Wake up to Gordon over 40%. Some traditions do never die <_<

---
RIP: Tnote827 - 4/4/2005
SC2k5 - 27/29

Leon's percentage has been rising slowly but steadily...could he finally crack 60% on Gordon?
---
"i started playing mario brothers as luigi and threw the controller down in disgust in the middle of the game because he lacked character"--gokujont

Ulti, you were calling Leonhart "X-stat blinders" earlier, but a more accurate name would be "Hindsight bias". Seriously, ever since Magus/Knuckles, he's done nothing but throw Sam Fisher in our faces, but beforehand, he thought he was being optimistic saying that Knuckles would crack 40%. That's hindsight bias at it's finest.
---
That... was her final mission. And like a true soldier, she saw it through the end.
ZSB [aX]

While Leon is gaining, he's actually lost a tiny bit of percentage in the last hour. Looks like my 59.99% Oracle pick won't be too bad. =)
---
Board 8: Where people treat each other right.

They say race car drivers lose half a second per lap once they get married... methinks Ulti's finding himself in a similar situation.
---
"You can't have gone over a kilometer up. 1500 meters, tops." -Samus Aran
"I'll express my love to you slowly, later." -Setzer Gabbiani

Ah. The universe will remain soundly running for one more year at least.
---
Go Seahawks!
Playing: Secret of Mana, Final Fantasy VII, Knights of the Old Republic

So based on this universe implosion theory, why does board 8 nominate GF every year? Or does the theory run that the universe will also implode if GF doesn't make it into the contest?
---
SSBM. Rock rock on!

Board 8 doesn't nominate Gordon.
---
"You can't have gone over a kilometer up. 1500 meters, tops." -Samus Aran
"I'll express my love to you slowly, later." -Setzer Gabbiani

I'll be happy when the stats don't have fanboys who try to adjust for every factor under the sun in an attempt to have them make sense whenever something goes wrong. When the stats predict something, all is peachy. When the stats are off by a fair amount, everyone scrambles around like ants in an effort to keep them looking as perfect and accurate as possible. The thought of voters simply changing their minds -- Or, get this! -- new voters coming to vote in the polls and throwing out the ridiculous assumption of transitivity is not allowed.

All of the stats are unreliable at best. They're a good medium for a general idea, but not much else. Fun to calculate, fun to look at, fun to debate, but they'll cause your bracket to go through hell and are void of factoring in human variables and ranges.

Ulti is my new god. My thoughts EXACTLY.
---
Chun-Li > Bowser book it - lol_internet
Vote for Kirby in SC2K5!

Is Ulti on drugs or something?

The stats are not "unreliable at best." Wow, they were off for two matches. That means the stats are ALWAYS inaccurate!

In reality, the stats are right on the nose nine times out of ten.
---
"Thank you, Mario. But our princess is in another castle."
-Toad in Super Mario Bros.

Oh no? Then GF is a Ceej input or something then I take it.
---
SSBM. Rock rock on!

You are aware that there is no actual set group of people who comprise "Board 8", and that all registered users can nominate characters, right?
---
"When I was young people in my village would ring a bell every time I was angry to tell everyone to go into their homes until I calmed down." - Igor

He's just bitter that Magus lost. I always figured that "if he loses, I want it to be to Knuckles" attitude was a bunch of bull.
---
"You can't have gone over a kilometer up. 1500 meters, tops." -Samus Aran
"I'll express my love to you slowly, later." -Setzer Gabbiani

Eh, I figured board 8 would have a good deal of the noms, but I guess that's true. (Like you guys got Manny in, right?)

Personally I've always missed the nominations, never seem to catch the news for them.
---
SSBM. Rock rock on!

