That was always infeasible and I made the post in jest; it made Kefka > Kirby, and, right now, Diablo > Zelda (heh). I would like to point out that it has climbed to 52.92% victory for Diablo by now.

Don't disregard the possibility.

Heil Dweebenheimer!!! For I am MasterMage119!!!

Heh, in my bracket, I have CJ and Leon in the same four-pack as well. I'm thinking about doing a fanfic, since I have loads of free time.
---
Sephiroth is good, if he wasn't good, sephiroth wouldn't have 8.000 fans around the world. ~Siimor11
ZSB [aX]

http://www.gamefaqs.com/shared/spr05/b28.jpg

That is one badass picture, I must say.

---
“You see, at this point… I’m pretty much the Queen ***** of the Universe. And not all of your little soldiers or space ships will stand in my way again.”

As for the picture: Ganondorf looks awesome and Robotnik looks like he's trying to eat his own boob. I think a better picture could have been found.
---
Sephiroth is good, if he wasn't good, sephiroth wouldn't have 8.000 fans around the world. ~Siimor11
ZSB [aX]

MM's right; again, I think Ganon's got it and there's probably not gonna be much question to the match...but between the arguements Chichiri has made for Magus's four-pack being over-rated and the general suspicion of that four-pack being fishy, CATS being our most consistant character to date thus far, and Bison/Diablo drawing more votes than either of Ganon's matches...it's not out of this world. I think Wesker beat Luca noticably more than he would have been expected to if they were indirectly compared though, thus screwin' Kefka and Diablo from any gains like that, but if Ganon really is at 25.9% on Link, Diablo might be able to make it closer than we thought.
---
Serious Reminder: Vote for Dr. Robotnik
"What awaits each person in Heaven is eternal peace, divine rest, and $10,000." --Cartman

CATS predicts Ganon at 28.43% on 2k3 Link.
---
If you go to him, he will destroy you. And your last moments will be of shadow and pain.

I was using 2k4 Link...I get the feeling we're still doin' something else different though, since I'm gettin' 27.77% for Ganon on Link 2k3...lemme clarify what I'm doin' and you tell me if I'm doin' it right or not:

Assuming CATS 2k4 = CATS 2k5, he's at 12.17% on Link 2k4. Ansem should then be at 13.24% (12.17/45.96*50=Ansem). Therefore, Ganon should be at 25.9% (13.24/25.56*50=Ganon), and that's what they should get on Link 2k4.
---
Serious Reminder: Vote for Dr. Robotnik
"What awaits each person in Heaven is eternal peace, divine rest, and $10,000." --Cartman

CATS predicts Ganon at 28.43% on 2k3 Link.

And that could very well be possible. I'd rather make assumptions based on a very constant character than make them on a close four-pack when close matches would logically cause a few more anamolies.

Heil Dweebenheimer!!! For I am MasterMage119!!!

I don't like basing Ganon off of 2k3's four-pack OR fodder like CATS, no matter how consistant he is; I just think we should look more at Ganon through CATS since at least he has a rep of being consistant and we're able to argue more and more against Magus's strength.
---
Serious Reminder: Vote for Dr. Robotnik
"What awaits each person in Heaven is eternal peace, divine rest, and $10,000." --Cartman

Doing anything assuming a constant CATS is just rediculous. CATS quite obviously overperformed in his match against Ansem due to bracket support and having his most recognizable picture, not to mention he was up against freakin' ANSEM, not a powerhouse like Link or Cloud.
---
For the sake of simplicity, I'm going to refer to Gordon Freeman versus Tanner as Schrodinger's Match. -Phoenix Flattener

Bah, I hate using 2k4 Link as a base. Anyway, to get that 2k4 Link to 2k3 Link, take 50 * 25.9 / (100 - 54.46) = 28.43 (my way), or 50 * 25.9 / (100 - 53.62) = 27.92 (cn's way). I think it's easier to just go with 2k3 Link.

