Board 8 > Today might be the day I finally block someone.

(+) Yeah! (+)

Superman is a perfectly fine/good character and iconic for a reason, but Man of Steel was bad (but only merely bad).
---

LeonhartFour posted...

Cloud/Seph is 51/49 in 2003 and 56/44 one year later

Comparing a 5% flip to a 10% (and being more realistic, more like 15%, since many were taking Samus to win comfortably) is a big difference though. And Mega Man vs Zero matching 9 years away is basically irrelevant because it's 9 years away. There's gonna be some variance-- hell you're probably gonna see a few percentages swing in non-SFF matches just from day to day, but not flipping a 60-40.
---

I'm just saying I don't think Mario 60/40s Samus every time.
---

What if you held the match today?
---

I wouldn't know what to do with myself because it would mean we're having a 1-on-1 Character Battle.
---

LeonhartFour posted...

Cloud/Seph is 51/49 in 2003 and 56/44 one year later.

Bracket voting and "inevitable" apathy.

God that 56/44 was heartbreaking for me.
---

the only negative thing I can say about Knuckles winning is it meant Squall didn't get to wipe the floor with Magus (although people would've chalked that up to SFF, too, so we still would've had Magus overestimation until he lost to Sandbag in 2008, which may have been even better had it played out that way)

it's the same problem with Vivi beating Mario because it meant Squall didn't get to

and Sonic getting announced in Brawl halfway through 2007 meant Squall didn't get to beat him that year either (and he came darn close anyway)

so many missed opportunities
---

Contest Stats and Discussion topic 832
---

LeonhartFour posted...

I'm just saying I don't think Mario 60/40s Samus every time.

Sure. I agree with that. Heck I think you could hold the match on back to back days and probably go anywhere from 55-65 for Mario. But I'm just saying in the years where Mario appeared "weaker" than Samus he probably still comfortably wins-- which may speak to SFF or it may speak to some Mario boost in 2005, but it could just speak to an error in the 2002, 2003, or 2004 stats and you wouldn't really know either way. It's just kinda presumptuous to say "well people who picked Samus were wrong but people who picked Mario to win easily were only right because Mario boosted" like you did.
---

it's kind of presumptuous to say Mario would easily win no matter when the match was held too but yeah
---

Well the result on paper being 60-40 means I'm significantly less presumptuous because we have first hand results saying that result not only happened but in a commanding fashion.
---

We've also seen results completely flip in pretty commanding fashion, too, like with Snake/Mega Man.

4-ways are significantly more volatile than 1-on-1, but we've even seen results flip from round to round there.
---

Scar, did I tell you Patricia McSamoff is a key NPC in a campaign I'm running?
---

Snake/Mega Man had actual reasons to change that people had speculated on before the contest even. It's not really the same as seeing a result and arbitrarily deciding that Mario had clearly boosted disproportionately to Samus.
---

It's also not arbitrary. It's based on data from past results. There may be some flaws in the interpretation of the data, but it's not arbitrary.

But demographics change year over year because the voter pool changes year over year. Who knows. Samus might even be able to beat Mario now because everything's so diluted. Chrono Trigger went from a game that couldn't even beat Mario 64 with Yoshi's Island in the poll to a game that appeared to be strong enough to win the entire contest, and there was zero reason for that to happen.
---

HanOfTheNekos posted...

Scar, did I tell you Patricia McSamoff is a key NPC in a campaign I'm running?

LMAO. That's hilarious. I'm amazed you remembered her!
---

LeonhartFour posted...

It's also not arbitrary. It's based on data from past results.

The data from past results is not arbitrary but deciding which of the past data to use, and how much to adjust (then) current data to line up better with past results is.

In addition making the decision that Mario boosted and Samus did not is indeed arbitrary and not based on anything but wishing to make a correction based on nothing other than to make the data more neat.
---

Lopen posted...

In addition making the decision that Mario boosted and Samus did not is indeed arbitrary and not based on anything but artificially wishing to make a correction to make the data more neat.

