|
It's been a while, so I thought I'd update with some stats about MLG I discovered.
Breakdown of MLG Orlando groups by total prize money won this season:
Group A - $ 4,900
Group B - $10,400
Group C - $ 2,200
Group D - $ 3,700
+ Show Spoiler [By Player] +A
IdrA 1900
Haypro 0
TLO 1000
Boxer 2000
Drewbie 0
4900
B
Kiwi 3000
Bomber5000
Inca 1400
Inc 0
Puma 1000
10,400
C
Slush 1000
Ret 500
Rain 700
Machine 0
HongUn 0
2200
D
HuK 1000
Hero 700
Tyler 0
MC 2000
July 0
3,700
For a sense of how this season's earning match up to IPL3 winnings, take a look at this.
No player has Top10'd all four MLG's this season.
Only three have Top16'd them all: Slush, Naniwa, and Sjow.
Only three players have Top10'd thrice: Naniwa, Idra, and Huk.
Twice: MMA, Kiwi, Slush, Sjow, DRG
And a somewhat odd group- players who have never placed higher than their current MLG points ranking: Idra, Slush, Sjow, Huk. And, trivially, Naniwa of course. I thought this was interesting because it somehow provides a measure of consistency and growth of a player rather than pure title-winning ability.
So many former GSL stalwarts are missing this season that Tyler, who is ranked 30th in MLG points, made it into the top 16 (after Bomber and and Boxer were taken out, as they are the GSL invites) to be seeded into groups.
One last funny thing that takes a bit to explain, so I'm spoilering it:
+ Show Spoiler +
at MLG Raleigh the placements in the middle of the groups did not predict well the placement in the tournament. That is, the 6th place players clustered around 25-21 and the first place players necessarily got 1-6, but rest were kind of scattered around in there. Specifically, the 3rd place players did much better than the 2nd place players, and the 4th place players did no better than the 5th place players. To help see this, here's a list of places followed by the group placement of the player who took that place. For instance, 15:3 means that the 15th place player took third in his group. O means open-bracket.
26: O
25: 6
24: 6
23: 5
22: 6
21: 6
20: 4
19: 4
18: 5
17: 4
16: 5
15: 3
14: O
13: 4
12: 2
11: 2
10: 2
09: 2
08: 3
07: 3
06: 5
05: 1
04: 3
03: 1
02: 1
01: 1
|
I've kept meaning to post this, but it's been such a crazy week I kept forgetting:
Here's an odd little statistic I ran that I call Games in Jeopardy. Essentially, you count up the total number of games in a tournament that a player could have been eliminated in, and it should give a measure of how much they have been dominating their opponents. For instance, if a players x-0's all his opponents, then he was never once put in jeopardy of elimination. If a player always goes to the last game of every match, then he will have lots of GiJs, meaning that he played many games that could have been his last.
I thought it up because I realized that somehow in my brain I subconsciously counted come-back wins for much more than solid beat downs when I tried to reckon players' relative skill, when in fact having lots of amazing comebacks is actually probably a bad predictor for long-term success.
Here is the list for all the GSL individual leagues, listed in order of (winner, loser):
+ Show Spoiler +Fruit 1 Rainbow 0
NesTea 0 MKP 4
MC 2 Rain 4
MVP 0 MarineKing 2 (S Jan)
MC 0 July 4
NesTea 5 IncA 1
NesTea 0 Losira 3
Mvp 0 Top 4
Mvp 1 MarineKing 3 (Champs)
Polt 5 MMA 6 (Super)
In this tournament MMA has risked elimination five times while Mvp was only put on notice once by teammate NesTea.
I'm still pretty on the fence about this statistic, and it is definitely not one to be blindly followed. I have no idea how reliable it could be, and perhaps it only really says "most tournaments have a dominate player that wins the finals," but I think it gives a neat insight into players' paths to the championships.
|
Now that the Up-And-Down matches are changing so dramatically, it makes sense to post some superlatives.
Most visits to the Up-And-Downs: Leenock and MarineKing at 4 apiece. Both had a promotion and a demotion, but MKP had two saves compared to Leenock's two failures to make Code S at all.
