournament so far...taking the mantle from Octopath Traveler > Undertale, at 21.58%. Undertale was the Guru favorite, too, and as for the Oracle...well, it seems that the Oracles were convinced that Octopath's win was a result of Undertale anti-votes, not its own strength. Very few people predicted a match this close, and most of the ones who did were Octopath pickers trying to mitigate the damage if they were wrong. I ended up with a top 5 for this match going over by nearly 3% on SK's percentage. So let's give a hand for everyone involved here. It's clear by the wild swings, at least relative to our vote intake, that this was a match where both sides were rallying, and it brings into question another outcome. It was, of course, noticed that Undertale had been trying to rally yet again, though it was far weaker than what it put up in 2015, and the common belief was that it just wasn't getting it done and couldn't get past the antivotes. But now, perhaps, one has to wonder if we weren't giving Undertale too much of the credit for the increased traffic. Maybe it just got outrallied?

Shovel Knight's run would be coming to an end the next round, of course, faced with Super Mario Odyssey. But it should be noted that outside of 64, 3D Mario has had a long and storied history of looking like utter trash in Games Contests. Whether it's Galaxy losing to CoD4 twice in 2009 or Sunshine and NSMBWii getting upset by Halo 3 and Marvel vs. Capcom 2 in 2010 or the Galaxy games losing to Super Metroid and Mass Effect in 2015, Mario just hasn't been worth much as a platformer franchise this century. Maybe he continues to stay so strong in character battles because this is RPGFAQs; the Paper Mario games have done just fine in contests, after all. (Of course this isn't actually the reason, but the point still stands; PM64 and TTYD are the only Mario games from either of the past two decades to do well in any contests.) It wouldn't surprise me one bit if Shovel Knight made Odyssey look bad heading into its match with Ultimate.

Match 76: Devil May Cry 5 vs. Super Mario Odyssey

DMC5 7917
SMO 19681

I may have spoken too soon. Holy hell, DK, what's wrong with you?! This is an ass-whooping entirely commensurate with what you'd expect a Mario game to be doing to a Devil May Cry game. For comparison's sake, the weaker of the two DMC games in the last GotD would be projected to break 41% on Galaxy, the strongest Mario game in that bracket. Super Ma

Match 77: Grand Theft Auto V vs. Cuphead

GTAV 16450
Cuphead 10849

GTA V continues to be casualbait, trailing only BotW and Witcher 3 in R2 prediction percentage. Cuphead's classic Disney aesthetic continues to perform well here, and while it never stood a chance of winning, this was a solid showing, more or less. It couldn't quite break 60%, though.

Match 78: Persona 4 Golden vs. Red Dead Redemption

P4G 13879
RDR 13428

Upset alert! Upset alert! Just 31.05% of brackets got this right, compared with 88.36% who picked P4G to win in Round 1. RDR's Round 1 prediction percentage was sixth-highest...and its sequel's was eighth-highest, so yeah, clearly getting a lot of respect from the voters. But do you really want to pick an action-adventure game over an RPG on RPGFAQs?

...Apparently, yes. This match broke a trend that had stood throughout the entire first round and into the second: Unregistered voters made up the majority of Persona 4 Golden's raw votes! Red Dead Redemption won over registered voters 4737-4620, but Persona 4 Golden smashed it with a 4664-3985 advantage with the unregistered voters. It made for a 562-vote win in raw votes, but just 451 on the scoreboard...wait, those numbers don't add up.

4620 4664 9284 13879
4737 3985 8722 13428

Okay, the third column is indeed the sum of the first two columns...but with the first column, the registered voters, counting double, shouldn't it be 13904-13459? Makes me wonder if the "registered voters" column is in fact counting all registered voters, and Allen has to manually adjust it to make sure that only accounts that were created before the start of the contest are getting their votes counted double. Well, fixing it would only make things even closer, so it's no big difference. You see...Persona 4 Golden was still extending its lead almost right up until the end. Its maximum lead was 458 votes with just 10 minutes left in the match. That's two fewer than our long-time record, the legendary Frog vs. Liquid Snake match from 2004. But no big deal, right? That record's been broken like 15 times now. In this contest alone, three other matches had smaller max leads.

Except Octopath led during its match. So did Hollow Knight. And Celeste. And Mega Man in both of his losses in 2018, and Mario in his loss to Cloud in the same contest, and all of the others, even the ones whose matches only lasted 12 hours. Red Dead Redemption? Didn't. Never even came close. Usually, the updater at least captures a single-digit lead because it takes its first snapshot only a few seconds in, but P4G was already up 22-10 by the time it could. It was up by 74 at the freeze and hit triple digits by the 15-minute mark. The lead entered the 200s by the end of the second hour. This was, in fact, never in doubt. It just happened to stay close. Heck, the lead didn't even hit 400 until only about 2 hours remained in the match. Could've really made a record that would be tough to break if P4G hadn't surged right at the end. This is a match that looks a lot better on paper than it actually is.

---

I picked RDR thinking P4G would underperform due to being a port. Obviously I was wrong.

You write really well, but while you're allowed to write however you want, it does seem a bit heavy on analysis of past stats.

---