Pre-Season Spring 2004 Contest Discussion - Part 3 |
: | | | | |
Page 1 of 10 | | |
From: solarshadow | Posted: 3/21/2004 9:56:58 AM | Message Detail |
If the discussion topics are this busy during the pre-season, they're going to explode once the contest starts. Only twelve more days...
The links:
Summer 2002 Contest Solarshadow's stats website: http://solarshadow2002.tripod.com The stats topic: http://solarshadow2002.tripod.com/page1.html
Summer 2003 Contest Solarshadow's stats website: http://solarshadow-stats.tripod.com First stats topic: http://solarshadow-stats.tripod.com/archive/topic1p1.html Second stats topic: MISSING Third stats topic: http://www.angelfire.com/games5/sum_contest/Page1.htm (See note)
Spring 2004 Contest Solarshadow's stats website: http://solarshadow-stats.tripod.com/sp2004/index.html
2003-2004 Off-Season First topic: http://membres.lycos.fr/shindohikaru/stats1.htm (See note) Second topic: MISSING Third topic: ? Fourth topic: ?
Spring 2004 Pre-Season First topic: http://geocities.com/cyber1166/gamefaqs/04SpringPreSeason1.html Second topic: http://cgi.gamefaqs.com/boards/genmessage.asp?board=8&topic=12997958 (Until purge)
Note: Don't use the links in the topic to browse through the pages, change the page number in the URL. --- Contest Stats: http://solarshadow-stats.tripod.com |
From: charmander6000 | Posted: 3/21/2004 10:08:53 AM | Message Detail |
The countdown continues --- LUE is a cancer, and if it can't be destroyed, it must be contained. - CjayC My site for the games contest http://www.freewebs.com/spring_game_contest/ |
From: bjgood0 | Posted: 3/21/2004 10:09:12 AM | Message Detail |
Wow, that was fast. |
From: bj good | Posted: 3/21/2004 10:10:26 AM | Message Detail |
Oh...heh...looks like it's been up a while. --- By the way, bj good, your posts speak of wisdom as well as your name. - sakbato88 |
From: ChichiriMuyo | Posted: 3/21/2004 10:19:14 AM | Message Detail |
2003-2004 Off-Season First topic: http://membres.lycos.fr/shindohikaru/stats1.htm (See note) Second topic: MISSING Third topic: ? Fourth topic: ?
The
second isn't missing, I have it on my pc. If anyone wants to upload it
anywhere just ask me and I'll email it to you or aim it to you. And I
know someone has the third one... just can't remember who. --- I`ve become sort of a Chichiri fan since then. - UltimaterializerX |
From: Heroic Tails | Posted: 3/21/2004 10:24:38 AM | Message Detail |
I
have Part 2, 3 and 4 of the off-season discussions, but I'm just too
lazy to edit (getting rig of ad space on the left mainly) and upload
them... I'll try to do it before my HD crashes. --- I'd rather be damned for my own beliefs, than be damned for someone else's. - Shake [This sig has been approved by UltimaterializerX] |
From: creativename | Posted: 3/21/2004 10:27:08 AM | Message Detail |
I'll take that personally. ;) I remember how silly I felt after that match.
Heh
:) Crono was obviously far stronger than just about everyone expected,
though it was clear from his high odds that he was underseeded. I
suppose many might've written it off as a case of him being overrated
by rabid Chrono Trigger fanboys though (ala Master Chief and Vercetti
in 2K3). And Dante had excellent odds as well. So there was no
systematic way to predict the popularity levels of these two characters
beforehand.
Yeah. It's going to take me a week to archive everything. ;)
How so? Just up your posts per page to 50, and you only have to archive 10 pages. --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: Ngamer64 | Posted: 3/21/2004 10:35:49 AM | Message Detail |
And don't let the purge get you down, I've got Pre-Season 2 saved away as well.
http://geocities.com/cyber1166/gamefaqs/04PreSeason2.html
Let's try to make this one last the full 12 days, shall we? As in, no wasting posts on arguing with a troll in here, etc.
--- the-elite.net Ngamer's Contest Archives: http://geocities.com/cyber1166/gamefaqs |
From: charmander6000 | Posted: 3/21/2004 12:16:16 PM | Message Detail |
I knew I should've posted this in the offseason, but I forgot so I'll just post this now.
Summer 2003 Contest Review in 1560 words
We
start off with Link last years champ and favor to win this year totally
destroying AiAi getting the biggest blowout award for this year. Fox
beats Pikachu by much more then we all expected his award to get beaten
by Link next round. Our first true nail-bitter match Ganondorf barley
beats Tidus, but this match got the most votes in this round. Magus the
character that says that will beat Link comfortably beats Sam Fisher in
his match. Mario’s brother Luigi now in this contest blows Ratchet out
of the water and people now say he may give Samus some trouble. Squall
pulls an upset beating Jill a.k.a. Kirby Killer making many of the
“experts” lost there perfects today. KOS-MOS unlike Ulala easily beats
Crash with a 20% vote gap. Samus Sonic’s rival easily beats Isaac
making people believe that Samus may also be a threat to Link. At the
beginning of Cloud vs. CATS, CATS was actually beating Cloud, which
only lasted for a few minutes, but that brought shivers down Final
Fantasy Fans back, but the match ended in a huge blowout in Cloud’s
favor. Tails loses again in the first round again to Auron making many
Sega fans say give Tails some easier opponents not these sweet 16
material ones. With many Earthbound fans saying that Bowser will lose
people actually believed that this match would be close, but it ended
up Bowser tripling Ness, which made people say Yoshi better put up the
same numbers if he wants to get past round 2. Yoshi does worse against
a worse character that Bowser had making people believe that it may be
all over for Yoshi next round. Aeris who got an 11 seed again easily
doubled Kingdom Hearts star Sora maybe CJayC shouldn’t put 2 characters
of the same game against each other in the first round. Master Chief
got anti-voted so much that Felix almost beated him, well that’s Xbox
for you. Zero defeats last years biggest upset Scorpion making people
believe that he only got to the elite 8 from a very easy road. This
also lessens Pac-Man’s chances of winning also. Sonic beats Ken easily,
but since Samus went up against Ken last year this match showed that
Sonic has gain in popularity. Mario runner-up last year delivers a
blowout which surprise us all since he barley gets 25% against the
weakest character. Shadow made a huge upset and has beaten Wario making
the whole board stunned. Kefka who’s suppose to totally beat Pac-Man
and give Crono a hard time in round 2 failed in his match with Pac-Man.
