| Forum Index > SC2 General |
|
Slasher
United States1095 Posts
Wins by Matchup: Entire Tournament: Race Breakdown - P: 85 - Terran: 87 - Zerg: 70 TvZ: Terran: 57% Zerg: 43% 154-115 PvZ: Protoss: 48% Zerg: 52% 106-114 PvT: Protoss: 51% Terran: 49% 149-143 APM (Calculated using in-game time): coL.DongRaeGu: 274 oGs.NaDa: 270 Here's the full infographic and numbers: http://www.majorleaguegaming.com/news/mlg-raleigh-starcraft-2-by-the-numbers | ||
|
Slasher
United States1095 Posts
Columbus: PvZ: Protoss: 46.51% Zerg: 53.49% PvT: Protoss: 54% Terran: 46% TvZ: Terran: 43.21% Zerg: 56.79% Anaheim: PvZ: Protoss: 51.1% Zerg: 48.9% 144-138 PvT: Protoss: 48.3% Terran: 51.7% 171-183 TvZ: Terran: 56.8% Zerg: 43.2% 142-108 | ||
|
ZAiNs
United Kingdom6525 Posts
 . | ||
|
teddyoojo
Germany22369 Posts
| ||
|
Grackodile
United States263 Posts
| ||
Kinky
United States4126 Posts
| ||
|
Amui
Canada10567 Posts
| ||
|
ohampatu
United States1448 Posts
Edit: id much rather see just the stats from group play and brackets. | ||
|
Elefanto
Switzerland3584 Posts
On September 16 2011 02:07 Grackodile wrote: I don't know why there is so much QQ lately about PvT and PvZ. I know it's just one tournament and all but the only real mu that varies a great deal is ZvT. These numbers aren't indicative of anything, because the discrepancy in skill is often so huge. | ||
|
Risen
United States7927 Posts
On September 16 2011 02:09 ohampatu wrote: I dont know how good these results are. Considering the top end of MLG has people of the likes as DRG/Hero, etc. While the bottom end has players at like my skill level. Edit: id much rather see just the stats from group play and brackets. This can't be emphasized enough. | ||
|
ohampatu
United States1448 Posts
On September 16 2011 02:10 Elefanto wrote: These numbers aren't indicative of anything, because the discrepancy in skill is often so huge. You make no sense there. These numbers are indicative of alot. I dont need to see the stats of July rolling through open bracket for instance. The stats are skewed because you have alot of good players beating alot of shitty players. Yes, apm, etc can still be found. But you cant honestly graph win/los percentage, when you have players like july/hero rolling open bracket. | ||
|
Madera
Sweden2672 Posts
| ||
|
Derez
Netherlands6068 Posts
On September 16 2011 02:07 Grackodile wrote: I don't know why there is so much QQ lately about PvT and PvZ. I know it's just one tournament and but the only real mu that varies a great deal is ZvT. The stats are incredibly skewed for TvZ anyhow. There were 10 matches between koreans T's and foreigner Z's, only 3 matches between koreans Z's and foreigner T's. All of the koreans won. Makes the win percentage rather meaningless. On September 16 2011 02:09 Amui wrote: The only matchup I think which needs some help is TvZ. There's something Terran gets that Zerg doesn't. Yeah, more good koreans T's playing foreigner Z's then the other way around. | ||
|
ohampatu
United States1448 Posts
Do this, but only do it from the group play and championship bracket. | ||
|
RinconH
United States512 Posts
The Koreans are going to win no matter what their race. GSL is the best thing to look at as the skill levels are more uniform. | ||
|
BroboCop
United States373 Posts
| ||
|
ohampatu
United States1448 Posts
On September 16 2011 02:38 BroboCop wrote: serious question: what can luckyfool do with 230 apm on 1base? Lol. You should see some of the old meme's. I remember a particular one that just had a picture of flash, top part said 300 apm, bottom part said 'only 5 scv's'. Cracked me up everytime i saw it. | ||
|
-_-
United States7081 Posts
On September 16 2011 02:38 BroboCop wrote: serious question: what can luckyfool do with 230 apm on 1base? loooooooooooooooooooooooooolllllllllllllllllllllllllllllllllllllll i like luckyfool, but that's damn funny :p | ||
|
genius_man16
United States749 Posts
On September 16 2011 02:38 BroboCop wrote: serious question: what can luckyfool do with 230 apm on 1base? Make his workers work extra hard? | ||
|
fuzzayy
United States99 Posts
lol yea | ||
|
windsupernova
Mexico5280 Posts
¨Mine harder, goddamnit!¨ | ||
|
Micket
United Kingdom2163 Posts
