EPTWinrates since WCS S3 (Mostly since October) - Page 5
| Forum Index > SC2 General |
|
9-BiT
United States1089 Posts
| ||
|
Qwerty85
Croatia5536 Posts
| ||
|
larse
1611 Posts
On December 09 2013 16:33 Qwerty85 wrote: So out of those 2.4k games, what is the sample size before and what is after the patch? Take a look at the google doc file in the OP | ||
|
Micro_Jackson
Germany2002 Posts
The TvZ numbers and what i watched recently shows that the Widowmine and Tank nerf/buff was a big failure. The "aim" was not to nerv, it was to make the game more diverse by making the tank more of an standard element beyond early game zerg allin situations. This didn´t happen, most pro tvz are still bio mine, just it is harder to do now with the nerf. In my opinion it did the opposite because the tank buff limited the zerg options. Roach/Hydra is (some say it was never) not viable on a pro level. If someone trys it, ask Hyun what happens, Teaja made him look like a diamond player at HSC finals when he tried Roach Hydra. So is the game balanced according to this numbers? Yea kinda everything in range of +-5% is ok. Is the game fun in every matchup? I don´t think so. TvT is way more gimmicky, TvP is bad as ever with Protoss adapting to speedmedivacs and Terrans fighting with the same weapons since the WOL beta, without any surprise (no wonder no one scouts anymore as P in TvP). It might be balanced by the numbers but it isn´t fun playing and watching in my opinion. PvZ looks kinda like the end of WoL, which means win on 2 base or die trying. Not that it is a bad thing for P, it´s more that it is a mistake to take ever a 3th base. | ||
|
vthree
Hong Kong8039 Posts
On December 09 2013 17:06 USvBleakill wrote: TvZ is kinda interesting, TvP the numbers doesn´t really matter. TvP will always have some weird numbers because there are way more successful foreign Protoss than Terrans. I think we have to look at why that is. We can't just assume the better foreigners just play Protoss. Also, strangely, I think the tournaments kind of balance themselves out in terms of win rate. There are very few big name foreign terrans so foreign teams end up picking up more Korean terrans and fly them to all the foreign events. I don't have enough them to gather all the data. But I would be interested to know that if we took the top 10 players of each race (using alilugac or something) and see how many events they have been to since the patch. I have a feeling that the top 10 terrans go to a lot more tournaments compare to the top 10 protoss. Therefore, they end up pulling the win rate closer to 50%. | ||
|
shadymmj
1906 Posts
| ||
|
vthree
Hong Kong8039 Posts
On December 09 2013 17:06 USvBleakill wrote: TvZ is kinda interesting, TvP the numbers doesn´t really matter. TvP will always have some weird numbers because there are way more successful foreign Protoss than Terrans. I think we have to look at why that is. We can't just assume the better foreigners just play Protoss. Also, strangely, I think the tournaments kind of balance themselves out in terms of win rate. There are very few big name foreign terrans so foreign teams end up picking up more Korean terrans and fly them to all the foreign events. I don't have enough them to gather all the data. But I would be interested to know that if we took the top 10 players of each race (using alilugac or something) and see how many events they have been to since the patch. I have a feeling that the top 10 terrans go to a lot more tournaments compare to the top 10 protoss. Therefore, they end up pulling the win rate closer to 50%. And I don't think we can throw out ladder data outright. Sure, if it is just for a single individual player, it might be because their style just works better when their opponent doesn't know their style(ie, mech in TvP, roach hydra in ZvT, etc). But blink all-ins, proxy oracles are pretty much in every protoss' arsenal, so if a terran player can be prepared for it in offline tournaments, they would be prepared on it on ladder. After Soulkey's comment about lack of terrans on KR ladder, I went to check and there are 3 terrans in the top 30. And I disagree that PvZ is like WoL. A lot of foreign protoss seem to prefer the 2 base all-ins. But Korean protoss are doing fine with taking a 3rd and defending it (although they still throw stuff like 2 base all-ins or each zealots (San!) to change things up). If you watch the Rain series, he has little to no problems taking his third. Yes, sometimes it gets cancelled/delayed but is not do or die. | ||
|
Deleted User 137586
7859 Posts
On December 09 2013 16:33 Qwerty85 wrote: So out of those 2.4k games, what is the sample size before and what is after the patch? It's tiny. For example, post-patch TvP is 205 games. Just to put that into perspective, someone beating another 4-0 gives a (sic!) 2% change in the winrate. The variance is so big, you'd literally need to update the winrate at a game-to-game basis because the numbers keep changing.  Thank you for the effort, but both the list 99 and this post-patch list just doesn't have enough games to tell us much. | ||
|
SirPinky
United States525 Posts