His picture looked more like a lettuce, but he did ended up beating
Pac-Man in a very close match. Crono who lost to Mario last year and he
may get his revenge in round 3 beats Tom Nook not as much as Mario beat
Captain Olimar, but that’s because Tom is a harder opponent. Alucard
easily beats Bomberman by more then Jill did last year making people
believe that it’s over for Kirby in round 2. Kirby gets a comfortable
win over Ramza who many people thought this would have been a close
match not Pac-Man vs. Kefka. Max beats Gordon in the match of the
weaklings this match also had the least number of votes in the whole
tournament. Semi-Finalist Sephiroth blows Raziel out of the water, but
it wasn’t as big as Mario’s and Crono’s win has Sephiroth been losing
steam? |
From: charmander6000 | Posted: 3/21/2004 12:19:05 PM | Message Detail |
(Part 2)
Snake
beats one of his own character in a big blowout. Knuckles beats Yuna
easily when all the odds looked like it could go either way. Ryu takes
out Duke showing that he can easily take out Dante next round. Dante
blows Ryo right out of the water making people believe that Ryu vs.
Dante may be closer then we all think. Donkey Kong returns with a lower
seed who easily takes out Vyse. Tommy Vercetti who most people thought
that he could take out Mega Man failed to impress anyone with a bad
performance against Kite making Donkey Kong vs. Vercetti a little more
exciting. Zelda takes out former one seed Lara Croft (so much for 1
seed) without ant trouble. Mega Man beats Mr. Resetti getting almost
90% of the votes ending the first round with a blowout. Since we
ended round 1 in a blowout why don’t we start round 2 with one too as
Link slaughters Fox it looks like Link isn’t losing any steam. Magus
who was supposed to easily defeat Ganondorf became another close match,
but this time Ganondorf lost so much for Magus as the “Link Killer”.
Squall pulled off another upset this time beating Luigi as much as he
did to Jill. Squall is now named the biggest upset and people now fear
that he may defeat Samus as well. Samus easily defeated KOS-MOS showing
that Crash is extremely weak and he shouldn’t be in this contest
anymore. Cloud easily takes out the last of the Final Fantasy X in a
blowout due to Same Fanbase Factor. People are calling this the match
of the round and Bowser takes out Yoshi in a somewhat close match
people are now saying that Bowser may give Cloud some hard times in
round 3. Aeris defeated Master Chief due to the hundreds of anti-votes
from the Xbox haters club, but that defeat wasn’t a big blowout as
people thought it would be. Sonic almost losing to some side character
made more people believed that Sonic has lost popularity then had gain
some. Mario also almost losing to Shadow making people hopes that Crono
will have a hard time with Kefka. Crono blows Kefka out of the water
making people believe that Mario is doomed in round 3. Alucard barley
beats Kirby in a very close match making this the second time Alucard
barley made it to the sweet 16. Sephiroth totally destroyed Max who
wasn’t even suppose to make it to the second round and only did based
on luck. Snake beat Knuckles for the second time in the second round,
but this time Snake did better then last year. Ryu vs. Dante was a very
good match, but as we all expected Ryu made it on top, but the match
was close. Vercetti who was suppose to easily beat Donkey Kong, but
turned out to be one of the closest match there were a few lead
changes, but Vercetti came through and barley beat Donkey Kong. Mega
Man only got 60% against Zelda which surprised many people and made
people think that Mega Man was losing steam. In round 3 Magus only
manages to get 35% against Link which wasn’t what we all expected, but
if you look at it this way Mario only got 37% so Magus did do really
good. Samus defeats the biggest upset Squall without too much trouble.
Cloud totally destroys Bowser making people start to look at his
strength. Sonic again almost loses to a side character if he can barley
defeat Aeris then it’s all over for him in round 4. --- LUE is a cancer, and if it can't be destroyed, it must be contained. - CjayC My site for the games contest http://www.freewebs.com/spring_game_contest/ |
From: charmander6000 | Posted: 3/21/2004 12:20:07 PM | Message Detail |
(Part 3)
Mario
vs. Crono the rematch was as great as the first match Crono had the
lead for most of the day, while Mario had the morning hours, but with
one hour left Crono is still winning by more then 200 votes will Crono
have his revenge, but wait Mario is making a comeback and he takes the
lead and defeats Crono with almost no time left on the clock this is
the second time Mario has beaten Crono with almost no time left people
were happy and people were mad how could’ve Mario won they all said.
Also that match broke the record of 129 000 votes they got 133 000
votes. With the Mario vs. Crono match excitement still on this match
wasn’t really notice, but Sephiroth easily takes out Alucard in a match
that was all in his favor. Snake took out Ryu in a slightly close
match, but that was no where near Mario vs. Crono close. The fear of
Mega Man losing steam has finally have been put to the test and Mega
Man still has all his steam easily beating Tommy Vercetti. As of now
there are only 1 and 2 seeds left. Round 4 has started and Samus the
last chance for defeating Link didn’t happen and this made people think
that Link will be the champion. Cloud almost doubles Sonic making
people start to think that Cloud or Sephiroth may beat Link. If you
still didn’t believe that Cloud or Sephiroth could beat Link well
Sephiroth just got almost 62% against the Nintendo legend Mario. Mega
Man beats Snake in a somewhat close match. Round 5 is here and the
impossible just happened Link has been defeated by Cloud with over 147
000 votes beating Mario vs. Crono vote total and over 17 000 brackets
are ruin with only 13.37% of the brackets predicting that Cloud will
win. The Sephiroth vs. Mega Man rematch wasn’t as exciting as last
years Sephiroth also beats Mega Man near the same percent as he did to
Mario. The finals are here and this is Nintendo’s worst nightmare an
all square final Cloud vs. Sephiroth. Cloud ended up winning due to he
was the main character and Sephiroth was the villain in Final Fantasy
VII.