Samus may have boosted or dropped from 2004 to 2005. It's possible. It doesn't really matter for the purposes of the result since it's SFF-induced anyway. The idea that Mario boosted in 2005 is not arbitrary at all. You've got a direct result with Crono two years in a row that can show you that something changed from 2004 to 2005, and by a pretty significant amount since Mario went from 46.76% to 53.85% (which is a 7% swing). You can argue the change was entirely on Crono's part (He did appear to be a little weaker compared to previous years but still strong enough to make the finals of the contest) or based on other outside factors like bracket votes, pictures, etc., but it's not just a random guess based on nothing.

Lopen posted...

The data from past results is not arbitrary but deciding which of the past data to use, and how much to adjust (then) current data to line up better with past results is.

Also not arbitrary. I may choose the wrong data to focus on and thus be incorrect in my picks, but it's not arbitrary. Going back and looking at past results to try to deduce where it went wrong is also not arbitrary. Analysis by nature is not arbitrary. It is often based on the wrong suppositions and can't take every possible factor into account.
---

Analysis based on as little data points as we have that's rooted entirely on the results we're given and not on the characters may as well be completely arbitrary when you're talking in the realm of "I predict this match that was 60-40 to flip on its head." I mean I constructed like 5 ways to interpret the data that also result in Mario putting up better numbers than you'd expect. There's also a random variance inherent to each match, which further muddies up the data.

If you want to say Mario boosted, fine, but do it for a reason that is based on something more than him getting more percentage points in a match than you were expecting. That's all I'm saying. I never bought the Mario boosting theory because it wasn't really based on anything but people being upset that he won harder than they were expecting. People accept SFF as a thing that exists but then don't adjust Mario at all in 2002 vs Link despite him beating Cloud outright that year. Does that make sense? People assume Sephiroth and Cloud are consistent between 2003 and 2005, despite FFVII winning everything in 2003 in early 2004, and looking much weaker by mid 2004. Does that make sense?

There are all sorts of things in the early stats that are just chosen at the whims of whoever's reading the stats.
---

Lopen posted...

People accept SFF as a thing that exists but then don't adjust Mario at all in 2002 vs Link.

SFF wasn't even widely accepted as a real thing for the first couple of contests. It's why the 2002 and 2003 X-Stats don't have any adjustments in them. It's partially why people were so flabbergasted Sora destroyed his fourpack in 2004 after he got doubled by Aerith in 2003.

Lopen posted...

If you want to say Mario boosted, fine, but do it for a reason that is based on something more than him getting more percentage points in a match than you were expecting. That's all I'm saying.

Say Mario boosted, but don't use match results as an argument? Okay then. Is the "eye test" good enough for you? The dude just looked better in 2005 than he did the previous two contests, and he's got multiple common opponents across those years to show it. I haven't even been discussing Mario/Samus when I talk about him being stronger in 2005.

Lopen posted...

People assume Sephiroth and Cloud are consistent between 2003 and 2005, despite FFVII winning everything in 2003 in early 2004, and looking much weaker by mid 2004. Does that make sense?

Yes, the X-Stats involve making assumptions that often turn out to be wrong. I'm not even arguing against that. We don't have access to all the information. Link had a pretty big reason to be stronger in 2004 though with TP hype. He may be the most popular character on the site, but it doesn't mean he can't get even stronger. People were projecting a Link victory just based off of his first round match with CATS. Cloud did a little worse against Link in 2005 than in 2004, but it wasn't even by that much. His strength probably isn't static from 2003 to 2005, but it's close enough that assuming he was won't make much of a difference because he's easily beating anyone who isn't Link anyway.

I don't even bother doing SFF-adjustments when I do X-Stats anymore. People can just look at the raw data and draw their own conclusions as far as I'm concerned. Also, we barely get contests anymore, so what's the point? We haven't had a Character Battle in 5 years, so even if I had made adjustments in 2013, they'd be worthless by now anyway.

Lopen posted...

I mean I constructed like 5 ways to interpret the data that also result in Mario putting up better numbers than you'd expect. There's also a random variance inherent to each match, which further muddies up the data.

