Board 8 > Contest Stats and Discussion - Part 1309

(+) Yeah! (+)

Somehow, the LAW would probably prevail in those contests with Toon Link for the previous decade and Linkle for this decade.
---

Ganondorf's results imply Vivi beats Dante and probably outright wins a Character of the 2000's contest at this point so yeah you missed a big one there!
---

*goes away for 2 days*

*they discuss not only BYIG, but also a console contest*

I'm so proud of you. Years was like a breath of fresh air to me, personally speaking, since it was a nice little contest with no rallies. Sure, it had its flaws, but it had Guru, Oracle and whatnot to make it more fun.

I'd actually dig a console contest. Obvious entrants get in first, then we have a vote-in poll. I wouldn't be surprised if Atari 2600 ends up entering as well. It would be a good mid-contest for 2019, preparing for GotD2 in 2020.
---

If there's ever a time for 4-way matches to come back, it would be a console contest. That would be a lot of fun to decipher. Especially since a lot of the 1v1 matches would be too obvious.
---

thankfully there isn't a time for 4-ways to come back
---

oh, right, hard to remember FFIX is a 2000s game sometimes.
---

ZenOfThunder posted...

charmander6000 posted...

Character of the decade

2B for champion?

@MetalmindStats this could be a good next project?

Maybe I'll make this bracket at some point, with 64 entrants. Prepare for tons of fodder if so, though!

If Mae Borowski can win a match, it will have been worth it.
---

MetalmindStats posted...

ZenOfThunder posted...

charmander6000 posted...

Character of the decade

2B for champion?

@MetalmindStats this could be a good next project?

Maybe I'll make this bracket at some point, with 64 entrants. Prepare for tons of fodder if so, though!

Can't wait for the 2B vs. Geralt finals.
---

Playrate poll for OoT today:

https://gamefaqs.gamespot.com/poll/7423-
---

_SecretSquirrel posted...

MetalmindStats posted...

ZenOfThunder posted...

charmander6000 posted...

Character of the decade

2B for champion?

@MetalmindStats this could be a good next project?

Maybe I'll make this bracket at some point, with 64 entrants. Prepare for tons of fodder if so, though!

Can't wait for the 2B vs. Geralt finals.

The Witcher actually came out back in 2007! I'm pretty sure this would just be the latest contest idea with an obvious winner.

MetalmindStats posted...

_SecretSquirrel posted...

MetalmindStats posted...

ZenOfThunder posted...

charmander6000 posted...

Character of the decade

2B for champion?

@MetalmindStats this could be a good next project?

Maybe I'll make this bracket at some point, with 64 entrants. Prepare for tons of fodder if so, though!

Can't wait for the 2B vs. Geralt finals.

The Witcher actually came out back in 2007! I'm pretty sure this would just be the latest contest idea with an obvious winner.

I feel like you'd have to put in some sort of relevancy clause just to make a '10s contest even remotely interesting.
---

Yeah, unless there's some sort of objective way to determine which pre-2010 characters belong, I don't even think it's worth having, because no one wants to see a near-doubling in a finals match.

Post-2000 characters would probably be more interesting than post-2010 characters. Sure, Vivi/Dante/Auron would probably be heavily favored against everyone else, but they're not ridiculously far ahead of the next strongest entrants. Plus you don't need nearly as much obvious fodder to fill out the bracket.
---

Don't forget Sora! lol lol my bracket
---

I always thought FF 9 was from 1999! So yeah! Sora wins of course!
---

FFVIII came out in 1999. FFIX came out in 2000.
---

Damn it you are right! I will never forget when a big eartquake happened in my country and I was playing FF 8 just a while before it happened! And that earthquake was so big and we had a lot of dead people and damages of course.
But I was young so I cared mostly about FF8 anyways.
---

red sox 777 posted...

So I've been reading the stats topic from the Link/Cloud 2010 final today. There were a lot of people arguing that Link was a 99% lock to win any contest going forward, that only a perfect storm of highly unusual circumstances could stop him. Well, it's more than 8 years later, and Link hasn't won a single additional contest in his own name. The best victory he's had since was a GOTD contest, and that was almost 8 years ago. 2011 was the last time Link/LoZ won anything in any format. Link's win percentage since that Link/Cloud match is 50% at best, and 0% at worst, depending on how you want to count the different formats.