--- LUE is a cancer, and if it can't be destroyed, it must be contained. - CjayC My site for the games contest http://www.freewebs.com/spring_game_contest/ |
From: UltimaterializerX | Posted: 3/21/2004 1:29:44 PM | Message Detail |
I`ve become sort of a Chichiri fan since then. - UltimaterializerX
I
remember making that statement, but wasn`t that a part of a bigger
paragraph? Something about how I used to not get along with you, and
how you kick more ass with each post you make? Something along those
lines... ~_^ --- MIASU!! My SC2K4 Petition: http://cgi.gamefaqs.com/boards/genmessage.asp?board=7&topic=12558738 |
From: cyko | Posted: 3/21/2004 3:33:18 PM | Message Detail |
lol, nice review, Charmander!
--- Please go nominate Secret of Mana for SNES. Right Now. A Cheese Legend of Trivia 12 |
From: swirldude | Posted: 3/21/2004 3:42:57 PM | Message Detail |
When do you all have SSBM losing? I have it losing to FFX in Round 3. --- The
bomb detonated, and they got to see a very annoyed CATS running down
the hill, with his hair on fire. "AUUGH! SOMEONE SET UP US THE BOMB!"
~BigCow |
From: Slowflake | Posted: 3/21/2004 3:54:21 PM | Message Detail |
According to the 58 brackets entered in the BOP:
4 have it losing in round 1 11 have it losing in round 2 22 have it losing in round 3 9 have it losing in the division finals 11 have it losing in the semifinals No one has it losing in the finals 1 has it as champion.
And
the timing is impeccable to ask Creativename a question. Using the same
kind of tables you did to calculate the odds, would it be possible to
create a table saying how many users have a given game losing in a
given round, sort of like above? --- SpC2K4 Status --- Winner: FF7 --- Finalist: CT --- Semifinalists: SMB3, LoZ:WW |
From: smitelf | Posted: 3/21/2004 3:57:27 PM | Message Detail |
When do you all have SSBM losing? I have it losing to FFX in Round 3.
SSBM
is one of the games that is going to make or break people's brackets.
Some have it going very far, to the final four; others have it losing
to Vice City in the second round. No pick can be discarded as absurd
but I personally have it losing to Vice City for reasons I've mentioned
several times in earlier discussion topics. --- "Your fate is sealed, and none but yours." -- Auron Proud Supporter of Starcraft in the Spring 2004 Contest |
From: nh82 | Posted: 3/21/2004 4:07:43 PM | Message Detail |
The
trouble with SSBM is that it's a total wildcard - it fits none of the
classic Gamefaqs popularity conditions, yet is still widely played and
discussed by users here.
Personally, I have had plenty of
trouble trying to work out if that popularity extends to the casual
voting user, or if it's just down to a large vocal support on the site.
I've
gone for pretty much the middle-of-the-road option with it losing to
FFX in round 3 as well. Feels justifiable, but it's more of an
intuitive pick than anything else.
Tbh, you could discuss the
128 division for months on end and never reach any firm conclusions -
you'd be just as well flipping a coin for all the matches from round 2
onwards (bar Halo/KH and FFTA/FFX - still, that leaves 5 matches well
up in the air...)
As is being said, without any experience of a
game popularity contest, we could be wildly wrong in our estimations of
who the giants are. Anyone got predictions on which match is the next
"Crono/Dante"? |
From: Slowflake | Posted: 3/21/2004 4:11:11 PM | Message Detail |
The next Crono/Dante? I was spot on last year with Zero/Scorpion, but this time around...
...I'll say LttP/FF6. The hype for this match is ridiculous, yet I think it'll be over with the first 50 votes. --- SpC2K4 Status --- Winner: FF7 --- Finalist: CT --- Semifinalists: SMB3, LoZ:WW |
From: creativename | Posted: 3/21/2004 5:16:39 PM | Message Detail |
Using
the same kind of tables you did to calculate the odds, would it be
possible to create a table saying how many users have a given game
losing in a given round, sort of like above?
Sure. Below is
how I would do it. There may be a simpler way, but though the
explanation is kinda long, it really doesn't take long to execute at
all:
*Make a column of the Round Number. For instance, just copy
a paste a "1" all the way down next to the round 1 data. Make sure you
name the column (e.g., Round). *Make a Pivot Table. --In the
Layout, put "Game" in the Row space. Make sure you click on "Show items
with no data", otherwise it won't list games that, you know, don't have
data for that round. You want it to list all games to make comparing
different rounds easier. --And put "Choices" (Sum of) in the Data space. --And put "Round" in the Page space. *Now
you'll have a PivotTable. The default Page will be "All". Thus, you
have a list of the total number of times a game was picked. Change the
Page to "1", and you'll have a list of the number of picks for each
game was picked in round 1. Change the Page to "2", and you have each
game's picks for Round 2. *To find the number of times a game was
picked to lose in round 2, just subtract the Page "2" data from the
Page "1" data. (if there are any negative numbers, there must be a data
entry error)
I suggest getting familiar with PivotTables. They
are very useful, and once you get the hang of them very powerful: you
can do time consuming tasks in very short periods of time. --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: creativename | Posted: 3/21/2004 5:41:19 PM | Message Detail |
I created a Word document (zipped) with screen captures that goes through the steps above.
http://home.nyc.rr.com/saqib/GameFAQs/losses.zip
It's 530K though. Let me know if it's any help.
Remember
that to create a PivotTable, go to Data-->PivotTable. Just hit next
for the "Step 1 of 3" window, highlight the data in "Step 2", and
choose the Layout by clicking "Layout" in "Step 3". --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: creativename | Posted: 3/21/2004 5:51:27 PM | Message Detail |
put "Game" in the Row space. Make sure you click on "Show items with no data",
Oops...forgot to mention that you have to double-click on "Game" after you drag it to the Row space.