I mean yeah you did and that's fine

I don't even care that I'm wrong. I'm just interested into looking into reasons why I was wrong. A Mario boost in 2005 seems like the most likely solution to me based on the data I've looked at over the years. Could I be wrong? Sure. Like I said, I don't have access to all the information, but that's part of what makes it so fun to try to figure out in the first place. Even with all the data we have, the contest can still surprise us.
---

well there was nothing to make adjustments off of in 2002 anyway

the fact that the 2004 Games Contest X-Stats had SFF adjustments is baffling to me because those truly are just arbitrary
---

I'm not talking directly about the 2002 or 2003 stats for the record. I'm talking about later stats that used unadjusted 2002 and 2003 stats as adjustment references without first wondering if those stats themselves were to be fully trusted.

There's just a lot of cherry picking going on across the board. Like the stats were made when a lot of people who made em were in middle school and didn't fully grasp the theory behind why extrapolation is useful very well. It's interesting to talk about, sure, but I'm just explaining why I can't really take you seriously when you say "well people who picked Mario may have been singing a different tune had he not massively boosted in 2005" like that's some sort of undisputed fact. Cause it's not. No one called for a Mario boost before the contest, and there's a reason for that. The assumption he boosted was based on results not lining up with what you expected-- to me that's analysis that's laced with bias going in. You at times seem to care more about what makes the data line up with what you expect with the least amount of adjustments than determining why the data isn't what you expect.

I ain't saying the X-Stats are worthless cause they're clearly not (well they are now because every contest is rallyville but I digress), but they should be used to get a baseline for character popularity on the site rather than as a straight hierarchy.
---

Lopen posted...

but I'm just explaining why I can't really take you seriously when you say "well people who picked Mario may have been singing a different tune had he not massively boosted in 2005" like that's some sort of undisputed fact.

I haven't seen anyone dispute Mario boosting in 2005 until literally right now 13 years later.

Lopen posted...

I ain't saying the X-Stats are worthless cause they're clearly not (well they are now because every contest is rallyville but I digress), but they should be used to get a baseline for character popularity on the site rather than as a straight hierarchy.

dude I pick against the X-Stats "hierarchy" all the time and you know that so I don't even know why you're saying this like it's something I do
---

LeonhartFour posted...

well there was nothing to make adjustments off of in 2002 anyway

Technically they could be adjusted using matches from 2003 now. I don't see how using stats from one year later is any worse than using stats from earlier years to adjust later ones.
---

They could be, and I'm pretty sure it's been done, even. People tried to come up with an explanation for the "Western Division Factor" in 2002. There just isn't much of a reason to make adjustments for 2002 after SC2K3 because they're essentially "obsolete" at that point other than just to make the numbers look neater for the sake of it.
---

scarletspeed7 posted...

Lopen posted...

scarletspeed7 posted...

Ulti impugning one of my favorite film franchises.

Now I wanna block you

Steven Soderbergh is super-underrated film maker.

Also @Pokewars protect my Ocean's.

I will cut off the NOSE of anyone who says ill towards the grand Ocean's. There could be PLAYS written about the sheer beauty of the films.

@scarletspeed7
---

LeonhartFour posted...

dude I pick against the X-Stats "hierarchy" all the time and you know that so I don't even know why you're saying this like it's something I do

You do, but there is still a heavy stats sanity check for it. Like "these two guys have a close match on paper according to stats, but I'm going to pick Squall because of warning sign X here and warning sign Y there in these matches"

But as a guy who has made some pretty pivotal bold choices and hit home on them with pre-contest logic that was largely not based on X-Stats but instead based on just existing matches as written and some thoughts on how voter minds work, I take some exception to implying Mario > Samus was something that depended on a Mario boost or something when the margin was that high.

And don't get me wrong I get a lot more wrong than I get right (mostly because I don't use the x-stats enough), but I do think "Mario boosted" is dumbing down the reasons that existed to go for the pick a lot and frankly could very well be rewriting history for that matter cause we don't really know if Mario boosted. And it's dumbing it down in a very x-stat apologist type way. The reason it's not debated that Mario boosted is because Mario supporters were content to just be happy with their win and not argue too hard on the details. Most of the people picking Mario aren't heavy stat heads anyway.

And as far as dumbing it down goes so is "it's freaking Mario" to a lesser extent but I digress.
---

Side note derailing a topic about Ulti into contest discussion may be the most glorious derailing of all time. I thank you for this.
---

scarletspeed7 posted...