And I've always wondered about this; what makes close matches more susceptible to anomalies?
---
If you go to him, he will destroy you. And your last moments will be of shadow and pain.

I'm not saying basing Ganon off of a steady CATS is right; I'm just saying that it may be closer to being right than basing Ganon off of Magus.
---
Serious Reminder: Vote for Dr. Robotnik
"What awaits each person in Heaven is eternal peace, divine rest, and $10,000." --Cartman

I'd think CATS would get some votes for his picture, as neither him nor Ansem are THAT well known or cared about by most.
---
SpC2k5: 32/36 Final four: Bowser, SEPHIROTH, Kefka, GANONDORF
*is Janus5000*

And I've always wondered about this; what makes close matches more susceptible to anomalies?

Not really sure, it just seems like it would make sense.

Heil Dweebenheimer!!! For I am MasterMage119!!!

Not really sure, it just seems like it would make sense.

Why does it make sense? Which side would it favor? Just like "Nintendo vs Square," this seems to pop up every now and again with no justification or even explanation of who it benefited.
---
If you go to him, he will destroy you. And your last moments will be of shadow and pain.

Why does it make sense? Which side would it favor? Just like "Nintendo vs Square," this seems to pop up every now and again with no justification or even explanation of who it benefited.

I'll try to put it into words. Just, not this second.

Heil Dweebenheimer!!! For I am MasterMage119!!!

And I've always wondered about this; what makes close matches more susceptible to anomalies?

Close matches should definitely be less susceptible to anomalies.

There can be more volatility at more extreme levels of performance; in close matches, there will be much less. e.g., Mario vs. Crono series.
---
www.SC2K5.com

Not really sure, it just seems like it would make sense.

It would defy sense.
---
www.SC2K5.com

I think close matches CAN cause anomolies if the one that's supposed to win doesn't, but since they're supposedly so close anyways, I think that only matters a little. I think what he's referring to is how people will rally for outside voters and/or cheat in closer matches, which it may be possible if the rallying/cheating isn't in the right proportion.

I'm not sure where you get the 54.46 or 53.62 values, though. I kinda took a shortcut when it came to converting from Link 2k4 to Link 2k3; I took a character with a nearby value in the 2k4 stats (Dante in this case) and divided what it showed him to get against Link 2k3 by what he was supposed to get against Link 2k4 (27.33/25.49=1.0721851...). I round it off afew digits (I used 1.0721) and multiplied it by what Ganon was expected to get against Link 2k4 (25.9 in this case), and I got approximately 27.77% for Ganon on Link 2k3. I'm not surprised this's is considerably closer to cn's way though, since I'm using his conversions from sc2k4.com right now, heh.
---
Serious Reminder: Vote for Dr. Robotnik
"What awaits each person in Heaven is eternal peace, divine rest, and $10,000." --Cartman

I think what he's referring to is how people will rally for outside voters and/or cheat in closer matches

We've never seen a case where this would alter the % much. The only possibility seems like Kuja vs. Master Hand—but it is clear that this was because of the day vote, not cheater deciding to rally behind Master Hand because of a close match.

Also, cheaters would give up once a good sized lead was established anyway. Not to mention that a good sized lead usually doesn't amount to much difference in percentage performance.

If close matches do cause cheating, and it is on both sites, the %es won't be affected much. Even if it is one-sided it shouldn't affect percentage performance much, unless you're talking about many thousands of one-way cheating votes.
---
www.SC2K5.com

Maybe because close matches are highly dependent on turnout? It's similar to an explanation I heard about politics. For example, America has a roughly even split between Democrats and Republicans. A presidential candidate who can mobilize his base more during election will win (cough Diablo cough), even though his opponent has a similar number of supporters.
---
A soda a day invites tooth decay.

It's not that there's a reason, it's just that it seems to be the case. Look at Liquid, dropped... like... rock. Look at Tidus, dropped... like... rock. Not to mention all of those close chains seem to be exceptionally high in the standings.