Likewise, you have to double-click on "Choices" after you drag it to the Data space and then choose "Sum of". --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: creativename | Posted: 3/21/2004 6:07:40 PM | Message Detail |
...I'll say LttP/FF6. The hype for this match is ridiculous, yet I think it'll be over with the first 50 votes.
50 votes??? It'd have to be 40-10 for that to be true. Perhaps 500, but no way 50.
I
just don't see either game in this match getting much more than 60%.
It's possible LttP can get up to like 63% or something, but I doubt it.
The winner should get around 55%, with a good chance of it being much
closer. These are both very beloved and respected games.
I'm
giving the edge to LttP, and am somewhat confident in that because of
your bracket work where LttP is the somewhat-clear favorite. But I
don't think this is going to be some surprise blowout like Crono vs.
Dante was. Of course, that's why these things are surprises...but
still. It's Final Fantasy VI. And I can't see LttP getting blown out
either. I actually think both these games would beat anything in the
128 division, with the possible exception of FFX. --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: StopPokingMe | Posted: 3/21/2004 6:19:10 PM | Message Detail |
Anyone got predictions on which match is the next "Crono/Dante"?
I'd
say Wind Waker-Metroid Prime, or maybe Phantasy Star-Contra. A while
ago I would have said Halo-Starcraft, but hype for that seems to have
died down. In fact, I still have Starcraft in my bracket, but that's a
point I'm willing to give away--Starcraft seems to have good "pot odds."
When do you all have SSBM losing?
Against Vice City, but similarly, that's 2 points I'm willing to give away on a gamble. --- The GameFAQs Summer Contest 2003 Fanfiction Project: http://crolapras.tripod.com/ffproj.html |
From: Qwaar | Posted: 3/21/2004 8:23:24 PM | Message Detail |
I'm willing to gamble.I have SSBM losing to Final Fantasy 7 in the Semi-Final. --- Proud Supporter of OoT and Starcraft In SC2K4! |
From: creativename | Posted: 3/21/2004 10:27:52 PM | Message Detail |
I
found out that GunMage is running a "spreads" based gambling topic.
This inspired me to work with CJayC's 2K2 odds to find out if they were
useful for predicting %es. I only worked with 1st round data, but the
results were IMO very interesting.
What I did was try to
predict vote %es based on "favorite ratio". I'll use Mario vs. Servbot
as an example. Mario had odds of 6:1, Servbot of 750:1. Thus, Mario's
"favorite ratio" was 750/6=125, and Servbot's "favorite ratio" was the
reciprocal of that--.008.
I then ran a few regressions in Excel
based on a column of "favorite ratios" compared to %es each character
got in the 1st round. Graphing the relationship showed a curved trend,
so I tried squaring the "favorite ratios"; but the best results were
actually obtained with a logarithmic transformation.
The Regression results were as follows: Intercept: ___Coefficient: 0.5 ___t-stat: 35.84 ___P-value: 3.76e-43 LN(Favorite Ratio): ___Coefficient: .0799 ___t-stat: 13.29 ___P-value: 8.25e-20
As
you can see, the t-stats are high and the corresponding P-values are
extremely low. (that's a good thing) I can put the Excel sheet up later
if anyone's interested.
For those not familiar with regression,
here's a very loose description: this is an equation generated by
inputing one or more sets of X-values (the data--in this case, the
natural log of the "favorite ratio" was the only input I found
necessary) to predict a set of Y-values (in this case, the % of votes
the character received in the first round).
In this case, the equation was: 0.5 + .0799*ln(favorite ratio)
Strange
that the intercept ended up exactly .5. Anyway, I used the "favorite
ratio" and other transformations of the favorite ratio as well in other
regressions, but with the LN of the favorite ratio present the P-values
for the other inputs weren't that high.
So, in this case, the equation would've predicted a Mario vote % of: .5 + .0799*ln(125) = 88.6%
Due
to the properties of the natural log function, this means Mario's
opponent Servbot's predicted % was the inverse of that--11.4% This was
very cool and it took me a second to realize why this was the case.
Mario's actual % was 74.1%.
The
standard error for the regression was .1116, and the Adjusted R-square
was .736. The correlation (e.g., Multiple R) between the predicted % an
the actual % was a very high .86.
The most overpredicted was
Terry Bogard: he had a predicted % of 51.8%, his actual was 35.2%. The
most underpredicted was Bomberman: his predicted was 51.5%, his actual
was 78%. While these and some other predictions were poor, the standard
error was 11.16%, so they aren't completely ridiculous.
NOTE: I
could actually add in data from later rounds and thus double my number
of data points. However, this would require a lot of manual data entry.
So I'm not gonna do that anytime soon.
I will post the predicted percentages of the Spring 2K4 1st round match-ups shortly. --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: ChichiriMuyo | Posted: 3/21/2004 10:50:10 PM | Message Detail |
[This message was deleted at the request of the original poster] |
From: ChichiriMuyo | Posted: 3/21/2004 10:50:46 PM | Message Detail |
I've only got room for 160 characters, so I have to cut out anything unimportant. --- I`ve become sort of a Chichiri fan since then. - UltimaterializerX "...you kick more ass with each post you make" - Ulti again talking about me. |
From: creativename | Posted: 3/21/2004 10:54:36 PM | Message Detail |
[This message was deleted at the request of the original poster] |
From: creativename | Posted: 3/21/2004 10:57:56 PM | Message Detail |
[This message was deleted at the request of the original poster] |
From: creativename | Posted: 3/21/2004 11:22:16 PM | Message Detail |
[I
originally posted the expected 1st round % breakdowns for each game,
but that looked very ugly. I posted my results in a gambling spread
format in GunMage's thread, and that looked much nicer, so I'm going to
repost that here]
A few notes: I used the current scoring system
weights as a proxy for CJayC's odds. I'm guessing it probably isn't too
different from the odds system Ceej used, but who knows.
