EPTI can understand that win rate in all the Zerg matchups because of 4 pool. But for PvT, if you do proxy gate, the first zealot arrives at Terran's base around 2:30-2:40, so the Terrans just leave the game immediately without a fight? But if that is the case, the win rate must be close to 100% for Protoss. So I guess the 30-40% Terran wins at that minute mark is due to Protoss leaving the game immediately after BBS. So no one bothers to pull probes and SCVs to micro?
Data analysis on 8 million games - Page 5
| Forum Index > BW General |
|
TMNT
3406 Posts
I can understand that win rate in all the Zerg matchups because of 4 pool. But for PvT, if you do proxy gate, the first zealot arrives at Terran's base around 2:30-2:40, so the Terrans just leave the game immediately without a fight? But if that is the case, the win rate must be close to 100% for Protoss. So I guess the 30-40% Terran wins at that minute mark is due to Protoss leaving the game immediately after BBS. So no one bothers to pull probes and SCVs to micro? | ||
|
sexyMIStrZZZ
Poland11 Posts
u dont have many tools to kill protoss as terran and u do have some tools as protoss, so occasionally protoss will win quick games more often than terran. Despite that Terran has overall >50% wr at 2000+MMR which is pretty crazy to think of if we include those lucky few min wins by Protoss. Basically you need to gamble and risk as much as possible early game in order to have a win from time to time in early games just to compensate for Terran being too strong later on. | ||
|
Dante08
Singapore4203 Posts
On October 16 2023 07:07 TMNT wrote: You know what, one thing that has been bugging me is the 60-70% win rate around 2-3 minutes in PvT at all mmr range. Because it makes little sense. I can understand that win rate in all the Zerg matchups because of 4 pool. But for PvT, if you do proxy gate, the first zealot arrives at Terran's base around 2:30-2:40, so the Terrans just leave the game immediately without a fight? But if that is the case, the win rate must be close to 100% for Protoss. So I guess the 30-40% Terran wins at that minute mark is due to Protoss leaving the game immediately after BBS. So no one bothers to pull probes and SCVs to micro? If you went rax expand and Protoss went zealots then yeah it makes sense to leave immediately lol. | ||
|
Comedy
487 Posts
On October 16 2023 06:48 sexyMIStrZZZ wrote: Arty if you're reading this you can always use any of the "valid" counterarguments below: a) Terrans train harder, Snow and Dewalt don't. Don't ask, they just don't. Just check the ladder, terrans have the most games, it's not like you can just delete an account and create a new one within 20 seconds. b) It's about the maps, the maps are the reason Terrans win c) Have you seen how many games do Protosses win at 5 minute mark? That's absolutely disgusting d) But when Jyj won ASL he played against Protoss only once and he lost that one game. Check-mate Protossists! e) Have you seen any Terran beat Protoss that is 400 MMR higher than them? f) Protoss has invisible units? That can't be fair, it's fucking bullshit g) I believed that P>T for 20 years, so there's nothing that can convince me, you're just wrong h) Someone messed up the data and its actually the other way around i) Did you see Snow beat Flash? Well, you just gotta train harder, if he could do that then you can do it too (completely forgetting Snow's performance in the last ASL, that didn't happen at all guys) Just remember - never talk about facts, at least not when they undoubtebly prove that Terran is overpowered at a high enough level (which here is already above 2000 MMR and probably even slightly below 2k if you check the stats closely). literally noone cares | ||
|
Kraekkling
768 Posts