But we can say that rarely, if ever, have people in this topic been more wrong than the people who said Link had a 99% chance of winning any contest.

You mean all one of them? Since that contest, we've had GotD (won by a Legend of Zelda game), Rivalry Rumble (won by Link/Ganondorf), CBIX (decided by an outside force, and Link still finished second in raw X-Stats), BGE3 (again decided by outside forces; OoT made the finals), and the Years contest (won by 1998, a year headlined by OoT--and while it would be wrong to say that Link won that one since it's also the year of Pokmon Red/Blue and Metal Gear Solid, he certainly didn't lose it.)

The way I see it, Link is still the dominant force on this board. The problem was that "highly unusual circumstances" happened. If anything, the rarity of contests may have made things worse, because the hungrier we are for contests, the easier it is for us to get salty when outsiders show up to "ruin" things. Though I'm not really sure I consider Draven to have "ruined" anything; Link-Draven was absolutely electric, and it's a true testament to just how strong Link is that he very nearly managed to beat Draven anyway despite the fact that most of the regulars are tired of him always winning and as such, he would seem to be a tough candidate to "rally" for. In fact, many people still don't consider Draven's win over Link legitimate because of the problems the site had that day. That seems like the definition of "a perfect storm of highly unusual circumstances".

8 years? Don't give me that crap. This is just the second character battle since that statement was made. I feel that the statement was accurate, but the very next character battle after said statement featured exactly the situation described as to how he might be defeated.
---

LeonhartFour posted...

FFVIII came out in 1999. FFIX came out in 2000.

VI 94
CHRONO TRIGGER 95
SUPER MARIO RPG 96
VII 1997
TACTICS 1998
VIII 1999
IX 2000
X 2001

Never Forget.
---

A true golden age
---

So I had a dream the other day, something I never told when 2B vs. Bowser happened IRL.

In my dream, after "waking up" the next day to check the results, it turned out that Bowser had barely beaten her by 50.9%, but 2B still went ahead to the next round instead of Bowser. She was beating her opponent by 52% IIRC, and unlike the previous rounds, she also had a really long name (supposedly her full name) in dark green letters, which was a reference to her color mix (white and dark green, as opposed to real-life's black and white) and it fitted her well.
---

Using 2B's real name would be quite the spoiler.
---

So in 5 years we had 4 main FF games. Man things have changed for sure.
---

Rank the following five characters in order of contest strength:

Master Chief
Commander Shepard
Lightning
Nathan Drake
Catwoman
---

Safer_777 posted...

So in 5 years we had 4 main FF games. Man things have changed for sure.

They cheated because FFIX wasn't supposed to be a main entry!

Which means we haven't had a mainline FF since 2010.
---

well Tactics wasn't supposed to be an FF game either but we allow it
---

Do you guys think K Rool could have won with this remix behind him?

https://www.youtube.com/watch?v=8zC6qJpBkcI" data-time="
---

maybe if you replaced all the lyrics with this:

https://www.youtube.com/watch?v=kZwhNFOn4ik" data-time="
---

LinkMarioSamus posted...

Rank the following five characters in order of contest strength:

Master Chief
Commander Shepard
Lightning
Nathan Drake
Catwoman

Lightning > Chief > Drake > Shepard >>>> i see what u did there
---

With regards to some inevitable repeat matches in the loser's bracket...

We've seen repeat matches before. In the case of Pokemon GSC vs. Xenogears, Xenogears won by more in the repeat match after the glitch allowing only registered voters was fixed. Of course, we can't glean anything from it since we know Pokemon GSC obviously was favored by the registered voters anyway.

The 2006 battle royale is another example. Granted, these rematches all happened in consecutive days, so I don't know how relevant these are, either. Still, it's notable that the wins who win are more likely to do better in the rematches. Link fans were increasingly enthusiastic over the first three days of the Battle Royale. Meanwhile, Mario and Sephiroth fans basically said "forget it" and stopped voting. Solid Snake's fans were surprisingly resilient, though, as Snake continued to do fairly well and ended up beating Sephiroth on the third day. To think that a Smash-boosted Snake, on the first day, couldn't compete with a Sephiroth being leeched by Cloud! Things were different back then....