On December 09 2013 07:14 Blargh wrote: Unfortunately, the game is still horribly designed with units such as the Oracle in the game. Do you really want a unit that flies at Nascar speeds and has the potential to kill a whole mineral line in about 10 seconds. It's just a unit in the game that really do not add anything truly "skill-requiring" to the game. Terran are stuck sitting with 5+ marines and a turret, and there is nothing else they can do about it. So, the win rates can be completely even, but the game is still sort of bad. Too many units lack skill-requiring mechanics, which ends up limiting the game's "reflectiveness" of skill (meaning, skill is not represented very well). @bo1b There is zero reason why Blizzard shouldn't be actively trying to improve their game other than the fact that they don't really make any money off of SC2 any more. And that's the worst developer excuse (at least for a company as well-off as Blizzard is........) ever. Instead of just trying to make it so all of the races win an equal amount, they should be trying to improve the mechanics and aspects of the game. They should change some skills around so that they do something more interesting than totally slow a huge area (Time Warp). There is really nothing the Terran can do once time-warped, as the spell is gigantic and has far too strong of an effect. The role of TW is also fulfilled by Force Fields, so there was not really much of a reason to add it in the first place. They have a TEST SERVER for this exact reason. Might as well have it constantly up with new changes. They could just throw totally random things up on the test server and it'd be better than how it is now. I completely agree. I can just see David Kim now in a managerial meeting "Guys! Look! Terran are only losing by 3% overall in TvP <Nevermind half the games requires an scv pull after mass turrets for oracle, blink, dt, vray, 4-gate, or 2 base immortal all-in>. What a punk. I read this guys bio and he is my age with BS undergrad who was seemingly good at SC1. How many koreans don't have that on their resume? What a poor choice as a game developer: His statements sound like every consultant I met that tries to marginalize a bad situation by not presenting the facts clearly (this, of course, is to preserve their employment after certain benchmarks are not met). I guess I have to give the guy credit for "spinning" the truth to preserve his job. Too bad for the game... | ||
|
Striker123
Canada14 Posts
The main problem that many terrans are addressing is the fact that due to design of Toss in general, ladder games can't be compared to LANs. Toss is a race that excels when one doesn't know how this person plays (a.k.a in ladder with so many barcodes in GM or randos that you haven't faced yet if you are in masters) due to the flexibility in early game. Now with the changes like MSC/Oracle Buff, it just adds on to so many things that toss can throw at you in terms of creativity (blink, dt, oracle, oracle into 3 gate, oracle into expo, greedy play with nexus cannon def, etc). This just makes TvP not FUN at all...it feels like when you win TvP that this toss tried to do something creative and failed horribly or you did not catch the timing attacks or the tech path toss went (in TvP no one can argue that Terran isn't the reactive race....if you don't choose the corresponding tech in forms of vikings vs col or ghosts vs ht, YOU ARE FUCKED. But if you guess it right, toss isn't always fucked and becomes a skill mu.) This doesn't happen as much in LANs since you know how that toss player plays and what kind of tricks they excel at in particular (which is why even MC changed his playstyle from all-in('timing') into more macro cuz you just can't win when everyone knows your style as a protoss unless you go macro style). HOWEVER, in ladder, it's just 5-6 different openers toss throws at you, and you trying to predict it. The games NEVER go back and forth like TvZ or TvT....which really reduces the quality of MU imo. This is why many terrans are begging blizzard to buff mech so Terran can also have a tech path they can choose and hopefully make toss react to Terran's opener depending on maps. | ||
|
TheBloodyDwarf
Finland7524 Posts
| ||
|
SirPinky
United States525 Posts
On December 09 2013 17:39 Striker123 wrote: I am surprised people are still talking about balance vs complaints on TvP....The problem lies in the design of TvP at the moment, not necessarily a balance issue. Obviously balance does come into play and many can see that by observing the ratio of toss vs terran players in Master and GM, and from that data, I would have to say that TvP is pretty balanced in pro scene atm (although many can argue that it's mainly due to like 2/3 korean terrans who are traditionally good since WoL days like Taeja). The main problem that many terrans are addressing is the fact that due to design of Toss in general, ladder games can't be compared to LANs. Toss is a race that excels when one doesn't know how this person plays (a.k.a in ladder with so many barcodes in GM or randos that you haven't faced yet if you are in masters) due to the flexibility in early