Only two matches? Out of all the matches involving two contest vets this year, the stats have been significantly off in:

Ganondorf/Yuna
Zelda/Vivi
Master Chief/CATS
Kefka/Vercetti
Magus/Knuckles
Luigi/KOS-MOS

That's all that involve two vets except Zero/Ryu and Mega Man/Conker.
---
That... was her final mission. And like a true soldier, she saw it through the end.
ZSB [aX]

So 4% is significantly now?
---
"You can't have gone over a kilometer up. 1500 meters, tops." -Samus Aran
"I'll express my love to you slowly, later." -Setzer Gabbiani

Wait a minute, I think I can safely say that where the X-Stats have applied in this contest, they have *not* correctly predicted the outcome within a reasonable margin 9 times out of 10. At least, not with the way we've been interpreting them.
---
"i started playing mario brothers as luigi and threw the controller down in disgust in the middle of the game because he lacked character"--gokujont

Zelda/Vivi was more of a Zelda increasing.

MC/Cats was more of CATS having his face pic. Use Spring 2005 CATS and MC's 2003 level and it predicts the match almost on point.

Kefka/Vercetti, using his unadjusted value, was already expected to be close, at least in the 50s%
---
"I dont know what I am talking about? I know more about gaming than 90 percent of the users on the internet." -xmagxus

4% is quite a large margin of error.

And I forgot DK/Sam, which was pretty close, actually.
---
That... was her final mission. And like a true soldier, she saw it through the end.
ZSB [aX]

What's the difference between adjusted and unadjusted? They both were in the 20XX division.

I'll agree on one point though, and that is, some of you people just expect too much out of the stats. I mean, to call Ganondorf/Yuna "way off", geesh... especially with FFX-2...
---
"You can't have gone over a kilometer up. 1500 meters, tops." -Samus Aran
"I'll express my love to you slowly, later." -Setzer Gabbiani

While I post that, someone just confirms what I'm saying.
---
"You can't have gone over a kilometer up. 1500 meters, tops." -Samus Aran
"I'll express my love to you slowly, later." -Setzer Gabbiani

Zelda/Vivi can't really be used to judge the stats, since this was the first time they were present in the same year.
---
"When I was young people in my village would ring a bell every time I was angry to tell everyone to go into their homes until I calmed down." - Igor

And to be honest, there's a big chance Ganondorf and Zelda both increased along with Link from 2003 to 2004, from the looks of it.

The stats are not just a guideling, they're a puzzle that need solving.
---
"You can't have gone over a kilometer up. 1500 meters, tops." -Samus Aran
"I'll express my love to you slowly, later." -Setzer Gabbiani

From: Acronym | Posted: 8/17/2005 11:30:58 AM | #322
Zelda/Vivi can't really be used to judge the stats, since this was the first time they were present in the same year.

Which is part of the point people like Ulti, Lucid Faia, and myself are trying to make. Some people try to force year-to-year transitivity where it's impossible to do so.
---
That... was her final mission. And like a true soldier, she saw it through the end.
ZSB [aX]

Ermm...isn't adjusting / explaining for the discrepancies in the xsts an indication that they are not as good of an indicator than expected?

Actually, to be purely objective, one should compare how the xsts do on a whole after the contest is over. So you have the xsts projected % on one side, the actual 2k5 matches on the other, and see how well they predict then.
---
SSBM. Rock rock on!

Darn, I keep forgetting they were in the same division.
---
"I dont know what I am talking about? I know more about gaming than 90 percent of the users on the internet." -xmagxus

Some people try to force year-to-year transitivity where it's impossible to do so.

Really? I've never seen anyone claim year-to-year transitivity. It's hard enough to claim it within a single year...
---
"When I was young people in my village would ring a bell every time I was angry to tell everyone to go into their homes until I calmed down." - Igor

That'd be very difficult. I suggested going with ranges for the X-Sts, but that was rejected.