Oh, yeah, and we've saw plenty of rematches in 2007-2009. Is there anything worth mentioning there?

I've heard the idea that the Noble Niners may be stronger because of pent-up strength (excitement of finally seeing these characters). I'm not convinced that will happen. We saw Mario do just fine against Sephiroth in the champions bracket in 2005, even though Sephiroth was waiting all that time. Granted, voters could have been pretty excited about Mario winning the entire bracket, so maybe it's possible that the excitement about Mario's championship cancelled out existing pent-up strength for Sephiroth.

Probably not a single thing I said will be of any help in figuring out how those rematches will go!
---

the difference is that most of those weren't 1-on-1 and the ones that were (you can include Hayabusa/Jill 2004) had mitigating circumstances as to why the original result wasn't valid and they never actually finished anyway
---

I think Phoenix would be a force in a best character of 2000-2009 contest
---

Yeah sure.
---

I could see pity votes being a factor in repeat matches, say a character got SFF in the first match, but the rematch has the character winning by less. Of course, that probably won't change the winner of any matches and may only be a factor in matches where the winner won by a lot already.
---

I could see people voting against Nintendo or Square in losers in an attempt to overturn the status quo. Like, allow Crono to flip against Cloud.

I don't see how that would work since both Crono and Cloud are Square characters.
---

I don't think most voters are that cognizant of what developers made which games.
---

Yeah, I'm not sure how many voters could say what developers made Nier: Automata or The Witcher III, and their characters got 1 seeds this contest.
---

LusterSoldier posted...

Link versus Cloud posted...

How high are people going for tiebreakers in second chance? Think we are going to get moar votals in the later matches?

It should be slightly higher compared to the first 4 rounds, but I wouldn't expect the average to be any higher than 23000. We're barely averaging over 20000 through the first 4 rounds. An ideal tiebreaker should be in the 610000 to 700000 range.

Even though the overall average is that low, the average for round 4 was close to 30,000 per match wasn't it? Isn't that what we should be comparing the final 16 to since it will be bigger matches with 2 per day format like those? I was thinking something closer to the 900,000 range.
---

Nier: Automata is made by Platinum Games like Bayonetta and Revengeance
---

What system is used when adjusting the raw x stats? I'm curious
---

Link versus Cloud posted...

Even though the overall average is that low, the average for round 4 was close to 30,000 per match wasn't it? Isn't that what we should be comparing the final 16 to since it will be bigger matches with 2 per day format like those? I was thinking something closer to the 900,000 range.

For the tiebreaker question, you need to put in a guess that does not factor in the registered user bonus. We were averaging about 20000 votes per match before the registered user bonus was applied to the raw votes. I guess the later rounds will probably have slightly higher vote totals, but I think it's pretty unlikely we'll average more than 23000 raw votes per match for the entire Legends/Loser bracket combined.

My recommended guessing range for the tiebreaker question is somewhere between 610000 and 700000.
---

Are we getting one or two matches per day tomorrow?
---

Advokaiser posted...

Are we getting one or two matches per day tomorrow?

Two matches every round, except for a couple semi-finals/finals one which will be one a day.

haloiscoolisbak posted...

What system is used when adjusting the raw x stats? I'm curious

Adjustments are largely made using results from previous contests

although it'll be a bit of crapshoot this year because the last Character Battle was 5 years ago and Draven made a mess of it, but at least the double elimination format will make things a bit easier
---

I think double elimination should make it easier because like, in Ganondorf's case you can just discard the match against Link and use the other one for stats.

Problem is that all other matches will actually have more data than they need, so they'll have to take the avarage between the two or something like that.

well it'll get confusing if a character actually flips a result and wins a rematch but that's part of the fun!
---

Well ideally matches will only flip a result if the first one was super close, so the characters will end close enough in the stats and we'll be able to atribute any future flip to intangibles.

But if a match flips from like 55-45 to 49-51... then I don't know what to think. Could be the final confirmation that xstats are a hack.

Which, considering the contest so far, doesn't seem like an unlikely scenario.