Also,
the system clearly isn't perfect. SMB3 probably won't get 93.5% of the
vote against Metal Gear, Chrono Trigger ain't gonna get 95.7%...and it
just so happens that since Final Fantasy VII is such a huge favorite,
this model actually predicts it'll get over 100% against Suikoden 2! :)
But anyway, here's the 1st round predicted %es.
Division 8 Super Mario Bros. 3 (-87.1%) vs. Metal Gear Metroid (-36.1%) vs. Pac-Man Contra (-3.3%) vs. Phantasy Star Final Fantasy (-58.1%) vs. Pitfall Donkey Kong (-12.1%) vs. Duck Hunt Legend of Zelda (-71.6%) vs. Adventure Pong (-19.7%) vs. River City Ransom Tetris (-59.9%) vs. Galaga
Division 16 Chrono Trigger (-91.5%) vs. Secret of Mana Super Mario RPG (-2%) vs. Street Fighter 2 Sonic the Hedgehog 2 (-42.8%) vs. Shining Force Super Mario World (-55.5%) vs. The Simpsons Legend of Zelda: A Link to the Past (-67.3%) vs. Gunstar Heroes Super Metroid (-46.7%) vs. Phantasy Star IV Doom (-2%) vs. Earthbound Final Fantasy 6 (-61.2%) vs. Mortal Kombat
Division 32-64 Final Fantasy 7 (-100.3%) vs. Suikoden II Xenogears (-4.8%) vs. Pokémon Gold/Silver/Crystal Dance Dance Revolution (-0.2%) vs. Final Fantasy Tactics Metal Gear Solid (-52.3%) vs. Resident Evil Castlevania: Symphony of the Night (-26.9%) vs. Perfect Dark Goldeneye (-50.2%) vs. Panzer Dragoon Saga Super Mario 64 (-39.2%) vs. NiGHTS into dreams Legend of Zelda: Ocarina of Time (-90.7%) vs. Fallout 2
Division 128 Halo: Combat Evolved vs. Starcraft : PUSH Kingdom Hearts (-57.2%) vs. Soul Calibur Legend of Zelda: The Wind Waker (-66.6%) vs. Skies of Arcadia Metroid Prime (-49.7%) vs. Half-Life Final Fantasy Tactics Advance (-36.5%) vs. Fire Emblem Final Fantasy 10 (-71.9%) vs. Shenmue Grand Theft Auto: Vice City (-18.4%) vs. Star Wars: Knights of the Old Republic Super Smash Bros. Melee (-57.9%) vs. Metal Gear Solid 2 --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: torey luvullo | Posted: 3/22/2004 5:39:31 AM | Message Detail |
[This message was deleted at the request of the original poster] |
From: torey luvullo | Posted: 3/22/2004 5:40:44 AM | Message Detail |
sorry
- what i meant to say was i did some analysis of the extremes in gun
mage's betting spread topic that i will not repeat here. suffice it to
say that the largest spread in both summer character contests was 84.6%. |
From: Slowflake | Posted: 3/22/2004 5:47:12 AM | Message Detail |
Thanks for the info. I'll try that out when I have the time, and that means, not this morning. ;)
LOL, that system of yours predicts DDR will beat FFT with 50.1% of the vote. --- SpC2K4 Status --- Winner: FF7 --- Finalist: CT --- Semifinalists: SMB3, LoZ:WW |
From: smitelf | Posted: 3/22/2004 9:12:57 AM | Message Detail |
LOL, that system of yours predicts DDR will beat FFT with 50.1% of the vote.
Hehe, yeah, I noticed that, too, but hey, if that's what's being predicted ;)...nice stats btw, creativename. --- "Your fate is sealed, and none but yours." -- Auron Proud Supporter of Starcraft in the Spring 2004 Contest |
From: creativename | Posted: 3/22/2004 10:07:34 AM | Message Detail |
LOL, that system of yours predicts DDR will beat FFT with 50.1% of the vote.
Yeah...but
that's only because of that one bracket that has DDR going all the way
;) That's the only bracket where DDR advances past the 3rd round. Take
out those picks for it winning more than 2 matches, and Tactics become
favored by 38%. Which I hope is a lot more reasonable.
I'd say
that Mario 64 is under-favored against NiGHTS (-39.2%), and SSBM is
highly over-favored against MGS2 (-57.9%). Mario 64 is under-favored
because almost nobody has it beating Ocarina of Time. And SSBM is
over-favored because no one has MGS2 going anywhere, but there are some
people who have SSBM going far (due to its "upside"). Other than that,
there's nothing that sticks out to me as being wildly off.
Off-topic, but I just found out what Adventure was.
http://www.cnspace.net/cge/atari2600/atarigames.html
From the looks of it, Legend of Zelda was a rip-off of Adventure?!
I had no idea. But I guess the basic Zelda formula (dungeons, keys,
different weapons/special items) all were inspired by Adventure. Wow.
Of course, it's still going to get no more than 15% against Legend of Zelda. --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: creativename | Posted: 3/22/2004 10:14:55 AM | Message Detail |
Other than that, there's nothing that sticks out to me as being wildly off.
That's
also including the aforementioned SMB3, Chrono Trigger, and especially
Final Fantasy VII spreads. Also throw in Ocarina of Time vs. Fallout 2
(Ocarina of Time is favored by 90.7% there).
I think it's not
unreasonable to think Suikoden II can get 25%. I'd be surprised if it
got significantly under 20%. The worst I see happening is Snake vs.
Raiden. I've never played it, but Suikoden II gets massive respect from
RPG fans.
Fallout 2 is also very beloved, despite the PC's
weakness around here. OoT isn't going to get close to 95% against it,
obviously. I don't think you can expect it to get more than 85% of the
vote against Fallout 2, if that. --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: solarshadow | Posted: 3/22/2004 12:00:30 PM | Message Detail |
How so? Just up your posts per page to 50, and you only have to archive 10 pages.
10
pages per topic times 2 (or more) topics times about 15 links per page
starts to add up to a lot of time spent in Tripod's crappy little HTML
editor.