Most failures: Leenock and JYP at 2. Both players made it to the Up-And-Downs twice without being promoted. JYP is still trying.
Most Saves without Demotion: Ensnare. Ensnare is the only player currently in Code S to have avoided demotion twice. Kyrix, Inca, Genius, FruitDealer, MarineKing, and Zenio all made two saves, but were eventually demoted.
Most Promotions: Sc is the only player to have successfully toughed it out through Code A and the and Up-And-Downs twice.
Ironman Award: NaDa and Clide STILL have never been to the Up-And-Downs
Bronzeman Award: Additionally, NesTea and Ensnare have never been demoted.
Tinman Award: Polt has never technically missed a season of Code S, despite being demoted once.
Persistance Award: Polt, Rain, Mvp, Leenock, MarineKing, Sc, MC- players who made it back into Code S after being demoted.
Here's the link to the spreadsheet I used to compile these.
|
Poor pullarius1, the new GSL format gonna make your stat-life hell.
Here's an interesting stat I found yesterday - before this season's Code A Prelim, only HerO ever made it through the prelim out of all the players on foreigner teams that tried. With this season, we can add iS and CrazymovING to this small list.
To give some perspective, here are Foreigner teams that had players previously in Code A/S: CoL:  NaNiwaEG: HuK,  IdrAFnatics: RainiS:  CrazymovINGmouZ: ThorZaINQxG: SaSeDignitas: SeleCT, SjoWLiquid: HayprO, HerO, Jinro, Ret, ShethIndependent: mOOnGLaDeNote that NaNiwa, ThorZaIN, SaSe, SeleCT, SjoW, Moonglade, HayprO, and Sheth's spots were gifted or earned through MLGs. So you're really lookig at EG, Fnatics, and Liquid as the only other teams to have qualified players through the prelims or from the Opens. Those three teams are good company to be in. For a more absurd stat - HerO is the only other player from the above list to have qualified from the Code A Prelims. + Show Spoiler [Previous players to have qualified thr…] +
Ace
Alicia
asd
Avenge
Bomber
BoxeR
Brown
Choya
CoCa
Creator
Curious
DongRaeGu
Dream
Extreme
finale
GanZi
GuMiho
Hack
Happy
hero
JYP
Jjakji
Jjun
JookTo
Keen
Leenock
Line
LosirA
Lucky
Lure
Luvsic
MMA
Maka
Maru
Min
Monster
Noblesse
Nuts
Oz
Puzzle
RevivaL
Ryung
SUPERSTAR
Sage
ShinyStar
Sniper
SocceR
Squirtle
SuperNoVa
TAiLS
Taeja
Tassadar
TheBest
TopClass
Violet
Virus
Weekend
Yoda
YuGiOh
|
Haha yeah, I think I might give up on that specific spreadsheet.
That has never clicked for me about foreign teams. I guess I because SO many get to the final round of their prelim group and lose :-/ Although perhaps it is more that the best foreigners get invited to the GSL anyway, and there are still so few Koreans on foreign team. Were you counting fOu under FXO as one?
|
Since this is something I've tracked from the beginning:
With the new format a player can go 5-5 from Code A an be promoted into Code S. (2-1, 0-2 in Code A 3-2, win breaker in Up-And-Down.) But a player can go 7-6 from Code S and be demoted to Code A! (2-0 Ro32, 1-2 Ro16, 1-2 Code A, 3-2 UpAndDown) These are definitely corner cases, but it seems pretty harsh that a player with an even record against mostly Code A players can get promoted while a player with a winning record against mostly Code S players gets demoted.
|
From what I can see, only Sheth and moonglade were ever Code A from the non-fOu side of FXO, and (un)luckly both have since moved on from the team, so my argument stand.