game. Now with the changes like MSC/Oracle Buff, it just adds on to so many things that toss can throw at you in terms of creativity (blink, dt, oracle, oracle into 3 gate, oracle into expo, greedy play with nexus cannon def, etc). This just makes TvP not FUN at all...it feels like when you win TvP that this toss tried to do something creative and failed horribly or you did not catch the timing attacks or the tech path toss went (in TvP no one can argue that Terran isn't the reactive race....if you don't choose the corresponding tech in forms of vikings vs col or ghosts vs ht, YOU ARE FUCKED. But if you guess it right, toss isn't always fucked and becomes a skill mu.) This doesn't happen as much in LANs since you know how that toss player plays and what kind of tricks they excel at in particular (which is why even MC changed his playstyle from all-in('timing') into more macro cuz you just can't win when everyone knows your style as a protoss unless you go macro style). HOWEVER, in ladder, it's just 5-6 different openers toss throws at you, and you trying to predict it. The games NEVER go back and forth like TvZ or TvT....which really reduces the quality of MU imo. This is why many terrans are begging blizzard to buff mech so Terran can also have a tech path they can choose and hopefully make toss react to Terran's opener depending on maps. Well put. The cost-benefit for Protoss far exceeds that of Terran: There are too many potentially "all-in" capability with far less risk for Protoss. Double Oracle fail? No problem, just keep the terran in base sniping unprotected sides of mineral lines while expanding. Blink all-in scouted? No problem, just make 5-7 stalkers and feign an all-in while you expand and your opponent make 3 bunkers for no reason and stays on 1-base. Or there is the good ol' fake aggression and have terran over produce defensive while Protoss takes double forge and every tech path in the tree. God forbid you try to be aggressive and with 10+ rines (after fail Oracle) and make them right-click on their Nexus. I guess we are supposed to scout that with 1 reaper right? O' and don't forget, with the new 4-player map pool, we need to save 10 scans to make sure we dont miss what Protoss is proxying. So to your point: Not to make this about balance, but knowing your opponents playstyle is key. With Top Protoss many Terrans know their openers and study replays, versus random ladder opponents it is FAR different. I don't see these statistics reflective of true gameplay on masters level or greater on ladder. | ||
|
Micro_Jackson
Germany2002 Posts
On December 09 2013 17:24 vthree wrote: I think we have to look at why that is. We can't just assume the better foreigners just play Protoss. Also, strangely, I think the tournaments kind of balance themselves out in terms of win rate. There are very few big name foreign terrans so foreign teams end up picking up more Korean terrans and fly them to all the foreign events. I don't have enough them to gather all the data. But I would be interested to know that if we took the top 10 players of each race (using alilugac or something) and see how many events they have been to since the patch. I have a feeling that the top 10 terrans go to a lot more tournaments compare to the top 10 protoss. Therefore, they end up pulling the win rate closer to 50%. And I don't think we can throw out ladder data outright. Sure, if it is just for a single individual player, it might be because their style just works better when their opponent doesn't know their style(ie, mech in TvP, roach hydra in ZvT, etc). But blink all-ins, proxy oracles are pretty much in every protoss' arsenal, so if a terran player can be prepared for it in offline tournaments, they would be prepared on it on ladder. After Soulkey's comment about lack of terrans on KR ladder, I went to check and there are 3 terrans in the top 30. And I disagree that PvZ is like WoL. A lot of foreign protoss seem to prefer the 2 base all-ins. But Korean protoss are doing fine with taking a 3rd and defending it (although they still throw stuff like 2 base all-ins or each zealots (San!) to change things up). If you watch the Rain series, he has little to no problems taking his third. Yes, sometimes it gets cancelled/delayed but is not do or die. To be honest i think we can assume that, to a certain amount, the best (speak european) foreigners play protoss. It is a relative sure fact that Terran benefits from being "fast" which is more a korean thing, but you have to keep in mind from what these players come. Warcraft3 was more popular in the recent years in the foreign scene and Protoss fits these players better than Terran or Zergs. | ||
|
mihajovics
179 Posts
please understand that these numbers mean (almost) ABSOLUTELY NOTHING!!! the sample size is just too small. like the guy above said, just losing ONE series 2-0 changes the winrates by 1% !!! it's fun to know these numbers but we can absolutely not determine anything balance wise (except that there isn't an enormous imbalance, like 60%, which would be highly unlikely, but my guess is that even with a real 55-45 distribution people would be raging) | ||
