Board 8 > Contest Stats and Discussion - Part 1270

(+) Yeah! (+)

Remember when all of the below stuff was relevant?

~*creativename's contest site (all things contest!)*~
http://www.gamefaqscontests.com/drupal/

~*The Board 8 Wiki (lots of useful contest and board information, including all past Post-Contest Analysis from Ulti, transience, Ed Bellis, and others)*~
http://board8.wikia.com/

~*List of All Polls (a search bar is at the bottom)*~
http://www.gamefaqs.com/poll/index.html

~*NGamer64's Archive Sites (good stuff!) and (LOL) X-Stats Sim*~
http://www.thengamer.com/
http://thengamer.com/xstats

~*GameFAQs Contests Hall of Fame*~
http://board8.wikia.com/wiki/GameFAQs_Contests_Hall_of_Fame
http://board8.wikia.com/wiki/GameFAQs_Contests_Match_Hall_of_Fame

~*Character Contest Histories*~
http://board8.wikia.com/wiki/Category:Contest_Histories

~*Simple Explanation of Extrapolated Standings*~
A = Strongest Character
B = Character Weaker than A
C = Character Weaker than B
To figure out a character's Xsts Percentage ---> [(CvB)(BvA)]/50 = CvA
To compare how C would do against B ---> [(CvA)/(BvA)]*50 = CvB
To figure out how B would do against A ---> [(CvA)/(CvB)]*50 = BvA

~*All the Match Pics*~
http://www.gamefaqscontests.com/gallery/index.php

~*Leonhart4's Trend Charts*~
https://spreadsheets.google.com/spreadsheet/ccc?hl=en&key=tOGmynfNIiNy5VukpEF-PdA&hl=en#gid=0

~*Daily Vote Trends - An Explanation for Dummies*~
http://board8.wikia.com/wiki/Daily_Vote_Trends

~*Acronyms and Percentages for Dummies*~
http://board8.wikia.com/wiki/Acronyms_and_Percentages

~*Say What? Common Stats Topic Lingo Defined*~

SFF (Same Fanbase Factor) - Same Fanbase Factor is the theory that, if two contestants share a common fanbase, the weaker of the two options will underperform in a direct matchup. For instance, Link was expected to defeat Ganon with 65% of the vote in 2004, based on their 2003 values. Instead, Link collected near 88% of the vote. This is the best example of SFF we've ever seen. However take some SFF labels with a grain of salt, as many people will slap it onto any match that doesn't make perfect sense.

Extrapolated Standings (X-Stats) - The mathematical "strength" of a contestant that's determined based on their performance relative to the rest of the field. This number is typically based on the contest entrant's loss, but adjustments are sometimes made. See above for a watered down explanation for how the stats are calculated.

LPF - Last Place Factor, which a term used in matches that have three or four entrants. Usually when there is an obvious choice for last place, that entrant performs weaker than usual since voters are smart enough these dates to put their votes on something that matters. A good example is Link/Shepard/Draven.

ASV - After school vote, which is usually the biggest trend in a poll other than the first hour and the morning vote.
---

Not_Wylvane posted...

Remember when all of the below stuff was relevant?

hey man I update those trend charts daily
---

I guess we need this now.
---

Do we just keep all of that text as a relic to the years gone by?
---

swirIdude posted...

Do we just keep all of that text as a relic to the years gone by?

Clearly, we need the explanation of Last Place Factor for this contest.
---

2006 is losing by enough that I can't say for certain it would win with a fair picture. Even if it would lose, I still wish 2006 was given a fair chance to win the match.
---

metaIslug posted...

doing pretty good .still perfecto including today

---

We can probably trim to the below and be fine. I'll leave it up to the rest of the topic for consensus, but the OP has been needlessly bloated for a while.




~*creativename's contest site (all things contest!)*~
http://www.gamefaqscontests.com/drupal/

~*The Board 8 Wiki (lots of useful contest and board information, including all past Post-Contest Analysis from Ulti, transience, Ed Bellis, and others)*~
http://board8.wikia.com/

~*List of All Polls (a search bar is at the bottom)*~
http://www.gamefaqs.com/poll/index.html

~*NGamer64's Archive Sites (good stuff!) and (LOL) X-Stats Sim*~
http://www.thengamer.com/
http://thengamer.com/xstats

~*Character Contest Histories*~
http://board8.wikia.com/wiki/Category:Contest_Histories

~*All the Match Pics*~
http://www.gamefaqscontests.com/gallery/index.php

~*Leonhart4's Trend Charts*~
https://spreadsheets.google.com/spreadsheet/ccc?hl=en&key=tOGmynfNIiNy5VukpEF-PdA&hl=en#gid=0