On October 16 2023 07:07 TMNT wrote: You know what, one thing that has been bugging me is the 60-70% win rate around 2-3 minutes in PvT at all mmr range. Because it makes little sense. I can understand that win rate in all the Zerg matchups because of 4 pool. But for PvT, if you do proxy gate, the first zealot arrives at Terran's base around 2:30-2:40, so the Terrans just leave the game immediately without a fight? But if that is the case, the win rate must be close to 100% for Protoss. So I guess the 30-40% Terran wins at that minute mark is due to Protoss leaving the game immediately after BBS. So no one bothers to pull probes and SCVs to micro? You could try and go to repmastered.app and examine this type of games, maybe you'll be able to find some pattern? Other than that, we're looking at a small numer of games overall. There also is some baseline probabilty of players just randomly quitting games or disconnecting, which results in a number of games within this period with a win rate of 50% for both races. This effect could be quantified by looking at the win rate at the 0-1 minute period, where we expect basically no player interaction at all. This information is not available to us, because games shorter than 2 minute were filtered out in the pre-selection of data. The only useful estimate which I can provide here is that the overall rate of disconnects is somewhere around 2% of all games. The win rate in PvT around 2-3 minutes is 58% for all players combined. In the MMR-bracket plot all data points except the one with the biggest error bar are close to 60%. Some speculative thoughts: + Show Spoiler + Scenarios that might make a Terran quit around 2-3 minutes: - scouting probe enters before depot finishes - barracks delayed due to scouting probe - scouting probe kills scv - gas stolen - manner pylon - Terran scouts proxy gate(s) - zealot enters Scenarios that might make a Protoss quit around 2-3 minutes: - something involving 12nex? - ??? | ||
|
Rainalcar
Croatia565 Posts
Wrong, and you should care too. So now we know for sure: While some religiously claim for years T>Z>P>T We actually have T>>Z>>P<T - enshrined racial imbalance at pro level. And this is oh so visibly reflected in titles won in the last 20+ years. | ||
|
TMNT
3406 Posts
On October 17 2023 02:17 Rainalcar wrote: So now we know for sure: While some religiously claim for years T>Z>P>T We actually have T>>Z>>P<T - enshrined racial imbalance at pro level. And this is oh so visibly reflected in titles won in the last 20+ years. The T>Z>P>T is still true at most levels below 2000 mmr according to the analysis though. That's where 90% of the players are. Keep in mind that the T>>Z>>P<T you just mentioned is from analysis of ladder games above 2000 mmr. That's not "pro level" stats, although pros account for a part of high rank ladder surely. Here's the actual pro level stats from eloboard (since 2021): ZvP 8448 - 7454 (53.1%) PvT 8221 - 8518 (49.1%) TvZ 9114 - 7426 (55.1%) Sponbbang had the stats from 2016-2021 but sadly the site is dead now. But iirc it's similar to this. The stats probably still lies somewhere in a tl thread though. So as much as Snow and Best carrying the PvT matchup for Protoss, i.e. no matter how many Bo9s they won against Royal and Rush, they can't skew the stats from 16k games for Protoss to have a positive win rate against Terran. | ||
|
Cryoc
Germany915 Posts
| ||
|
Rainalcar
Croatia565 Posts
On October 17 2023 02:53 TMNT wrote: The T>Z>P>T is still true at most levels below 2000 mmr according to the analysis though. That's where 90% of the players are. Keep in mind that the T>>Z>>P<T you just mentioned is from analysis of ladder games above 2000 mmr. That's not "pro level" stats, although pros account for a part of high rank ladder surely. Here's the actual pro level stats from eloboard (since 2021): ZvP 8448 - 7454 (53.1%) PvT 8221 - 8518 (49.1%) TvZ 9114 - 7426 (55.1%) Sponbbang had the stats from 2016-2021 but sadly the site is dead now. But iirc it's similar to this. The stats probably still lies somewhere in a tl thread though. So as much as Snow and Best carrying the PvT matchup for Protoss, i.e. no matter how many Bo9s they won against Royal and Rush, they can't skew the stats from 16k games for Protoss to have a positive win rate against Terran. You are right, it's not all pros. Still, eloboard results are not that far off. P is still visibly weakest, while T has an inherent advantage. | ||
|
TT1
Canada10104 Posts
it gives a good idea of the general skill lvl among the higher MMRs but like i mentioned before, progamers play a different game, and even among them there's huge discrepancies at the highest lvl (Flash in all MUs, Snow PvT etc) | ||
|
sexyMIStrZZZ
Poland11 Posts