I'm calling it... Ganondorf, dropped... like... rock.
---
He needs nominations, 61! Nominate Raiden (Metal Gear Solid 2) for SC2K5!
http://boards.gamefaqs.com/gfaqs/genmessage.php?board=8&topic=21155016

Aw, looks like Diablo should be ending around that 54% area. Very, very nice for him to perform this well though.

---
“You see, at this point… I’m pretty much the Queen ***** of the Universe. And not all of your little soldiers or space ships will stand in my way again.”

Oh, and the whole PGC thing with Cloud/Mario. But even that would be only a difference of 3 points; the volatility of blowouts seems to be much higher than that, in terms of impact.

Rmember that a 3 point swing in a 50/50 match isn't as big a deal as a 1 point swing in an 85/15 match.
---
www.SC2K5.com

It's not that there's a reason, it's just that it seems to be the case. Look at Liquid, dropped... like... rock. Look at Tidus, dropped... like... rock. Not to mention all of those close chains seem to be exceptionally high in the standings.

Where do you get this from? That doesn't seem like the explanation one should find there.

A series of close matches will result in more trustworthy extrapolated results than a series of blowouts.
---
www.SC2K5.com

What other series of close matches are there? Are there any? I want to know to see if they dropped... like... rock. (okay I'll stop that now)
---
He needs nominations, 61! Nominate Raiden (Metal Gear Solid 2) for SC2K5!
http://boards.gamefaqs.com/gfaqs/genmessage.php?board=8&topic=21155016

Christ... Ganondorf looks like a freakin' gargoyle...

---
RIP: Tnote827 - 4/4/05
Team Dragostea Din Tei for Trivia XV

What I'm saying is, for as far as we know, how do we know matches like FFT/MGS or Liquid/Frog had the right winner? The cheating and rallying may have been pretty close to equal, but both of those matches were decided by 38 and 93 votes (respectively). Now, their vote totals weren't really all that different from other matches there, though they were in the upper end of vote totals as far as other matches in the round said match was in...but 38-93 votes is hardly anything; our stats don't account for wrongful winners (except for Crono > Magus...that's my weddin' gift to ya Ulti, enjoy).

Again, I think if they were that close to begin with, then it won't make for much of a difference though there will still be some irregularities (such is the way of the stats). I'm just saying that, in all likelyhood, there IS that difference somewhere in probably a good number of our matches that are within, say, 52.5-47.5.
---
Serious Reminder: Vote for Dr. Robotnik
"What awaits each person in Heaven is eternal peace, divine rest, and $10,000." --Cartman

I'm calling it... Ganondorf, dropped... like... rock.

Ganon's value is 2 years old. He had a pig picture against Ansem. CATS had his best pic ever. All these things were mentioned as reasons for Ganon to underperform the straight-up expectation before the Ganon vs. Ansem match.

Ganon will most likely underperform relative to the straight-up expectation vs. Sephiroth as well, as he doesn't seem to be at 34.72% against 2K3 Link—maybe 30%. And Sephiroth seems to be stronger than his value in 2K4; probably closer to his 2K3 value. Take Ganon at 30% and Sephiroth at close to 50%, and you're approaching 70% for Seph. The straight-up SFF-adjusted 2K4 values put Sephiroth at only 61.8%. I would be quite surprised if Ganon held Sephiroth so low.
---
www.SC2K5.com

http://www.gamefaqs.com/shared/spr05/b28.jpg

Oh my....I <3 Ganon. He looks AWESOME in that picture! Robotnik isn't have bad either.
---
Out of space....
It's MOOOOOOLLLLLTTTAAAARRRR!!!!

how do we know matches like FFT/MGS or Liquid/Frog had the right winner?

The winner is irrelevant; this doesn't matter at all. Super-close matches are flukes in terms of who wins, and it doesn't matter for the extrapolated method. Switch the winners' and losers' percentages for all these matches and any effects would be trivial.
---
www.SC2K5.com

Robotnik's pic last round was better. This one is sort of weird.