~*Daily Vote Trends - An Explanation for Dummies*~
http://board8.wikia.com/wiki/Daily_Vote_Trends

~*Acronyms and Percentages for Dummies*~
http://board8.wikia.com/wiki/Acronyms_and_Percentages
---

I consider this similar (but not as bad) as the Vincent picture in Characters 2013. Not good but changing to something good wouldn't change the outcome (but people will still complain anyway)
---

I think the explanations of SFF and the X-stats should stay since we actually still use those terms, but the Character Histories could probably go (has that even been updated since 2008?).
---

I would like to suggest a modification to part of the opening post. The URL to access the list of all polls could be trimmed down a bit. Here is my suggested change:

~*List of All Polls (a search bar is at the bottom)*~
https://www.gamefaqs.com/poll
---

on the pics link I see 2 for 2015. I assume one was for the wildcard round(I forget the pic they used that round) and one for htis round. One has undertale and one doesn't. Is undertale on this round's pic lineup?
---

haloiscoolisbak posted...

on the pics link I see 2 for 2015. I assume one was for the wildcard round(I forget the pic they used that round) and one for htis round. One has undertale and one doesn't. Is undertale on this round's pic lineup?

The 2013 match picture has an error and says 2015 in the picture instead of 2013. Looks like I might have to PM SBAllen about this to get the mistake corrected before the 2013 match starts.
---

NowItsAngeTime posted...

I consider this similar (but not as bad) as the Vincent picture in Characters 2013. Not good but changing to something good wouldn't change the outcome (but people will still complain anyway)

What was his pic in 2013?

http://www.gamefaqscontests.com/gallery/albums/cb9/5229.jpg
---

LeonhartFour posted...

http://www.gamefaqscontests.com/gallery/albums/cb9/5229.jpg

The quote underneath makes it so much better/worse.
---

I totally forgot the quote.

I still remember Tharja's while she looked at Aerith.
---

charmander6000 posted...

I totally forgot the quote.

I still remember Tharja's while she looked at Aerith.

Mega Man X vowing to stop this madness as Draven just ran right over him was great too.
---

I think 2006 would win with a better picture just because I think pictures are more important here. Don't make the symbols Oblivion vs. Skyrim / include KH2 instead of Gears and I think 2006 takes it.
---

LeonhartFour posted...

http://www.gamefaqscontests.com/gallery/albums/cb9/5229.jpg

Ah thanks.

He looks a lot like Keanu Reeves. Did people note that back then?

creativename posted...

LeonhartFour posted...

http://www.gamefaqscontests.com/gallery/albums/cb9/5229.jpg

Ah thanks.

He looks a lot like Keanu Reeves. Did people note that back then?

I thought he looked more like Phoenix Wright than Phoenix Wright did.
---

That picture of Vincent was so bad that it affected his trends. He did not have the usual terrible FFVII board vote and even his night vote was much weaker than it usually would be. Seems like a lot of voters didn't think it was FFVII Vincent in that picture.
---

I like SFF and X-stats in the opening post. I'm not sure when LFF was removed (and LPF was inserted?) but LFF is also an extremely useful term.
---

FFVII trends have been becoming less pronounced for years and Vincent never quite had Cloud/Sephiroth trends - remember he didn't lose a single hour to Ganondorf in 2006 despite only winning with 52%.
---

Trends in general are becoming less pronounced lately now that we don't have the younger voters necessary to support an ASV. The strong trends we used to have prior to 2011 were often produced by a combination of Age based trends and Geolocation trends.

There are 3 known ways of producing trends in polls:

Age based trends - The concept that a person's age strongly affects how they vote. The best known example of age based trends would be the age poll itself, but even that poll's trends have become less noticeable as the site's average age has continued to increase.

Time based trends - The concept that trends can vary quite a bit depending on what time of the day you vote. These trends should not be confused for age based trends or Geolocation based trends because Time based trends apply to all parts of the world. If a person's age makes them unlikely to vote at a particular time of the day, this type of voting does not fall under Time based trends.

Time based trends don't really appear in contest matches. The best known example of Time based trends is a poll asking what time of the day you usually wake up every day. In that kind of poll, a person who generally wakes up at 7:00-7:59 AM wouldn't be voting for that option at 3:00 AM as you would expect them to be sleeping at that time (and this applies to voters of all regions, not just North America).

Geolocation based trends - The most common form of trends that appear in polls now that Age based trends are mostly dead at this point. These trends reflect the voting differences between different regions. It is now considered the primary driver of trends over the course of a poll, since the geographic distribution of votes can change very sharply over the course of a poll. The continent poll best represents Geolocation based trends.
---

red sox 777 posted...

