Race Performance - Page 2
Forum Index > BW General |
Telcontar
United Kingdom16710 Posts
| ||
matjlav
Germany2435 Posts
PvP is basically the same thing as ZvZ, but longer and with more luck-based tech builds.... Otherwise, not too surprising, but an interesting way to present the data. TvZ really is the most imba matchup by far, haha. I feel like the Terran always has to fuck up somehow to lose a TvZ... | ||
Scarecrow
Korea (South)9172 Posts
On June 17 2011 03:16 Release wrote: It's unfair to just say P>T>Z>P... Bisu(P) >Z Flash(T) > Z,P Jaedong(Z) > T Why do you think these 3 are so popular? Everyone likes heroes who buck the trend | ||
FlaShFTW
United States9936 Posts
| ||
Huragius
Lithuania1506 Posts
| ||
Severedevil
United States4821 Posts
On June 17 2011 04:11 Huragius wrote: T>P>Z<T- remove Flash, and this will be incredibly wrong. Yes... remove Flash but not Jaedong or Bisu or Stork, that's a fair comparison. | ||
Xiphos
Canada7507 Posts
On June 17 2011 03:22 matjlav wrote: PvP has always been the true coinflip matchup. I don't know why ZvZ got that reputation. PvP is basically the same thing as ZvZ, but longer and with more luck-based tech builds.... Otherwise, not too surprising, but an interesting way to present the data. TvZ really is the most imba matchup by far, haha. I feel like the Terran always has to fuck up somehow to lose a TvZ... This man speaks the truth. In ZvZ, you can go for early expand or 9 pool speed rush but at the end of the day, you will STILL go for that Muta/Ling combo and comebacks are more likely. With PvP, pffff, there are much more diversity and plus it all depends on that one battle (Huh, kinda like SC2 I guess lol). Terran vs Terran is just who have better multitask. | ||
Huragius
Lithuania1506 Posts
On June 17 2011 04:15 Severedevil wrote: Yes... remove Flash but not Jaedong or Bisu or Stork, that's a fair comparison. I don't think that by removing Jaedong/Bisu or Stork would change the results so drastically like by removing Flash. | ||
Chef
10810 Posts
What if you took the top 30 players of each race, and compared #1 vs #1 vs #1, and #2 vs #2 vs #2? That would make a lot more sense to me. Or maybe that's what you did... I'm not really sure. If that's what you did it's truly surprising that there are no anomalies outside the #1 spot. | ||
hitthat
Poland2232 Posts
| ||
]343[
United States10328 Posts
On June 18 2011 00:34 Chef wrote: I feel like you are manipulating the data a little too much to come to the conclusion you did. What if you took the top 30 players of each race, and compared #1 vs #1 vs #1, and #2 vs #2 vs #2? That would make a lot more sense to me. Or maybe that's what you did... I'm not really sure. If that's what you did it's truly surprising that there are no anomalies outside the #1 spot. Yeah, I think that's what he did (compared #1 ZvT to #1 TvZ, etc.) Doesn't it show in the graph? | ||
djbhINDI
United States372 Posts
On June 17 2011 03:16 Release wrote: It's unfair to just say P>T>Z>P... Bisu(P) >Z Flash(T) > Z,P Jaedong(Z) > T Oh no, you're talking about a general trend over dozens of pro gameers when THREE buck the trend? Oh, noes, you are unfair. | ||
xarthaz
1704 Posts
On June 18 2011 00:53 ]343[ wrote: Yes, that is correct. Or to be percise, the rankings are normalized, ie rank 30 terran compared to rank 34 protoss because there were more protoss players.Yeah, I think that's what he did (compared #1 ZvT to #1 TvZ, etc.) Doesn't it show in the graph? | ||
ImbaTosS
United Kingdom1666 Posts
On June 17 2011 03:22 matjlav wrote: PvP has always been the true coinflip matchup. I don't know why ZvZ got that reputation. PvP is basically the same thing as ZvZ, but longer and with more luck-based tech builds.... Otherwise, not too surprising, but an interesting way to present the data. TvZ really is the most imba matchup by far, haha. I feel like the Terran always has to fuck up somehow to lose a TvZ... If your statements on PvP are truly correct, explain why Horang2 is pushing above a 70% winrate against P. He can win almost every game. That can only be huge skill in the matchup, allowing him to be elevated above everyone else. | ||
Caos2
United States1728 Posts
| ||
xarthaz
1704 Posts
On June 21 2011 06:03 Caos2 wrote: Nice graphs, but can we get a raw dump of the data? I'd love to poke around. Its just TLPD advanced player chart copy-pasted into excel(+ excel auto sort function), i did basically nothing to it | ||
Crunchums
United States11143 Posts
On June 24 2011 00:36 xarthaz wrote: The race with the highest mirror matchup ELO is in aggregate, the best performing race in the game. This is because comparitively, that race has the best performance against iself, compared to the performance of other races against itself(ie, terrans high vT elo compared to zerg's vZ elo must mean that protoss and zerg have lower vT elo in aggregate, than terran and protoss do vZ, thus suggesting the aggregate superiority of T>Z). I do not understand - are you saying that if TvT skill is higher than ZvZ skill in aggregate that means that Terrans are more skilled than Zergs, or are you saying something else? | ||
xarthaz
1704 Posts
| ||
figq
12519 Posts
| ||
djbhINDI
United States372 Posts
On June 24 2011 00:45 Crunchums wrote: I do not understand - are you saying that if TvT skill is higher than ZvZ skill in aggregate that means that Terrans are more skilled than Zergs, or are you saying something else? Well, if the curve is smooth all the way across, then that matchup is less "skill based" - if the worst Terran's TVT ELO is really low, that means he loses a lot and if the best Terran's TvT ELO is really high, that means he wins a lot. If the best wins a lot and the worst loses a lot, then you can say this matchup is skill-based. On the other hand, the ZvZ and PvP ELO's are a bit smoother than TvT - the best doesn't win as much and the worst doesn't lose as much. This is why these two are considered more "coin flip" - skill level doesn't affect games won/lost as much as in TvT. The guy you are responding to either doesn't know what he's saying or doesn't know how to say it - his post is really unclear. This is kinda what I think he might have been wanting to say. | ||
| ||