Ganon's is decent. Though he looks Scottish.
---
www.SC2K5.com

Robotnik's picture was better last round, but his picture this round does look good.

---
“You see, at this point… I’m pretty much the Queen ***** of the Universe. And not all of your little soldiers or space ships will stand in my way again.”

OMG!!! SC Wars Episode V: Attack of the Stats!!! Witness the transformation of Sephiroth into Dark Lord of the Stats!

I'm actually hoping that Bowser and Ganondorf screw up the stats even more against Sephiroth.
---
Graduation: April 30, 2005 "There are only 10 kinds of people in this world, those that read binary and those that don't."

The winner is irrelevant; this doesn't matter at all. Super-close matches are flukes in terms of who wins, and it doesn't matter for the extrapolated method. Switch the winners' and losers' percentages for all these matches and any effects would be trivial.

Not really true. If the opponent in the next round (e.g. Crono) shares a significant fanbase with one of the contenders (e.g. Magus) but not the other (e.g. Ganondorf), the result of the next match can be vastly different, depending on the winner.
---
For the sake of simplicity, I'm going to refer to Gordon Freeman versus Tanner as Schrodinger's Match. -Phoenix Flattener

Not really true. If the opponent in the next round (e.g. Crono) shares a significant fanbase with one of the contenders (e.g. Magus) but not the other (e.g. Ganondorf), the result of the next match can be vastly different, depending on the winner.

We can switch the percentages around arbitrarily. It would just be question of altering strength numbers. In this case, simply flipping around the values for the winner/loser. If this were done for close matches, there would be virtually no difference in the standings' values.
---
www.SC2K5.com

hooray for Ulti!! *cries for Kefka*


No, it was just that we had no proof that Link overperformed on Mega Man, and the Blue Bomber should have NEVER been automatically given his 2003 number, especially since that was the only year we had anything absolute on him. For him to receive that number again with no debate whatsoever is ridiculous.

i remember saying something like this about the 20xx Factor some time ago. i still agree that it was unusual to count that as an undisputed adjustment. it does seem odd to me that Snake could drop so much, though. but still.....


Who knows? Perhaps the angel picture did make a difference against Wesker. Still, I wouldn't have thought it would make him look like Kirby (not literally, of course).

i continue to stand by this. i believe that Kefka would be much closer to Diablo if he got that pic instead of this ugly clown pic. i'm not sure if it would have been enough to win, but it would have been very close.


I'm not sold on Seymour being stronger than Sin for some reason. He just comes off as being the FFX version of Kefka.

if you ask me, he comes off as being the FFX version of Kuja.

I hope we get nominations for Summer Contest the day after the Spring Contest ends so we dont try to hard to fit the stats from the Villians Contest in.

i hear that, man. i could do without a month of nothing between contests.

Heh. Ganondorf might have a crappy performance tommorow if there's any merit to Wind Waker Link being weak.

cel-shaded Ganon still looks cool and has casual appeal, unlike, the big-eyed kid version of Link. heck, i wouldn't be surprised in the least if a Young Link pic hurt Link's performance a little bit.

---
Official Co-Host for PUB TRIVIA XV
http://boards.gamefaqs.com/gfaqs/gentopic.php?board=579093

>_>

Mario got mass support to beat Cloud in 2002, because PGCN linked there, right?

Well...

Couldn't that also be why Link was so strong in 2002? I mean, he's already stronger than anyone on this site, but with that added support, it may have been why he did so well against everyone.

Or did they only link during Mario matches?
---
"The PS3 isn't designed to lean towards games. It's not a computer for children."- Ken Kutaragi

We can switch the percentages around arbitrarily. It would just be question of altering strength numbers. In this case, simply flipping around the values for the winner/loser. If this were done for close matches, there would be virtually no difference in the standings' values.

I'm not sure you understand my point. Let's say we have Ganondorf vs. Magus with the winner going up against Crono.