I have Part 2, 3 and 4 of the off-season discussions,
but I'm just too lazy to edit (getting rig of ad space on the left
mainly) and upload them... I'll try to do it before my HD crashes.
Is
that a crack at me? :) My own laziness at getting the second topic up
last year and my subsequent HD crash is why that topic seems to be gone
for good.
I suggest getting familiar with PivotTables. They
are very useful, and once you get the hang of them very powerful: you
can do time consuming tasks in very short periods of time.
Oh,
very cool. I've been trying to figure out how to do stuff like that in
Excel. You are so much more handy than the Office Assistant. :)
This inspired me to work with CJayC's 2K2 odds to find out if they were useful for predicting %es.
This
is facinating stuff. Could you use a similar technique to estimate
prediction percentages? If you could find something that was reasonably
accurate we might be able to figure out how CJayC determined the odds.
I've played around with a couple of possibilities but haven't been able
to figure it out. The closest I got was pred%*16764^x (where x is the
round number and 16764 is of course the number of brackets). With
reasonable estimates in the final rounds I could get it to fairly
accurately predict the odds, but because only the final percentage
actually affected the odds calculation significantly, it really didn't
turn out to be very objective... (or a good weighting system). ;) --- Contest Stats: http://solarshadow-stats.tripod.com |
From: Heroic Tails | Posted: 3/22/2004 12:03:30 PM | Message Detail |
Is
that a crack at me? :) My own laziness at getting the second topic up
last year and my subsequent HD crash is why that topic seems to be gone
for good.
Absolutely not. I didn't even know that happened. O_o --- I'd rather be damned for my own beliefs, than be damned for someone else's. - Shake [This sig has been approved by UltimaterializerX] |
From: Haste2 | Posted: 3/22/2004 2:32:49 PM | Message Detail |
Well,
I actually did use the odds of the 2002 contest to predict what the
prediction percentages were a long time ago...I admit that I simply
assumed that odds 1:x was 1/x because of the same reason creativename
came up with. All I did was use the odds of the winner and divide by
the sum of all the odds in that "section." Of course, this is all
assume that the characters have the exact same support for every round,
which obviously isn't the case. As exepcted, the predicted prediction
(sorry for being so redundant) percentages seem to get closer to the
real prediction percentages as the later rounds go by. Mario and
Sephiroth seem pretty wacked, though.
I hate it when I spend a
bunch of time typing something up, only to realize that it probably
won't be of much worth. Well, I hope someone will care about this....
Mario vs. Servbot - 99.21% +3.34 Spyro the Dragon vs. Morrigan Aensland - 64.00% +21.72 Terry Bogard vs. Aya Brea - 44.44% -11.83 Donkey Kong vs. Bub - 87.72% -6.46 Alucard vs. Tails - 81.08% +29.67 Duke Nukem vs. Iori Yagami - 75.00% -8.87 Pikachu vs. PaRappa the Rapper - 80.65% +15.83 Cloud Strife vs. Fox McCloud - 96.62% +16.42 Lara Croft vs. Chop Chop Master Onion - 92.31% +3.80 Ryo Hazuki vs. Guybrush - 54.55% -22.72 Crono vs. Simon Belmont - 94.79% +18.40 Dante vs. Q*Bert - 94.34% +19.85 Kasumi vs. Aeris - 73.53% -6.09 Gordon Freeman vs. Tina Armstrong - 31.03% -12.68 Knuckles vs. Akira Yuki - 76.92% +7.9 Solid Snake vs. Squall - 94.20% +10.32 Pac-Man vs. Goemon - 92.59% +10.9 Abe vs. Kyo Kusanagi - 62.5% +11.99 Scorpion vs. Kazuya - 66.67% -6.22 Max Payne vs. Dirk the Daring - 96.15% +4.23 Bomberman vs. Kane - 54.55% -16.39 Jill Valentine vs. Kirby - 35.71% -13.26 Strider Hiryu vs. Raziel - 39.02% -16.77 Link vs. Little Mac - 98.81% +2.42 Sonic the Hedgehog vs. Harry - 97.80% +2.21 Tidus vs. Claire Redfield - 81.63% +20.83 Samus Aran vs. Ken Masters - 91.74% +11.62 Ryu vs. CATS - 80.36% -6.62 Mega Man vs. Ms. Pac-Man - 95.69% + 1.61 Serious Sam vs. Mr. Driller - 71.43% +6.56 Sephiroth vs. Gabe Logan - 99.11% +8.94 Crash vs. Ulala - 75.00% - 6.34
Round 2: Mario vs. Morrigan - 95.26% +13.55 Aya vs. DK - 41.67% -27.91 Alucard vs. Duke Nukem - 47.55% +18.77 Pikachu vs. Cloud - 84.77% +15.26 Lara Croft vs. Ryo Hazuki - 68.97% -5.16 Crono vs. Dante - 49.08% -2.69 Aeris vs. Tina Armstrong - 51.55% -1.93 Knuckles vs. Snake - 90.27% -14.09 Pac-Man vs. Kyo Kusanagi - 71.43% +6.58 Scorpion vs. Max - 28.99% -4.63 Bomberman vs. Jill - 24.72% -7.73 Strider vs. Link - 90.56% +10.07 Sonic vs. Tidus - 61.16% +.39 Samus vs. Ryu - 52.51% -5.51 Mega Man vs. Serious Sam - 79.68% -4.98 Sephiroth vs. Crash - 94.61% +17.49
Round 3: Mario vs. DK - 86.77% +14.25 Alucard vs. Cloud - 62.49% +8.96 Lara Croft vs. Crono - 33.39% +.34 Aeris vs. Snake - 78.11% +10.6 Pac-Man vs. Scorpion - 14.39% -2.94 Jill vs. Link - 76.32% +7.31 Sonic vs. Samus - 31.13% -.62 Mega Man vs. Sephiroth - 76.45% +20.71
Round 4: Mario vs. Cloud - 64.11% +19.87 Crono vs. Solid Snake - 12.08% -2.55 Scorpion vs. Link - 57.90% +.89 Samus vs. Sephiroth - 47.37% +9.03
Round 5: Mario vs. Crono - 29.91% -1.05 Link vs. Sephiroth - 26.08% -4.55
Round 6: Mario vs. Link - 11.30% -1.47
--- "Ah, a party! We haven't had one of those. It could be fun! So...what is a party?" "Well, you drink punch and eat CAKE! ...I think." |
From: creativename | Posted: 3/22/2004 5:18:16 PM | Message Detail |
10
pages per topic times 2 (or more) topics times about 15 links per page
starts to add up to a lot of time spent in Tripod's crappy little HTML
editor.