Fact of the matter is, in 2003, following the X-Sts put you on the leaderboard prior to Link vs. Cloud. Trust me, I know. They didn't really falter in 2004 either.
---
"You can't have gone over a kilometer up. 1500 meters, tops." -Samus Aran
"I'll express my love to you slowly, later." -Setzer Gabbiani

To be fair, a lot of '04 was guess the SSF.
---
Ahem! There's LETTUCE on my boots!
http://www.geocities.com/lettucekefka/

I'm gonna make a statement to see how well people agree with this sentiment, 'cuz it's how I feel. Ahem:


The extropolated stats are, barring SFF and other things we discover in hindsight (bad adjustments, for example), generally very accurate...for the contest they're derived from; it's when they're carried over to the next contest that puts them more in question. I do believe that, last summer, Master Chief and Liquid Snake would have also had a very close match much like Frog did with the both of them. However, with this summer, I would doubt MC/Liquid would be nearly as close this year as it could have been last year, despite Halo 2 seemingly doing nothing and Liquid seemingly under-performing against Sephiroth...close, but nothing like it probably would have been last year.

Even then, though, despite using old contest data for new contests, we can usually solve the problems pretty easily...sometimes too late, but nonetheless. A big example's Magus/Knuckles; we've had misleading data on Magus since 2k3 simply because of ONE match of his against Link...since then, Tidus has arguably shown in his last two matches (KoaC and Mega Man) that he was over-rated, and Sam Fisher has done the same in his last three matches (Donker Kong, Samus, and Mr. GFNW). With Magus just under Knuckles, the Kefka four-pack lines up much better too. The only thing to challenge this at all is that Ganon still kept close to his strength he was assumed to have...and one can argue that Ganon got a good-sized boost like Link and Zelda did. It's just that we never thought seriously enough about all of this because of Link/Magus (which was a good showing, no doubt).

So, in other words, I think the x-stats are great for the contest they're created from, and they're not bad for a contest or two after that. However, they should still always be taken with a grain of salt and should have everything considered so you can make a valid judgment for yourself as you check them out so you don't make as big of mistakes.


Sorry for the Great Wall Of Text.
---
Hillary Clinton: We need to treat violent video games the way we treat tobacco or alcohol.
Lewis Black: Namely as valued contributors to our re-election fund.

Seriously, ever since Magus/Knuckles, he's done nothing but throw Sam Fisher in our faces, but beforehand, he thought he was being optimistic saying that Knuckles would crack 40%.

No, I didn't, otherwise I'd have never taken Squall over Magus in the second round. I said Magus was overrated before this contest began, and used Sam Fisher, Luca Blight, and Tidus all as examples to prove it. I said right before the match started that Knuckles would break at least 45%.

Either you don't remember what I've said very well, or you're confusing me with somebody else. I never thought 40% was "optimistic." If Magus had broken 60%, I would've been surprised.
---
It's just not the same without this name.
Faithful supporter of Squall Leonhart in SC2K5

Leonhart speaks the truth. I was also preaching that Magus was overrated since the contest began, just not as vocally.
---
For the sake of simplicity, I'm going to refer to Gordon Freeman versus Tanner as Schrodinger's Match. -Phoenix Flattener

So, how do we feel about today's results? Leon's a flop? Half-Life 2 gave Gordon a nice boost? Gordon Freeman would break 40% on a Link/Cloud/Sephiroth splice?
---
Board 8: Where people treat each other right.

GFNW is a puzzling force indeed. I can't tell what to make of it. I want to see Gordon face an established midcarder next year (somebody like Knuckles) to see how he'd do.
---
It's just not the same without this name.
Faithful supporter of Squall Leonhart in SC2K5

He'd break 40%.
---
Board 8: Where people treat each other right.

I, for one, would like to see GFNW simultaneously coincide with Gordon Freeman ranking in the top five characters... <_<
---
"i started playing mario brothers as luigi and threw the controller down in disgust in the middle of the game because he lacked character"--gokujont

Many of us preached Magus to be over-rated...none of us enough though. Heh, despite me thinkin' he was over-rated, I still took him to win the division...shame on me.
---
Hillary Clinton: We need to treat violent video games the way we treat tobacco or alcohol.
Lewis Black: Namely as valued contributors to our re-election fund.