FFVII trends have been becoming less pronounced for years and Vincent never quite had Cloud/Sephiroth trends - remember he didn't lose a single hour to Ganondorf in 2006 despite only winning with 52%.

I miss those days.

Also, is Vincent/Ganondorf still the most loaded round 1 match ever?

I guess 1994/2004 is up there now, too...!
---

http://www.gamefaqscontests.com/gallery/albums/bestyear/r1_2016.jpg

This is so weird to me. I look at it and think "Where's Doom? Where's Titanfall 2? Where's The Witness?"

And then I remember this is GameFAQs, and that Dark Souls 3, FFXV and SFV actually probably would beat all those games on this site. Well, maybe not Street Fighter. C'mon!
---

needs Uncharted 4
---

Wait. How the shit is 2006 losing?
---

LeonhartFour posted...

red sox 777 posted...

FFVII trends have been becoming less pronounced for years and Vincent never quite had Cloud/Sephiroth trends - remember he didn't lose a single hour to Ganondorf in 2006 despite only winning with 52%.

I miss those days.

Also, is Vincent/Ganondorf still the most loaded round 1 match ever?

I guess 1994/2004 is up there now, too...!

Crono/Missingno is the only one that comes to mind.
---

LusterSoldier posted...

Trends in general are becoming less pronounced lately now that we don't have the younger voters necessary to support an ASV. The strong trends we used to have prior to 2011 were often produced by a combination of Age based trends and Geolocation trends.

There are 3 known ways of producing trends in polls:

Age based trends - The concept that a person's age strongly affects how they vote. The best known example of age based trends would be the age poll itself, but even that poll's trends have become less noticeable as the site's average age has continued to increase.

Time based trends - The concept that trends can vary quite a bit depending on what time of the day you vote. These trends should not be confused for age based trends or Geolocation based trends because Time based trends apply to all parts of the world. If a person's age makes them unlikely to vote at a particular time of the day, this type of voting does not fall under Time based trends.

Time based trends don't really appear in contest matches. The best known example of Time based trends is a poll asking what time of the day you usually wake up every day. In that kind of poll, a person who generally wakes up at 7:00-7:59 AM wouldn't be voting for that option at 3:00 AM as you would expect them to be sleeping at that time (and this applies to voters of all regions, not just North America).

Geolocation based trends - The most common form of trends that appear in polls now that Age based trends are mostly dead at this point. These trends reflect the voting differences between different regions. It is now considered the primary driver of trends over the course of a poll, since the geographic distribution of votes can change very sharply over the course of a poll. The continent poll best represents Geolocation based trends.

Dont forget rally trends
---

Big Bob posted...

http://www.gamefaqscontests.com/gallery/albums/bestyear/r1_2016.jpg

This is so weird to me. I look at it and think "Where's Doom? Where's Titanfall 2? Where's The Witness?"

And then I remember this is GameFAQs, and that Dark Souls 3, FFXV and SFV actually probably would beat all those games on this site. Well, maybe not Street Fighter. C'mon!

Doom and Titanfall 2 were fantastic games, and probably the first FPS games in at least 5 years to even excite me enough to play. The Witness had potential, but the complete lack of a colour-blind mode meant it was unplayable for me at a certain point
---

NowItsAngeTime posted...

Dont forget rally trends

Well yeah, rally trends are a thing, but they are extremely uncommon and only happen in a very small percentage of polls. I didn't include rally trends as they appear very infrequently.
---

I suck at photoshop. Someone make a 2017 match pic with Zelda, Persona, Horizon, and Nier for me.
---

what "trends" would you even put for rallies

"if they happen the contest is over"
---

Big Bob posted...

I suck at photoshop. Someone make a 2017 match pic with Zelda, Persona, Horizon, and Nier for me.

Challenge accepted
---

1994/2004 could be this year's Spyro/Morrigan. If the prediction percentage isn't below 50%, it is likely under 60%.
---

Who thinks 1998 will finish above 90%?
---

Whoa, 2004 was the very slight favorite.

2011, too.
---

another 37-0 run to open a match

this contest
---

39 votes
---

1998 went about 40 votes without a single dissenter. It happened again!
---

huh, 2016 did well with the early vote.

so much for pity votes against 1998
---

Aww yeah, just saw it at 28-0, then 40-1.
---

41-1. Ouch.
---

1992 lookin' kinda bad

or 2016 looking really good
---

this kinda makes me feel like 1995 might be in trouble? probably not logical
---

huh, 1994 had less brackets than 2006
---