I'm not sure what you mean by 15 links per page. Are
you talking about changing the "next page" links? Perhaps just using
frames and telling people to use a navigation frame, rather than
clicking on the "next page" links, would be better.
You are so much more handy than the Office Assistant. :)
Well, I would hope so! ^_^
Could
you use a similar technique to estimate prediction percentages? If you
could find something that was reasonably accurate we might be able to
figure out how CJayC determined the odds.
I don't see how
this would be useful for trying to reverse engineer CJayC's odds
system, but I did run a regression trying to predict user prediction
percentages. I thought it might be interesting. The input was once
again LN(Favorite Ratio) [as well the signed square of it; this extra
factor didn't really help the model, but I was getting some curvature
in the function so I decided to use it] I'm not going to go into the
stats details here, I'll just give the results.
This actually
took me a very long time to do...and now that I've done it, I don't
believe that it has very much utility. And while it's a good fit
model-wise (the Multiple R is .825), it probably isn't really all that
accurate. A big reason for that is that the odds sometimes don't even
match up to the prediction percentages: for instance, Morrigan had
better odds to win than Spyro did, yet Spyro was picked by more people
to win the match. The same happened with Terry/Aya, Jill/Kirby, and
Strider/Raziel.
Below is an estimate of how many people would've
picked each character, in each match that took place (for instance, the
estimate is that in a Mario vs. Link match, 56.7% would've picked
Mario, and 43.3% would've picked Link):
Round 1: Mario (94.2%) VS. Servbot (5.8%) Spyro the Dragon (40.7%) VS. Morrigan Aensland (59.3%) Terry Bogard (53.7%) VS. Aya Brea (46.3%) Donkey Kong (77.3%) VS. Bub (22.7%) Alucard (71.4%) VS. Miles "Tails" Prower (28.6%) Duke Nukem (65.5%) VS. Iori Yagami (34.5%) Pikachu (71.1%) VS. PaRappa The Rapper (28.9%) Cloud Strife (88.9%) VS. Fox McCloud (11.1%) Lara Croft (82.4%) VS. Chop Chop Master Onion (17.6%) Ryo Hazuki (53.1%) VS. Guybrush Threepwood (46.9%) Crono (84.4%) VS. Simon Belmont (15.6%) Dante (85.1%) VS. Q*Bert (14.9%) Kasumi (34.2%) VS. Aeris (65.8%) Gordon Freeman (62.6%) VS. Tina Armstrong (37.4%) Knuckles the Echidna (68.2%) VS. Akira Yuki (31.8%) Solid Snake (84.9%) VS. Squall Leonhart (15.1%) Pac-Man (82.7%) VS. Goemon (17.3%) Abe (41.7%) VS. Kyo Kusanagi (58.3%) Scorpion (61.1%) VS. Kazuya Mishima (38.9%) Max Payne (88%) VS. Dirk the Daring (12%) Bomberman (53.1%) VS. Kane (46.9%) Jill Valentine (40.5%) VS. Kirby (59.5%) Strider Hiryu (42.7%) VS. Raziel (57.3%) Link (93.4%) VS. Little Mac (6.6%) Sonic the Hedgehog (91.2%) VS. Pitfall Harry (8.8%) Tidus (71.9%) VS. Claire Redfield (28.1%) Samus Aran (81.7%) VS. Ken Masters (18.3%) Ryu (70.9%) VS. CATS (29.1%) Mega Man (87.2%) VS. Ms. Pac-Man (12.8%) Serious Sam (64.3%) VS. Mr. Driller (35.7%) Sephiroth (94%) VS. Gabe Logan (6%) Crash Bandicoot (66.8%) VS. Ulala (33.2%) --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: creativename | Posted: 3/22/2004 5:18:43 PM | Message Detail |
Round 2: Mario (90.4%) VS. Morrigan Aensland (9.6%) Aya Brea (40.6%) VS. Donkey Kong (59.4%) Alucard (57.4%) VS. Duke Nukem (42.6%) Pikachu (20.8%) VS. Cloud Strife (79.2%) Lara Croft (73.3%) VS. Ryo Hazuki (26.7%) Crono (51.5%) VS. Dante (48.5%) Aeris (74.5%) VS. Tina Armstrong (25.5%) Knuckles the Echidna (11.2%) VS. Solid Snake (88.8%) Pac-Man (73.3%) VS. Kyo Kusanagi (26.7%) Scorpion (39.9%) VS. Max Payne (60.1%) Bomberman (44.5%) VS. Jill Valentine (55.5%) Strider Hiryu (11.3%) VS. Link (88.7%) Sonic the Hedgehog (61.1%) VS. Tidus (38.9%) Samus Aran (57%) VS. Ryu (43%) Mega Man (76.6%) VS. Serious Sam (23.4%) Sephiroth (88.7%) VS. Crash Bandicoot (11.3%)
Round 3: Mario (87.6%) VS. Donkey Kong (12.4%) Alucard (26.7%) VS. Cloud Strife (73.3%) Lara Croft (43.2%) VS. Crono (56.8%) Aeris (18.2%) VS. Solid Snake (81.8%) Pac-Man (64.3%) VS. Scorpion (35.7%) Jill Valentine (13.5%) VS. Link (86.5%) Sonic the Hedgehog (46.3%) VS. Samus Aran (53.7%) Mega Man (26.7%) VS. Sephiroth (73.3%)
Round 4: Mario (63.3%) VS. Cloud Strife (36.7%) Crono (29%) VS. Solid Snake (71%) Scorpion (14.9%) VS. Link (85.1%) Samus Aran (29.4%) VS. Sephiroth (70.6%)
Round 5: Mario (74.4%) VS. Crono (25.6%) Link (50%) VS. Sephiroth (50%)
Round 6: Mario (56.7%) VS. Link (43.3%) --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: creativename | Posted: 3/22/2004 5:26:19 PM | Message Detail |
Trying
to map this information into estimates of how many people picked each
character to go a certain distance could be done (I think), however it
would be very complex, and require a lot of Conditional
Probability/Bayesian calculations. I'm not even sure if I can do that,
but I'll give it a try sometime.