On October 17 2023 07:33 TT1 wrote: damn good stuff, ty for the MMR stats Kraekkling it gives a good idea of the general skill lvl among the higher MMRs but like i mentioned before, progamers play a different game, and even among them there's huge discrepancies at the highest lvl (Flash in all MUs, Snow PvT etc) They may play a different game, but the balance stats of their games suggest the exact same thing as 2000+ mmr results in this database, which gives even more evidence that Protoss is the weakest race in the game. | ||
|
TT1
Canada10104 Posts
On October 17 2023 07:43 sexyMIStrZZZ wrote: They may play a different game, but the balance stats of their games suggest the exact same thing as 2000+ mmr results in this database, which gives even more evidence that Protoss is the weakest race in the game. well the trends u see in winrates get amplified even more the higher up u go in skill, so ZvP and TvP keeps getting worse :D but there's still some outliers where pure skill flips everything on its head, like Snow PvT, it would be nice to see Flash at his best vs Snow personally if we were to talk about actual balance (where the players are playing the game at the highest lvl humanly possible) we would have to run games between the top 1-2 players of each race and check those stats, everything else is pretty meaningless (even midd/low lvl pro games, even A tier progamers to a certain extent) anything below that is just a whine fest of ppl just crying about their own experiences, the truth about BW is balance only really matters to the top .1% of players in this game, the skill cap of this game is so big that u can overcome it with skill, the higher up u go that cap becomes tighter and tighter tho | ||
|
sexyMIStrZZZ
Poland11 Posts
On October 17 2023 08:01 TT1 wrote: well the trends u see in winrates get amplified even more the higher up u go in skill, so ZvP and TvP keeps getting worse :D but there's still some outliers where pure skill flips everything on its head, like Snow PvT, it would be nice to see Flash at his best vs Snow personally if we were to talk about actual balance (where the players are playing the game at the highest lvl humanly possible) we would have to run games between the top 1-2 players of each race and check those stats, everything else is pretty meaningless (even midd/low lvl pro games, even A tier progamers to a certain extent) anything below that is just a whine fest of ppl just crying about their own experiences, the truth about BW is balance only really matters to the top .1% of players in this game, the skill cap of this game is so big that u can overcome it with skill, the higher up u go that cap becomes tighter and tighter tho There's some truth to that, but watching how the Z>P and T>P advantage is highly correlated with mmr one could assume that if we had the absolute best P ever play infinite games against the absolute best T ever, T would win vast majority of the time. Same with T>Z and Z>P. I think that the correlation of how T's wr goes up with mmr against P means more than the actual winrate itself. And the wr/mmr correlation is positive for T>Z Z>P and T>P. TBH I think Probes should have 41 hp so that Vultures wouldnt kill them with 2 shots which is absolutely ridiculous and then Reaver should be fixed so that it would always hit the target. Maybe increase damage to bio with storm so that PvZ would get fixed and then weaken the stimpack to fix TvZ. I also think that low mmr balance stats should be disregarded given that low mmr players could easily improve to become high mmr players while high mmr players do not have that opportunity and are much closer to their skill ceiling and any improvement of their game is minor and doesn't affect their winrates significantly. Low mmr stats should be used solely in order to show the trend of how the wr develops with increase of mmr. | ||
|
Dante08
Singapore4203 Posts
On October 17 2023 08:01 TT1 wrote: well the trends u see in winrates get amplified even more the higher up u go in skill, so ZvP and TvP keeps getting worse :D but there's still some outliers where pure skill flips everything on its head, like Snow PvT, it would be nice to see Flash at his best vs Snow personally if we were to talk about actual balance (where the players are playing the game at the highest lvl humanly possible) we would have to run games between the top 1-2 players of each race and check those stats, everything else is pretty meaningless (even midd/low lvl pro games, even A tier progamers to a certain extent) anything below that is just a whine fest of ppl just crying about their own experiences, the truth about BW is balance only really matters to the top .1% of players in this game, the skill cap of this game is so big that u can overcome it with skill, the higher up u go that cap becomes tighter and tighter tho Flash at his best vs Snow resulted in a 4-0 whoop. | ||