I, for one, would like to see GFNW simultaneously coincide with Gordon Freeman ranking in the top five characters...

...in a five-character contest.
---
Hillary Clinton: We need to treat violent video games the way we treat tobacco or alcohol.
Lewis Black: Namely as valued contributors to our re-election fund.

Well, the only problem with us not seeing Magus being overrated "enough" is that Tidus ranked above Knuckles last year. At least to me it was.
---
Squall Leonhart's Road to the Summer Contest 2005 Championship
Round 2: (3) Mag...Er...(6) Knuckles the Echidna

I, for one, would like to see GFNW simultaneously coincide with Gordon Freeman ranking in the top five characters... <_<

It can easily happen, but would require a 16 seed in order to face Link or Cloud in Round 1. I just can't imagine Gordon seeded 16th. >_<
---
Board 8: Where people treat each other right.

Unless the Xsts predict a match down to a perfect percentage, they are not correct.

From: Slowflake | Posted: 8/17/2005 11:45:31 AM | Message Detail | #305
They say race car drivers lose half a second per lap once they get married... methinks Ulti's finding himself in a similar situation.

??????????

And by the way, I was quite serious about the Knuckles thing. Don't you remember that Knuckles > Tails > Sonic > Shadow topic I made?

~*ST*~
---
Winner of the Spring 2004 "Best. Game. Ever." Contest
"If my fate is to be destroyed... I must simply laugh!" -Magus

From: Slowflake | Posted: 8/17/2005 12:34:56 PM | Message Detail | #328
That'd be very difficult. I suggested going with ranges for the X-Sts, but that was rejected.

I only rejected that because Xst whining is hilarious to read. But if you want me to be serious, how is it even possible to put ranges in the stats at all? The only thing I can think of would be to list each charactter's lowest and highest strength as their range. Anything other than that is a guess.

~*ST*~
---
Winner of the Spring 2004 "Best. Game. Ever." Contest
"If my fate is to be destroyed... I must simply laugh!" -Magus

Unless the Xsts predict a match down to a perfect percentage, they are not correct.

You know that's not going to happen. With so many unknown factors and not every single person votes in the same poll over and over again you and I both know they won't be exactly the same.
---
I've been Pwned by KoaC. He really is the king of all cosmos.

You know that's not going to happen. With so many unknown factors and not every single person votes in the same poll over and over again you and I both know they won't be exactly the same.

Which is exactly my point. The stats are decent for picking matches where two characters are really far apart on them, but their lack of transitivity makes them incorrect in almost every case. Have the stats ever been dead on when picking the percentages?

~*ST*~
---
Winner of the Spring 2004 "Best. Game. Ever." Contest
"If my fate is to be destroyed... I must simply laugh!" -Magus

Which is a fair enough point, Leonhart, and Tidus certainly made it easy to believe he should have been stronger than that due to him and Shadow lining up almost perfectly. I also wanna say maybe Magus dropped with age, but Crono actually gained that year; while not impossible, you would think that Magus wouldn't drop as Crono gained. I'm gonna try to wait until we get some more matches before we delve much deeper, but for now there's not much of an answer.


Kind of off-topic, but I think it's weird how this looks to be Magus's worst year without a doubt DESPITE the following:

--Crono making the Final Four in 2k2 and BARELY losing to Mario in controversy.
--Magus beating Sam Fisher and Ganon before over-performing against Link in 2k3, wrecking quite afew brackets along the way.
--Crono destroying Tom Nook and Kefka before barely losing to Mario again in controversial fashion, this time in the Sweet Sixteen.
--Chrono Trigger being the runner-up in the Game Contest in 2k4.
--Magus destroying cult-favorite Luca Blight before "gettin' SFF'd" by Crono in 2k4.
--Crono beats down Conker and doubles Magus before BEATING Mario in the Sweet Sixteen in 2k4, then goes on to lose respectably to Link.
--Frog makes his magical run of barely defeating both Liquid Snake and Master Chief in 2k4 before putting up a hell of a fight against Solid Snake (which all signs lead on an over-performance, but nonetheless).
--Lavos puts up decent numbers against Liquid Snake in the Villain Contest after beating Mother Brain.