That's what I was actually
intending to do before I realized how complex it would be, since for
each round, you have to match up each character with all their possible
opponents. It'd be ridiculous in straight Excel, but perhaps it could
be done in a programming language with arrays.
I probably will attempt this, because I'm curious, but I don't think it's going to be useful for anything. --- Remember to nominate Frog from Chrono Trigger for Summer Contest 2K4! |
From: Slowflake | Posted: 3/22/2004 5:29:35 PM | Message Detail |
So
Scorpion was only the favorite in his first round match, and not even
by much at that against a complete jobber. Funny how it was thought
that MAX PAYNE would beat him. --- SpC2K4 Status --- Winner: FF7 --- Finalist: CT --- Semifinalists: SMB3, LoZ:WW |
From: ChichiriMuyo | Posted: 3/22/2004 6:20:05 PM | Message Detail |
From
the looks of it, Legend of Zelda was a rip-off of Adventure?! I had no
idea. But I guess the basic Zelda formula (dungeons, keys, different
weapons/special items) all were inspired by Adventure. Wow.
Yeah,
which is why I said that so many of the matchups look like complete
setups... how many spots was adventure raised above it's deserving
seeding (which probably could fall out of the top 16 8-bit games if you
take actual strength) just to get it to match up with Zelda?
Oh, and even though I saw the outcome I'm still surprised that Scorp was even favored over Kazuya. --- I`ve become sort of a Chichiri fan since then. - UltimaterializerX "...you kick more ass with each post you make" - Ulti again talking about me. |
From: solarshadow | Posted: 3/22/2004 10:19:28 PM | Message Detail |
I'm not sure what you mean by 15 links per page. Are you talking about changing the "next page" links?
Yeah.
Each of the page 1-10 links, first page, previous page, next page, last
page, and an index link (like I did for the first topic last year). It
takes a while. And is probably unnecessary, but makes navigation easier
(for those few times when I want to reference something from a middle
page). I should probably just do previous and next and let people
change the url to find a specific page. I don't really want to get into
frames.
Absolutely not. I didn't even know that happened. O_o
Oh. Interesting coincidence then.
I
hate it when I spend a bunch of time typing something up, only to
realize that it probably won't be of much worth. Well, I hope someone
will care about this....
I always care. :)
Crono vs. Simon Belmont - 94.79% +18.40
Yikes. Poor Simon. I think my initial bracket actually had him winning this match. Everyone loves Castlevania, right?
I
don't see how this would be useful for trying to reverse engineer
CJayC's odds system, but I did run a regression trying to predict user
prediction percentages.
I guess I figured that if we could
reasonably accurately estimate the prediction percentages we could then
construct an approximation of all the brackets (which we would then
attempt to match to the odds). I wasn't thinking very clearly earlier
today (I'm still not... must sleep). The fact that we would still be
missing a significant amount of the necessary information didn't occur
to me. All of the magic you were performing on the numbers led me to
imagine you could conjure up anything. ;)
From the looks of it, Legend of Zelda was a rip-off of Adventure?!
I
wouldn't say Zelda was a rip-off of Adventure, per se. Adventure
essentially gave birth to the adventure genre. Zelda sort of defined
the genre (think of Wolfenstein 3D vs. Doom as a similar example). --- Contest Stats: http://solarshadow-stats.tripod.com |
From: Starion | Posted: 3/23/2004 12:50:07 AM | Message Detail |
Solar,
when do you officially start collecting predictions for the Spring
Oracle challenge? The contest is only about a week away.
Castlevania
is popular enough but the early Belmonts were rather bland and
replaceable. Alucard, on the other hand, is different. The fact that he
is associated with the most popular Castlevania game doesn't hurt. --- Spring Contest 2004 - ????? VS. ????? Points - 0/192 |
From: UltimaterializerX | Posted: 3/23/2004 2:03:06 AM | Message Detail |
I`ve been sigged twice in one sig? Holy ****!
And nice analysis up there. Pity I`ve said all of those exact same things already ~_^ --- MIASU!! My SC2K4 Petition: http://cgi.gamefaqs.com/boards/genmessage.asp?board=7&topic=12558738 |
From: UltimaterializerX | Posted: 3/23/2004 4:59:30 AM | Message Detail |
I just realized something. Our little contest is pointless without the list of entrants:
1. UltimaterializerX 2. Shake 3. Starion 4. DomaDragoon 5. Ngamer 6. ChichiriMuyo 7. cyko 8. Z1mzum 9. Haste2 10. Neoatomtaco 11. Heroic Mario 12. i am vishnu 2 13. Yesmar 14. solarshadow 15. StopPokingMe 16. charmander6000 17. torey luvullo
You can thank me later, because I`m pretty sure the other topic would have been purged tonight. --- MIASU!! My SC2K4 Petition: http://cgi.gamefaqs.com/boards/genmessage.asp?board=7&topic=12558738 |
From: Sir Chris | Posted: 3/23/2004 5:05:06 AM | Message Detail |
bracket contest? oh sure, why not --- "Give a pog a rum, and he'll get drunk for an hour. Teach a pog how to make rum, and he'll die within 2 hours." ~ PoG |
From: swirldude | Posted: 3/23/2004 5:40:28 AM | Message Detail |
what's the list for? --- 0oT.
Of course both games are excellent and have equal followings, but
GameFAQs seems to have more RPG fanboys. ~bbb on OoT vs FFVII |
Jump to Page: 1 | | | | | | | | | |