|
AttackZerg
United States7696 Posts
A strategy guide written from the perspective of "when you will die" and what from sounds funny. "At this point, if you are wrong, 14 zerglings with speed will destroy you - micro very hard before resigning" "... and if your scan reveals there to be five completed sunken colonies in front of your marine and medic army, press f2 and watch the two turrets in your mineral line fall to 8 mutalisks" "Congratulations you reached the late-game, you will die in 2 minutes and 15 seconds from guardian harass that will misplace your army allowing a defiler-push on your third or forth expansion, you played a good game" | ||
Excalibur_Z
United States12273 Posts
As for the actual data, I see a lot of conclusions being drawn about balance over time at "pro/semipro-level" MMR and frankly I find the sample size too low to say much of anything for that cursed yellow line. The variance as depicted by the I-bars is extremely high, particularly beyond minute 15. Even the 2200+ data I think has too much variance to really be useful in concluding much of anything. I think when you get into the 2000-2200 MMR range that's where you'll find pretty good sample sizes, but then by that point you're having a different discussion about a different cohort. Still, I'd say that's our best lead. I also see a lot of people trying to explain various parts of the graph (makes sense, people love interpreting data) but I think you can only maybe do that for openers. After say minute 10, the game state is so highly dependent upon the effectiveness of the openers of both players that virtually anything could happen. There are so many potential outcomes resulting from openers that it is impossible to make any meaningful attribution - all we know for sure is X number of games tend to end at Y time with Q race being the victor. Very cool post. | ||
|
sas.Sziky
Hungary341 Posts
with the ladder statistic just a joking, and makes no sense. Listen the guys who is 100x better as we us and know more we us. ( Flash,Mini, Jaedong and etc. ). | ||
|
Dakota_Fanning
Hungary2371 Posts
On October 17 2023 17:03 sas.Sziky wrote: First of all, congratulations, it must have been a lot of work and time. Second i want to say about the apm. i checked my id (StarCast_Sziky ) on the remastered ap and i saw the next : avg 233 eapm. what is not possible. or this is the a difffernt program what i know the last 10 years. + only 165 games? i am not remember correctly but it was much more than that everywhere. ( so I don't understand which games are included in the data and which are not) Third talk about the balance with the ladder statistic just a joking, and makes no sense. Listen the guys who is 100x better as we us and know more we us. ( Flash,Mini, Jaedong and etc. ). Obviously repmastered.app doesn't have all ladder replays, "just" about 8.5 million (that are longer than 1:50). You say it's not possible that you have 233 EAPM, why not? Is it too low or too high? Note this is EAPM (effective APM), not APM. Also you obviously had multiple accounts, many which we probably don't know about (not public). There are 35 players on repmastered.app that have "sziky" in its name, here's the combined player profile of all those 35: 35 combined players containing "sziky" These combined players have 4,494 games total, having 327 APM and 211 EAPM. Only 320 games are from Blizzard ladder. | ||
|
sas.Sziky
Hungary341 Posts
| ||
|
Dakota_Fanning
Hungary2371 Posts
On October 17 2023 20:26 sas.Sziky wrote: Too high, yes. i mean my max eapm around 240. avarage 200++. so the second data it's more real + 320 ladder games very few minimum X10? or something like that:D I just mentioned that on this because there could be problems here again from a statistic point of view, but its okay, and a good job!!! Unfortunately I can only get the latest 25 matches from ladder at any given time (although I'm collecting replays for some time now). Older matches are only available if someone uploads them. | ||
| ||