Still trying to get my head around the new format, my gut feeling thinks that there may be circumstances where it's easier to get to code S from one route than another in a counter-intuitive fashion.
|
On October 28 2011 01:07 pullarius1 wrote:Now that the Up-And-Down matches are changing so dramatically, it makes sense to post some superlatives. Most visits to the Up-And-Downs: Leenock and MarineKing at 4 apiece. Both had a promotion and a demotion, but MKP had two saves compared to Leenock's two failures to make Code S at all. Most failures: Leenock and JYP at 2. Both players made it to the Up-And-Downs twice without being promoted. JYP is still trying. Most Saves without Demotion: Ensnare. Ensnare is the only player currently in Code S to have avoided demotion twice. Kyrix, Inca, Genius, FruitDealer, MarineKing, and Zenio all made two saves, but were eventually demoted. Most Promotions: Sc is the only player to have successfully toughed it out through Code A and the and Up-And-Downs twice. Ironman Award: NaDa and Clide STILL have never been to the Up-And-Downs Bronzeman Award: Additionally, NesTea and Ensnare have never been demoted. Tinman Award: Polt has never technically missed a season of Code S, despite being demoted once. Persistance Award: Polt, Rain, Mvp, Leenock, MarineKing, Sc, MC- players who made it back into Code S after being demoted. Here's the link to the spreadsheet I used to compile these.
Very weird to see Ensnare and NesTea ever being put in the same category. 
You should give an award to YuGiOh. Six seasons in Code A before making it into Code S.
|
Thank you so much for doing this I love reading these stats.. but I'm too lazy to look them up on my own unless it's in relation for a post/writing or something.
|
Now that we have had a huge change in the GSL format, I'm thinking about resetting the statistics I keep. As such, here are a bunch of the long-term stats I've kept:
Running percentages of races by round (Ro32, Ro16, Ro8, Ro4)
P: 25%, 26%, 30%, 20%
T: 40%, 49%, 47%, 60% (!!!!!)
Z: 26%, 27%, 22%, 18%
And then for total spots (so a Terran that gets three rounds counts for 3)
Protoss: 26% Terran: 48% Zerg: 27%
So terrans account for 48% of all player*games in Code S. Even more disturbing is that terrans have taken a whopping 60% of all semifinals slots, which just feel absurd despite the relatively small sample size. Protoss and zerg still seem fairly even, except protoss are oddly better at getting to the quarter finals than zergs.
Win Percentages:
PvT: 136 - 166 (45%)
PvZ: 94 - 115 (45%)
TvZ: 152-155 (50%)
Shouldn't be all that surprising. These statistics are a tad misleading, though, due to the fact that the race of the two finalists has an inordinate effect in the winning percentages. That sounds odd to say, but I think Protoss's plight is better understood in the previous set of numbers. It should be noted, though, that Protoss has had only three winning seasons: The two seasons MC won, and, oddly, the World Champs.
Number of mirror matches by race:
Protoss: 64 (17%)
Terran: 258 (69%)
Zerg: 53 (14%)
So a whopping 69% of all mirror matches in Code S have been TvT. In the last tournament >50% of ALL matches were TvT. I'm tempted to calculate that number for the entirety of the GSL. Maybe later.
Here's an odd statistic that I've posted a few times on here: whether or not a race has lost >50 <50 or exactly 50% of its players from one round to the next. Here's the raw data:
+ Show Spoiler + B = more than 50% lost
L = less than 50% lost
H = exactly half lost
0 = only one or less of that race remained in that round
NB: only logged up until the semifinals or each round
Toss:
BLBB0
BBB00
HLLLH
BH00
LLLB
LHB
HLB0
BB000
LLB0
HHB0
B000
Terr
LBLLB
BBLLB
LLBLH
LBLL
BBB0
HLL
HBLB
LLLHH
BBH0
LLLL
HLLH
Zerg:
BBH00
LLBB0
BBHB0
BLH0
HB00
BB0
HHH0
BLB00
HLLL
BB00
LBB0
Toss: BLBB0|BBB00|HLLLH|BH00|LLLB|LHB|HLB0|BB000|LLB0|HHB0|B000
Terr: LBLLB|BBLLB|LLBLH|LBLL|BBB0|HLL|HBLB|LLLHH|BBH0|LLLL|HLLH
Zerg: BBH00|LLBB0|BBHB0|BLH0|HB00|BB0|HHH0|BLB00|HLLL|BB00|LBB0
Number of rounds in which each race had one or less player:
Protoss - 14
Terran - 2
Zerg- 14
So Terran has had at least two players in every round before the finals except for two, while the other two races have fourteen in which at most a sole representative was present.