I just fail to see how you can look at Xelnaga, and say that it is balanced. Best 2 rax map, best hellion map, best map for denying scouting (tiny main), amazingly OP gold base for T, difficult 3rd for Z, very difficult to counter attack due to short rush distance; just everything on that map favours Terran except easier drop defence. Sure, the map can be ok if Terran does a standard 2 base all in/timing and it gets absolutely mauled, but if Terran plays the map the way MVP plays it, I just can't see what you can do. | ||
|
Lysergic
United States355 Posts
On September 16 2011 02:14 MaderA wrote: Awesome stats, thank you. It's weird how TvZ went from 43.21% in Columbus to 57% in Raleigh. Columbus had Losira, Moon, and July (July in open bracket). Idra, Ret, and Slush also had a lot of wins. Only korean terran was MMA I think (Select doesn't live in korea). 5 Zergs in the top 8. Anaheim had Ganzi, Mvp, MMA, Boxer, Rain (Rain and Ganzi in open bracket). Only korean zerg was DRG. This was also when the new BF Hellion style was revealed. 5 Terrans and 1 Zerg in the top 6, all korean. Raleigh had Puma, Bomber, Rain, Nada, Noblesse. Puma and Noblesse were in the open bracket, and even Trimaster beat a lot of zergs, including Idra and Haypro. Only korean zergs were Coca and DRG. Look at Raleigh's Pool B results alone, there was a 10% ZvT match winrate. (1-9 only because Haypro beat Sjow). It's easy why you can't take the matchup statistics seriously. The open bracket obviously has such a wide skill gap, and the koreans (such as Hero, Puma, Noblesse) that blaze through the open bracket skew the results. The Championship/Pool statistics carry no meaning either, because there are too few games and the results have a lot to do with what koreans are there. The players who keep getting reseeded (such as incontrol, machine, haypro, moonan) make it even more meaningless. I'm guessing a lot of people are still gonna say dumb things like "lol toss complaining about PvT" | ||
|
Montana[TK]
1624 Posts
| ||
|
ohampatu
United States1448 Posts
On September 16 2011 03:20 Montana[TK] wrote: can't wait for the APM changes in 1.4 so the stat gains a little more meaning That stat doesn't really have any meaning. Even after the changes you can still spam it up just like before. Just now you can't go '123, 123, 123, 123, 123'. People will still spam. I generally wont even look at apm, and if i do, i wait untill past 50 food so that i get a reasonble result. | ||
Noam
Israel2209 Posts
On September 16 2011 02:04 Slasher wrote: Race Breakdown - P: 85 - Terran: 87 - Zerg: 70 This stat confuses me a bit as it suggests the sum of players participating is 242. According to the MLG website there were 235 open bracket players (256 - 21 Byes). If you add the 16 Pro Circuit players and 4 GSL invites this results in 255 players total. So either 13 players are missing from the race stats or they are not accounting for all the players for whatever reason. I'm having a hard time finding a possible explanation for this as scenarios like players not showing up are properly documented in the bracket as DNS instead of Bye. | ||
|
Exarl25
1887 Posts
On September 16 2011 03:20 Montana[TK] wrote: can't wait for the APM changes in 1.4 so the stat gains a little more meaning It will have less meaning, not more. Legitimately cycling through control groups will no longer count but clicking 50 times to move an SCV will still cause your to APM skyrocket. | ||
|
alepov
Netherlands1132 Posts
On September 16 2011 02:09 ohampatu wrote: I dont know how good these results are. Considering the top end of MLG has people of the likes as DRG/Hero, etc. While the bottom end has players at like my skill level. Edit: id much rather see just the stats from group play and brackets. Same skill difference in groups between people like bomber and incontrol etc. | ||
|
HwangjaeTerran
Finland5967 Posts
On September 16 2011 03:39 Noam wrote: This stat confuses me a bit as it suggests the sum of players participating is 242. According to the MLG website there were 235 open bracket players (256 - 21 Byes). If you add the 16 Pro Circuit players and 4 GSL invites this results in 255 players total. So either 13 players are missing from the race stats or they are not accounting for all the players for whatever reason. I'm having a hard time finding a possible explanation for this as scenarios like players not showing up are properly documented in the bracket as DNS instead of Bye. Well numberwise 13 randoms wouldn't be too far fetched? | ||