Case 1: Magus beats Ganondorf 50.30% / 49.70%, and then Crono beats Magus 69.78% / 30.22%. Extrapolated results against Crono are then
Crono: 50.00%
Magus: 30.22%
Ganondorf: 30.04%

Case 2: Ganondorf beats Magus 50.30% / 49.70%, and then Crono beats Ganondorf 56.46% / 43.54% (the results I get using the SFF-adjusted standings, but this number is arbitrary - all that matters is that it is significantly different from 69.78%). The extrapolated results against Crono are then
Crono: 50.00%
Ganondorf: 43.54%
Magus: 43.28%

As you can see, a tiny swing in one match that alters the winner can have a huge effect on the results of the next match, and thus on the extrapolated standings, due to one characer having SFF but the other not.
---
For the sake of simplicity, I'm going to refer to Gordon Freeman versus Tanner as Schrodinger's Match. -Phoenix Flattener

if you ask me, he comes off as being the FFX version of Kuja.

Good comparison. I've never believed the Seymour>Sin stuff myself. I don't think Seymour would be as strong as Sin.
---
www.SC2K5.com

As you can see, a tiny swing in one match that alters the winner can have a huge effect on the results of the next match, and thus on the extrapolated standings, due to one characer having SFF but the other not.

To be fair, we always adjust for those situations.

Heil Dweebenheimer!!! For I am MasterMage119!!!

Aprosent, that doesn't affect what we're talking about. We already make (controversial) SFF adjustments to the best of our abilities. "What if" scenarios don't enter the picture anywhere.

Though if you wanted, you could certainly make a make-believe SFF adjustment in some cases—e.g., if Mario did face Wario. That's no problem (as long as you don't expect to be accurate in terms of what would've actually happened...). That's not related to our discussion though.
---
www.SC2K5.com

Ok, I guess we're just arguing about different things then. What we can agree on is that the extrapolated standings are only accurate and useful when SFF does not exist to a significant extent. Given the results of one match not involving SFF, we can only make an educated guess at the extrapolated results of a match involving SFF or vice-versa.
---
For the sake of simplicity, I'm going to refer to Gordon Freeman versus Tanner as Schrodinger's Match. -Phoenix Flattener

The reason I think that close matches are less likely to be very inaccurate is more or less just... a feeling... I mean, I think that two characters permorming equally against each other would not perform equally against any other common opponent. I feel that they happen to have a unique situation when they will perform almost equally against each other. But I could be completely wrong, and probably am.

Heil Dweebenheimer!!! For I am MasterMage119!!!

My thoughts exactly. I was just writing a post explaining my thoughts on that. I think Master Chief/Frog would be a good example of that....
---
Graduation: April 30, 2005 "There are only 10 kinds of people in this world, those that read binary and those that don't."

I wanted to say Frog/MC, but I didn't due to there being some linkage. It's one thing to rally or cheat, but linkin' (while similar) is something above those two I think, so I used Liquid/Frog instead.
---
Serious Reminder: Vote for Dr. Robotnik
"What awaits each person in Heaven is eternal peace, divine rest, and $10,000." --Cartman

We need Donkey Kong/Tails in the first round this summer to test MasterMage's theory.
---
O, flesh, flesh, how art thou fishified! ~Mercutio

How would that match prove his theory? You would need two people who HAVE faced each other in a close match go against a common opponent....
---
Graduation: April 30, 2005 "There are only 10 kinds of people in this world, those that read binary and those that don't."

...I misread his theory.

But we still must have that match.
---
O, flesh, flesh, how art thou fishified! ~Mercutio

How would that match prove his theory? You would need two people who HAVE faced each other in a close match go against a common opponent....

Exactly. And we'd need two who have probably not changed over the course of a contest, so Master Chief is out. Basically, we need a situation where Frog and Liquid Snake both end up facing Link in the same contest.

Heil Dweebenheimer!!! For I am MasterMage119!!!