|
AxionSteel
United States7754 Posts
| ||
|
Rollora
2450 Posts
On December 09 2013 18:20 AxionSteel wrote: TvZ stats will continue to decline, TvP will always probably be around the same. It always seems somewhat balanced at the top level, no matter how frustrating it is for ladder terrans! yeah, thats the pity of starcraft 2 being an e-sport title. It is balanced around the very best, and not for the rest of the 99.9 %. Which is fine, but ofc frustrating to some. When I check out my TvP stats in Sc2Gears (I play random but just for the interest), I see an almost 50% winrate, but with Toss having about a third of the APM, it feels like you have to put more effort into the game as T, than as P. So it will always be frustrating and I will always understand the complaining from the Terran side of things. | ||
|
Micro_Jackson
Germany2002 Posts
On December 09 2013 18:06 mihajovics wrote: guys, guys... please understand that these numbers mean (almost) ABSOLUTELY NOTHING!!! the sample size is just too small. like the guy above said, just losing ONE series 2-0 changes the winrates by 1% !!! it's fun to know these numbers but we can absolutely not determine anything balance wise (except that there isn't an enormous imbalance, like 60%, which would be highly unlikely, but my guess is that even with a real 55-45 distribution people would be raging) I think too that we shouldn't look too much at these numbers as long as they are not going out of the +-5% window, but saying they mean nothing is wrong in my opinion. Stats like this alone doesn´t tell to much without some background knowledge like every statistic or study. On the other hand is the "to small sample size" wrong. Almost every scientific study works with smaller sample size that is just calculated to a trend. If we would cover all games played in the last month you could argue that "99% of all players are to bad to take them into account" and you wouldn´t be wrong. But thats all we have to discuss and the past statistics show that they are pretty consistent data. | ||
|
TheBloodyDwarf
Finland7524 Posts
On December 09 2013 18:06 mihajovics wrote: guys, guys... please understand that these numbers mean (almost) ABSOLUTELY NOTHING!!! the sample size is just too small. like the guy above said, just losing ONE series 2-0 changes the winrates by 1% !!! it's fun to know these numbers but we can absolutely not determine anything balance wise (except that there isn't an enormous imbalance, like 60%, which would be highly unlikely, but my guess is that even with a real 55-45 distribution people would be raging) Luci is 0-12 w/l in last two tournament Edit: sorry, lucifron has 14 losses and 0 wins in last 2 tournaments. | ||
|
Grovbolle
Denmark3813 Posts
On December 09 2013 18:53 TheBloodyDwarf wrote: Luci is 0-12 w/l in last two tournament Edit: sorry, lucifron has 14 losses and 0 wins in last 2 tournaments. Results for   LucifroN after 2013-11-01.Games: 15.79% (3-16) Matches: 27.27% (3-8) Current Form: W W L L L L L L L L Recent Matches: LucifroN 0 – 3  jjakji LucifroN 0 – 3   elfi LucifroN 0 – 2  NaNiwa LucifroN 0 – 2 Patience LucifroN 0 – 2   TLO LucifroN 0 – 2 HyuN LucifroN 0 – 1 Alicia LucifroN 0 – 1 Sting LucifroN 1 – 0 Daisy LucifroN 1 – 0 StarNaNFilters: + Show Spoiler + Opponent Race: all Stats by Aligulac. Link. | ||
|
vthree
Hong Kong8039 Posts
On December 09 2013 17:39 Striker123 wrote: This doesn't happen as much in LANs since you know how that toss player plays and what kind of tricks they excel at in particular (which is why even MC changed his playstyle from all-in('timing') into more macro cuz you just can't win when everyone knows your style as a protoss unless you go macro style). HOWEVER, in ladder, it's just 5-6 different openers toss throws at you, and you trying to predict it. The games NEVER go back and forth like TvZ or TvT....which really reduces the quality of MU imo. This is why many terrans are begging blizzard to buff mech so Terran can also have a tech path they can choose and hopefully make toss react to Terran's opener depending on maps. I am not sure there is that much difference between LAN and ladder now. Maybe when you first have builds come out like 1-1-1, soul train, Stephano roach max, etc, only some player could execute them well and so in LAN, you can disregard the possibility. But as time goes on, these builds that are shown to be powerful and reliable and they are added to the arsenal of every top tier player. The 2 biggest all-in/aggressive builds in PvT is the proxy oracle and blink all-in. Is there any top protoss is the world where if a top terran were to play them, they could say 'well, this protoss won't do proxy oracle, or won't do blink all-in?' Maybe someone like Rain. But for pretty much everyone else, the terran has to know it is possible they will pull out these builds so I don't think it is that different from ladder. It is like 1-1-1 pre nerf and 11-11 in WoL, every top terran can and WILL throw it out there. Granted, some players will do it more than others. But I think it is common enough that pro terrans have take it into consideration whether in LAN/ladder. | ||
| ||