|
Noam
Israel2209 Posts
On September 16 2011 03:47 HwangjaeTerran wrote: Well numberwise 13 randoms wouldn't be too far fetched? That's actually a very good idea didn't even consider those random players.If its true its unfortunate that MLG didn't mention that there were 13 random players signed up, that's a fairly large number imo. | ||
|
ohampatu
United States1448 Posts
On September 16 2011 03:59 Noam wrote: That's actually a very good idea didn't even consider those random players.If its true its unfortunate that MLG didn't mention that there were 13 random players signed up, that's a fairly large number imo. Why would MLG mention 13 non-pro randoms? | ||
|
Burns
United States2300 Posts
| ||
|
Cadgers
United States514 Posts
| ||
|
Sadistx
Zimbabwe5568 Posts
| ||
|
rfoster
United States1005 Posts
| ||
|
CatNzHat
United States1599 Posts
Would be totally epic. Btw, anyone know what software they're using the parse the replay files in bulk? sc2 gears? | ||
|
arbitrageur
Australia1202 Posts
- Most of the highest APMs are Koreans. - Best players are Koreans. Conclusion: can't make a conclusion... but yeah. Koreans have far far better macro and they also happen to have higher APM. | ||
|
Probe1
United States17920 Posts
Case in point- Idra had at least two games he SHOULD have won but left early. So don't overanalyze the stats, just enjoy them! edit: arbitrageur we tell plat and below players apm doesn't matter. Of course the higher apm you have the better you'll do when you're talking about pros. | ||
|
Ryuu314
United States12679 Posts
On September 16 2011 10:27 arbitrageur wrote: Someone did an analysis of last MLG and found a statistically significant correlation between APM and win rate.... so it does mean something. - Most of the highest APMs are Koreans. - Best players are Koreans. Conclusion: can't make a conclusion... but yeah. Koreans have far far better macro and they also happen to have higher APM. I've found that in BW people really like to overstate the value of APM, while in SC2, people tend to understate the value of APM. APM isn't a "meaningless number." It does mean something. A player who can achieve 400 APM will naturally be able to do more stuff than a player who can only achieve 100 APM. The question becomes whether or not the actions they're doing are meaningful. For the most part, lower level players will inflate their APM through spam, so big numbers aren't indicative of much. But, for higher level players, where few, if any, actions are wasted, APM does say quite a lot. If pro A can do 300 meaningful actions a minute, while pro B can only manage 200, all else constant, it's very likely pro A will be able to gain the upper hand simply by being able to make more shit happen at once. | ||
|
goldendwarf
Canada170 Posts
| ||
|
Kujawa
United States137 Posts
| ||
|
arbitrageur
Australia1202 Posts
On September 16 2011 10:36 Ryuu314 wrote: I've found that in BW people really like to overstate the value of APM, while in SC2, people tend to understate the value of APM. APM isn't a "meaningless number." It does mean something. A player who can achieve 400 APM will naturally be able to do more stuff than a player who can only achieve 100 APM. The question becomes whether or not the actions they're doing are meaningful. For the most part, lower level players will inflate their APM through spam, so big numbers aren't indicative of much. But, for higher level players, where few, if any, actions are wasted, APM does say quite a lot. If pro A can do 300 meaningful actions a minute, while pro B can only manage 200, all else constant, it's very likely pro A will be able to gain the upper hand simply by being able to make more shit happen at once. voice of reason | ||
|
Perseverance
Japan2800 Posts
What's the formula to covert in-game time apm to real time apm? | ||
|
Lysergic
United States355 Posts
On September 16 2011 10:54 Perseverance wrote: Thanks for the stats ^^ What's the formula to covert in-game time apm to real time apm? Multiply the in-game APM by 1.38 DRG's 274 APM = 379 real APM. And the new APM change is going to make APM calculations meaningless. DRG/Losira/Nada and most pro's are going to have significantly lower APM. | ||
|
partisan
United States783 Posts
The large sample size tends to help but I think there could be a clever solution to weeding out some of the more obvious outliers. Unfortunately I am not a clever person, so I'll leave that someone else. | ||