HanOfTheNekos posted...

Scar, did I tell you Patricia McSamoff is a key NPC in a campaign I'm running?

LMAO. That's hilarious. I'm amazed you remembered her!

The whole family is involved. Skip died @masterplum
---

UltimaterializerX posted...

Even had he beat Knuckles Squall would have pasted him.

Too bad we'll never know.
---

Also this Lopen/Leon discussion has made me want another contest real bad.
---

Sometimes I want another contest

Then I realized we haven't had a contest that wasn't guess the rally in the last 4 or 5 tries
---

Lopen posted...

You do, but there is still a heavy stats sanity check for it. Like "these two guys have a close match on paper according to stats, but I'm going to pick Squall because of warning sign X here and warning sign Y there in these matches"

Nah, I make my picks first and look at the stats for my rationale second, and I very rarely change my picks even after looking at stats. The stats are supports, not the foundation. I argue stats because it's fun, especially if you're making a pick that goes against them, not because I think they're some kind of authority. I look for reasons to pick favorites more than anything else.

Lopen posted...

I take some exception to implying Mario > Samus was something that depended on a Mario boost or something when the margin was that high.

Yeah, I didn't do that. I said it might have been possible Samus could have won under different circumstances. I never said anything definitive. You don't think there's any chance Samus could ever beat Mario under any circumstances, but I'm not quite that absolute on it.
---

To be clear I think Samus could beat Mario in some circumstances

Not circumstances that existed in 2003, 2004, or 2005, though.

If we had a new contest and we got our first new Metroid game in like 20 years to be released when the bracket came out? Yeah I'll take Samus.
---

Lopen posted...

Side note derailing a topic about Ulti into contest discussion may be the most glorious derailing of all time. I thank you for this.

This is why we need another contest imhotbqh

the great Leonhart/Lopen rivalry is forced to lay dormant far too long these days

Lopen posted...

Sometimes I want another contest

Then I realized we haven't had a contest that wasn't guess the rally in the last 4 or 5 tries

but yeah then I remember this

we haven't had a good contest in 8 years

well the 2015 Games Contest was mostly good outside of the rallies but you can't evaluate it without them
---

We only had rallies in two contests guys.

And despite them being horrendous garbage, we STILL almost managed to defeat Draven and had some bomb ass matches with Link and with Snake.
---

I will admit that David Hayter rallying for Snake has made me happy in a way few things have in contest history but even then we weren't allowed to have nice things for long

but even the contests where we didn't have rallies (Rivalries and Years) were two of the worst contests ever

although I think the 2013 contest was on pace to be pretty bad even without Draven
---

I still can't believe SBallen literally helped Draven win.

I just

It still enrages me to this day.
---

Like the only thing I can think of is that he was actively working with the rally. Even if the site wasn't down due to DDoS like he says (bullshit), there is ZERO REASON for him to not extend the match. We have precedent for it. We've even started matches over.

It was completely ludicrous and it's the worst thing that's ever been done by someone on this site.
---

Forget it, it's the worst thing ever done on this planet. Fuck Draven.
---

I like how irrelevant Draven is now.
---

HanOfTheNekos posted...

scarletspeed7 posted...

HanOfTheNekos posted...

Scar, did I tell you Patricia McSamoff is a key NPC in a campaign I'm running?

LMAO. That's hilarious. I'm amazed you remembered her!

The whole family is involved. Skip died @masterplum

Damn I was the one who died in the campaign
---

If I blocked people, Ulti would be the only one I blocked. And maybe Full Throttle.
---

I think it's pretty understandable at this point. And again, I just need a break from someone being so disingenuous. It doesn't make for an enjoyable b8 experience.
---

Also Corrik, what do you think about Jurassic World and Incredibles 2 being only the second pair of films in history to have back-to-back $100 million dollar weekends?
---

scarletspeed7 posted...

I think it's pretty understandable at this point. And again, I just need a break from someone being so disingenuous. It doesn't make for an enjoyable b8 experience.

I do suggest you mark every Ulti post in which any insult is made. I was moderated by a terrible moderator because I said Ulti was a "troll". Calling a user a troll is "flaming". No joke.

He is abusing alts to keep flaming people and eventually that will catch up to him.
---