Percentage of rounds with >2 players in which less than half were lost:
Protoss - 39%
Terran - 55%
Zerg - 29%
Terran is MORE LIKELY THAN NOT to lose less than half of its players in any given round of the GSL. This, I felt, showed that it was not simply the amount of Terran players that accounted for their success in the GSL.
Finally, an update on the Lineal Championship, which has finally moved back to Code S! We left off with Tails upsetting Mvp to claim the crown in the GSTL.
+ Show Spoiler [Lineage] +Tails
Tassadar
Keen
Seal
Genius
BboongBboong
Keen
Oz
Yugioh
|
This thread always amaze me.
Thank you <3
|
Surprising fact that could have some serious repercussions in the new Code S format: Only six players have ever Top 8'd two Code S's in a row: NaDa, Anypro, sC, Losira, NesTea, and Mvp. Only NaDa has strung together three in a row.
In fact, since July, when nobody from the June quarterfinals managed to Top 8, only Mvp has managed to hit the Top8 twice in a row.
The most Top 8 repeaters were in March>May with three.
This is pretty crazy considering that only the Top 8 of Code S are safe this season.
|
So basically, Terran is too strong. Anyone that follows the game already knows that. =P
Thanks for more numbers to support our hypothesis though!
|
Hey all! Sorry it's been so long since I updated. The reason is that I've started writing for MLG! I have a pretty regular column there called "SC2 By the Numbers" along with other stuff throughout the week. Here's a sampling of the stuff I've done this week:
Some NASL-related stats
Some stats on GSL Mirror matches
Code A stuff
and some silliness with the MLG Viewership Numbers
If you want to see more, you can just search for my real name, since the MLG search is really finicky.
I'll still try to post tidbits here or on twitter, but the really juicy stuff will go to MLG and SC Report :-)
|
Congrats on MLG, pullarius!
Finally, someone who can cite numbers without making me want to dig my eyes out.
|
|
|
Oh you don't know, pullarius. Last night someone went on lo3 as "the numbers guy" and it just made me raged so hard.
HACKKKKKKKKKKKKKKKKKKKKK
|
On December 10 2011 07:52 pullarius1 wrote:Hey all! Sorry it's been so long since I updated. The reason is that I've started writing for MLG! I have a pretty regular column there called "SC2 By the Numbers" along with other stuff throughout the week. Here's a sampling of the stuff I've done this week: Some NASL-related statsSome stats on GSL Mirror matchesCode A stuffand some silliness with the MLG Viewership NumbersIf you want to see more, you can just search for my real name, since the MLG search is really finicky. I'll still try to post tidbits here or on twitter, but the really juicy stuff will go to MLG and SC Report :-)
Ha, grats, was it you who posted the protoss in GSL article ?
|
On December 22 2011 01:59 MrCon wrote:Show nested quote +On December 10 2011 07:52 pullarius1 wrote:Hey all! Sorry it's been so long since I updated. The reason is that I've started writing for MLG! I have a pretty regular column there called "SC2 By the Numbers" along with other stuff throughout the week. Here's a sampling of the stuff I've done this week: Some NASL-related statsSome stats on GSL Mirror matchesCode A stuffand some silliness with the MLG Viewership NumbersIf you want to see more, you can just search for my real name, since the MLG search is really finicky. I'll still try to post tidbits here or on twitter, but the really juicy stuff will go to MLG and SC Report :-) Ha, grats, was it you who posted the protoss in GSL article ?
Yes it was :-)
MLG is cutting back on written stuff now, though, so look for more stuff on ESFI.
|
Well... that was short :-/ You'll still be able to find my stuff on ESFI though.
Anyway, a quick stat: over ALL the individual SC2 tournaments Gom has run (So All Code S, Code A, Opens and special tournaments) only 5 Protoss players have ever made any finals: MC (Sx2), Inca (S), Puzzle(A), Oz(A), and Tassadar(A).
Over 22 tournaments there have been 44 finalists. Of those, 24 have been Terran, 14 have been Zerg, and 6 have been Protoss. Or, by percentage, a 55/32/14 split, meaning that Terrans have had four times more finalists than have Protoss players.
|
|
|
|
|
|
EPT