|
ChaosTerran
Austria844 Posts
On September 16 2011 02:56 Micket wrote: best map for denying scouting (tiny main), . Just no, no and no. A small main actually makes it very easy for zerg to scout, if you have a big main it's much harder for zerg to actually scout what the terran is doing. | ||
|
Tachion
Canada8573 Posts
On September 16 2011 10:36 Ryuu314 wrote: I've found that in BW people really like to overstate the value of APM, while in SC2, people tend to understate the value of APM. APM isn't a "meaningless number." It does mean something. A player who can achieve 400 APM will naturally be able to do more stuff than a player who can only achieve 100 APM. The question becomes whether or not the actions they're doing are meaningful. For the most part, lower level players will inflate their APM through spam, so big numbers aren't indicative of much. But, for higher level players, where few, if any, actions are wasted, APM does say quite a lot. If pro A can do 300 meaningful actions a minute, while pro B can only manage 200, all else constant, it's very likely pro A will be able to gain the upper hand simply by being able to make more shit happen at once. Eh I don't think it was overstated in BW. The minimum apm required just to macro efficiently was higher in BW than SC2, plus with the pathing and more control groups etc, you NEEDED more apm to do the same tasks you do in SC2. Not saying that apm isn't important in SC2 though. There was a good point made about strelock, by who I can't remember, but he gets by with 90 apm cause of his builds and strategic decisions, but when the game gets more fleshed out, and other people start doing the same thing, his apm will inevitably force him to fall behind. | ||
|
oDieN[Siege]
United States2905 Posts
| ||
|
Vindicare605
United States16121 Posts
No but seriously I wish there was a way to get stats this detailed from the ladder. | ||
|
Mangix
United States115 Posts
| ||
|
usethis2
2164 Posts
On September 16 2011 03:44 Exarl25 wrote: It will have less meaning, not more. Legitimately cycling through control groups will no longer count but clicking 50 times to move an SCV will still cause your to APM skyrocket. It will have more objective meaning. Clicking actually takes screen time during which you can't really do anything else. That's why you don't see anyone clicking 50 times to move an SCV. IMO, If someone does the same quality macro/micro with less control group spamming, then it's not illogical to value that more. For example: - player A does manage constant SCV production from 3 CCs and it takes her/him avg. 10 cycling through the CCs - player B does manage constant SCV production from 3 CCs and it takes her/him avg. 5 cycling through the CCs - player C does manage constant SCV production from 3 CCs and she/he doesn't need to cycle around because s/he is simply so well trained to the point that s/he never forgets it without even looking at CCs. What you get is 3 SCV production every 25 secs, and those are the "Actions" that are measured. How you get there will differ per individual, and I think it's not too absurd to assume, at pro level, if you can do the same/more with less, you're probably a better player. And if you're player C, you may be able to produce marines, marauders, tanks, etc. and better engage in the battle (i.e. more "actions") while player A is busy cycling through CCs to produce SCVs. It's definitely more objective way of measuring APM, and it will be much more useful/meaningful than current way of measuring. | ||
|
FlamingTurd
United States1059 Posts
| ||
|
MonkSEA
Australia1227 Posts
On September 16 2011 12:17 Mangix wrote: Shouldn't it be SK.Nada? :O It's based off the site sign-up, and Nada signed up with the ogs tag ^^; | ||
| ||
StarCraft 2 StarCraft: Brood War Britney Dota 2  23635Calm  5597Mini 907Horang2 828firebathero  426BeSt 342actioN 207EffOrt 181Soulkey 101Leta 71[ Show more ] Counter-Strike Heroes of the Storm Other Games gofns10995 Grubby2303 FrodaN1666 B2W.Neo905 Beastyqt379 ArmadaUGS146 crisheroes126 Fuzer  122KnowMe84 QueenE69 ToD68 Trikslyr61 |
|
Replay Cast
RSL Revival
Maru vs MaxPax
BSL
RSL Revival
Cure vs Rogue
uThermal 2v2 Circuit
BSL
Afreeca Starleague
Wardi Open
Replay Cast
Sparkling Tuna Cup
[ Show More ] |
|
|
EPT