Look at that list; every CT-related character or the game itself has made the Sweet Sixteen before. Every CT-related character or the game itself has lost admirably before. Every CT-related character or the game itself has had at least a five-seed before. Crono has looked like he's gained strength every year thus far as well. CT is already quite the force, but it's amazing to me to think that, after all of this, literally anybody who keeps an eye out for these contests hasn't played Chrono Trigger; it has made a huge name for itself here. Apparently not though, apparently not.
---
Hillary Clinton: We need to treat violent video games the way we treat tobacco or alcohol.
Lewis Black: Namely as valued contributors to our re-election fund.

Have the stats ever been dead on when picking the percentages?

To my knowledge, no, and anytime they do if they ever do would most likely be pure coincidence instead of there being a perfect carry-over.
---
Hillary Clinton: We need to treat violent video games the way we treat tobacco or alcohol.
Lewis Black: Namely as valued contributors to our re-election fund.

People who are good with stats, what are the chances that we will see Albert Wesker>Leon in the X-stats?
---
When this happened, all of us got PWND by Smurf
http://www.freewebs.com/vgcg/Knuckles.jpg

From: UltimaterializerX | Posted: 8/17/2005 6:29:10 PM | Message Detail
Which is exactly my point. The stats are decent for picking matches where two characters are really far apart on them, but their lack of transitivity makes them incorrect in almost every case. Have the stats ever been dead on when picking the percentages?

They're not meant to be dead on. They're just an estimate. The stats, are just a nice reference and a way of ranking the characters each year. Of course, they're not that accurate , but, saying they're completely useless is just wrong, they have been off in some match, but they have provided a nice reference for others.

I'm saying the XStats are the only thing that should be used, of course they're not, you should think about it, but, you can't discredit them completely.
---
Ahem! There's LETTUCE on my boots!
http://www.geocities.com/lettucekefka/

I wanna argue one more thing before I start my real-life, heh.


Unless the Xsts predict a match down to a perfect percentage, they are not correct.

Let's assume the x-stats ARE meant to be able to perfectly predict matches down to the very percentage. Afterall, outside of direct matches, it's some of the best information we can use. If our x-stats can't predict matches perfectly, nothing can, right?

Let's compare this situation to basketball analysts.

Here, we have people being paid more than probably any five of us combined and put on television to give us their predictions on a regular basis. We have well over a hundred characters we examine as opposed to their, what, 32 teams I think? We have had contests since 2002...while I have no statistics, I would wager thata majority of the players in the NBA have been playing since 2002 or earlier. We see a total of sixty-three matches per contest, whereas each team plays 90-100 games...BEFORE THE TOURNAMENT. Then, once we're in the tournament, we see best-of-five/best-of-seven rules be used (depending on the round). By the time we get to the championship, there's still at least four games that has to be played, and it can go up to seven games...against the same opponents.

How many times does said basketball analysts get their predictions exactly right?

Despite given almost-infinitely more data and much more recent data than what we're ever given, despite it being much more of a year-around thing, despite its analysts being PAID to predict these games and being cast on television for the world to see...they can't do it either, and especially on a regular basis. If they can't do it on rematches where they last played one-two days ago, of course a simple math formula shouldn't be able to precisely drill match-after-match either a year later. It's completely and absurdly unreasonable to think they should, too.
---
Hillary Clinton: We need to treat violent video games the way we treat tobacco or alcohol.
Lewis Black: Namely as valued contributors to our re-election fund.

I've said it once and I'll say it again: the Stats are only there for reference, they are not hard fact as to the respective strengths of characters. The cannot take into account boosts or drops in the year between contests. Use them at your own risk.
---
Today's subliminal thought is: