|
United Kingdom20344 Posts
On January 02 2012 23:01 Markwerf wrote:
Don't assume too much from this data, a non-parametric regression like this one is very sensative to each months data which in some graphs might not be that much. For example PvT korea december is probably not played that much so if T did better there it doesn't mean that much..
Also you have to consider that only the top games are taken so unlike ladder stats for example there isn't corrected for rank here. For example the last patch might have buffed PvZ and PvT, which in turn can cause more P to qualify for the tournaments based on their PvZ for example. In tournaments the PvT results could actually go down then because PvZ was buffed more then PvT. Just an example but many, many effects like these can happen. A narrow observational study like this simply can't reveal things like matchup balance well. Also you can not really assume trends from these lines too much because they are very swingy by nature, the exact line also depends very much on the bin width you choose (month here).
Contrasting this with ladder statistics would be nice, ladder stats corrected for rank are the best way to look at balance still if that's your goal imo.
I know several masters players, and im pretty sure all of them would drop 16-0 in a best of 31 with MVP without issues. Balancing around ladder stats even for the highest league (GM is a broken system, so it would be masters) is wrong, even on korean server, and ESPECIALLY with stats that are "corrected" to change the results in ways that we do not know about.
It would be in blizzards best interests to lie, or change the results in a roundabout way, if something went really wrong.
|
On January 02 2012 22:05 Big J wrote:Show nested quote +On January 02 2012 22:02 SeaSwift wrote:On January 02 2012 21:59 StarcraftNerd1547 wrote:
I think the thing about TvP is that unless you have godly micro as T it's incredibly hard to win a late game engagement as terran. The reason why terran though is favored in these charts though is that in pro-level terrans has that micro which i was talking about and protoss can't do much about about it as TvP engagements is mostly about what terran does. If terran does a shit job and misses all EMP and gets his vikings kiled before defeating the toss army he loses, if he hits all EMP and kills every collosi before they can do significant damage he wins. This is why pros with good micro wins TvP and gold-level scrubs loses it.
I feel like the matchup needs a rework. Completely agree. All-ins are still really good as Terran, but Protoss is favoured slightly in passive macro games. Terran has a higher skill requirement, but also a higher skill ceiling, whereas Protoss has a lower skill requirement but a lower skill ceiling, too. I don't think the matchup will get any better until HotS. imo terran skill requirement is rather low compared to Ps... only judging from offracing both, but my terran is way better than my Protoss...
I played Terran for a year before i started playing random. I can easily say that i win more PvT's lategame then TvP's.
Of course it's individual but i think most people agree that terran has a higher skill requirement/skill ceiling in that phase of the game. Which is problematic for every Terran who isn't a pro, but also problematic for the pro Protoss who face the terran pro's.
|
So they should change the match up so you can miss your EMPs and lose your vikings, but still win the fight? I don't think they should change the game to make easier. Now if you want to talk about the PvT match up being to unstable for both sides, that's a discussion worth having. That match up is filled with one sided victories due to one missed forced field, emp, storm, stim or blink.
I didn't mean that they should change Bio, i want them to change so that there are other viable(maybe less fragile in lategame) strategies.
|
On January 02 2012 23:17 StarcraftNerd1547 wrote:Show nested quote +So they should change the match up so you can miss your EMPs and lose your vikings, but still win the fight? I don't think they should change the game to make easier. Now if you want to talk about the PvT match up being to unstable for both sides, that's a discussion worth having. That match up is filled with one sided victories due to one missed forced field, emp, storm, stim or blink. I didn't mean that they should change Bio, i want them to change so that there are other viable(maybe less fragile in lategame) strategies.
Actually I'd like it if they made lategame P more micro intensive instead... As it is now, micro in that matchup is solely the terrans requirement, and sadly it feels like HotS is aiming to make Terran require little micro as well with the battle hellions...
Edit: And the more I play against P, the more I dislike warp gates as a mechanic, but that one will probably never change...
|
On January 02 2012 23:00 SeaSwift wrote:Show nested quote +On January 02 2012 22:55 R!! wrote:On January 02 2012 22:47 Plansix wrote:On January 02 2012 21:59 StarcraftNerd1547 wrote:
I think the thing about TvP is that unless you have godly micro as T it's incredibly hard to win a late game engagement as terran. The reason why terran though is favored in these charts though is that in pro-level terrans has that micro which i was talking about and protoss can't do much about about it as TvP engagements is mostly about what terran does. If terran does a shit job and misses all EMP and gets his vikings kiled before defeating the toss army he loses, if he hits all EMP and kills every collosi before they can do significant damage he wins. This is why pros with good micro wins TvP and gold-level scrubs loses it.
I feel like the matchup needs a rework. So they should change the match up so you can miss your EMPs and lose your vikings, but still win the fight? I don't think they should change the game to make easier. Now if you want to talk about the PvT match up being to unstable for both sides, that's a discussion worth having. That match up is filled with one sided victories due to one missed forced field, emp, storm, stim or blink. The win ratio of 30 zealots with 3/3 and charge seems pretty solid to me, there's a reason why a PvT monster like Polt says that he feels like he has to either 14cc or 1/1/1 against random ladder scrubs in the new big maps. Source?
he wrote it on irc
03:24 @TSLPolt • toss is OP
03:24 @TSLPolt • but
03:24 @TSLPolt • If you use 1/1/1 you can win 
to be fair bomber says on his stream the only big maps are problem
|
Awesome! Great way to see trends in the game for each matchup!
|
On January 02 2012 22:47 Badfatpanda wrote:Show nested quote +On January 02 2012 20:15 ToastieNL wrote:
TvP is likely changing because
A- people learned to aim with smaller EMP
B- Skymehci s being developped and P needs to adapt
C- The core problem of the matchup (Colossi strong gateway weak) hasn't been attacked
D- Terran goes 2 engi bay faster than 2 forges for P
What? these are from professional matches. Skymech hasn't been seen being used other than here and there in a random bo3 since beta against protoss. People aren't learning to aim with EMP any more than 3 months ago, they've just been getting more ghosts. The core problem of the matchup is warpgate remax and zealots ability to tank damage for Protoss AOE units so well. And Terran getting ebays earlier than Protoss STILL end up with Protoss hitting 3/3 first because of chronoboost. Come on man.
very nice post, you are actually spot on mate.
the ability to crank a million zealots out in no time to tank damage for fucking high templars+colossi+archons is riddiculous.
|
On January 02 2012 23:00 ZenithM wrote:Show nested quote +On January 02 2012 22:47 Badfatpanda wrote:On January 02 2012 20:15 ToastieNL wrote:
TvP is likely changing because
A- people learned to aim with smaller EMP
B- Skymehci s being developped and P needs to adapt
C- The core problem of the matchup (Colossi strong gateway weak) hasn't been attacked
D- Terran goes 2 engi bay faster than 2 forges for P
What? these are from professional matches. Skymech hasn't been seen being used other than here and there in a random bo3 since beta against protoss. People aren't learning to aim with EMP any more than 3 months ago, they've just been getting more ghosts. The core problem of the matchup is warpgate remax and zealots ability to tank damage for Protoss AOE units so well. And Terran getting ebays earlier than Protoss STILL end up with Protoss hitting 3/3 first because of chronoboost. Come on man. So, if there are only problems coming from Protoss OPness, how do you explain the even winrate? Better, Terran is actually ahead. My only problem with Terran is the strength of their one base play because of minerals oversaturation with mules. This is really problematic, design wise. Why would they be allowed to have 25% more income on one base than Protoss or Zerg (you saturate with roughly 16-20 SCV and a MULE mines like 4 SCVs so it's literally 20-25% more mining)? It's gigantic and emphasizes all-in play, because of the momentum you gain over the over races. The rest I'm ready to acknowledge everything you want, from "chargelots a-move boohoohoo OP, I must stutter step and it's hard  " to "mech is so bad, why can't I go mech even though Protoss can't go stargate either :'(". But I'd like the one base terran imbalance (in my humble opinion) to be dealt with, first. Then, nerf Protoss to the ground if you want.
I DIDN'T SAY THERE WERE ONLY PROBLEMS. Please ffs just read what I was responding to. He wrote why TvP was changing from P favored to T favored, I said that his reasons didn't hold up. I don't believe in these win rates I think the majority of the system data is inflated due to the nature that TLPD accumulates it's data, the skill difference will be quite large. And in addition look at what Protoss has to offer in Korea among the top tier of players. What have they done recently and what have their Terran equivalents won?
It's not due to imbalance it's due to a difference in player skill that has never really been examined and cannot be quantified into little graphs that come up every month.
|
I wonder what TvP statistics would look like if you took out the all ins. Imho the matchup is balanced terribly at the moment. Protoss dominates most straight up games while Terran keeps crushing them with all ins.
|
On January 02 2012 23:45 Ermac wrote:
I wonder what TvP statistics would look like if you took out the all ins. Imho the matchup is balanced terribly at the moment. Protoss dominates most straight up games while Terran keeps crushing them with all ins.
+1
User was warned for this post
|
Ye from what I have seen on most pro streams, they often mention one race or another being OP but they do it jokingly. Another thing I noticed on Polt's stream whenever he beats a protoss people are always like well thats understandable because he is a beast where as whenever he loses its not because his opponent is good (which he undoubtedly is because his MMR matches him against Polt) instead it is that Protoss is OP. People need to identify their own racial biases.
If you think back to all the times when you won a game and your opponent called your race OP, you normally understand where they lost the game and it wasn't just that you had OP units. Its normally better macro, better micro, better composition, larger economy, better positioning or better upgrades Unless you look back over the replay and cant see where you could have won there is no point complaining about balance.
It would also be helpful if people didn't underrate the skill required by their opponents constantly. Just some examples: TvP (A-Move more! Zealots dont die! Late game so broken!) PvT (All T does is spam marauders! How is it fair that my T3 army loses to T1 !?! All you do is stim and its GG!) PvZ (srsrsrsrsrsrsrsrsr so skilled! Mutas so imba im glad blizz are fixing them in HotS!) ZvP (Herp derp a move colos = win! All protoss do is cheese!)
|
I'm glad to see that I'm not the only one with a lot of problems in zvp :D lets hope for some new breakthrough by the koreans this year though
|
love these charts, thanks !
|
On January 02 2012 23:39 Badfatpanda wrote:Show nested quote +On January 02 2012 23:00 ZenithM wrote:On January 02 2012 22:47 Badfatpanda wrote:On January 02 2012 20:15 ToastieNL wrote:
TvP is likely changing because
A- people learned to aim with smaller EMP
B- Skymehci s being developped and P needs to adapt
C- The core problem of the matchup (Colossi strong gateway weak) hasn't been attacked
D- Terran goes 2 engi bay faster than 2 forges for P
What? these are from professional matches. Skymech hasn't been seen being used other than here and there in a random bo3 since beta against protoss. People aren't learning to aim with EMP any more than 3 months ago, they've just been getting more ghosts. The core problem of the matchup is warpgate remax and zealots ability to tank damage for Protoss AOE units so well. And Terran getting ebays earlier than Protoss STILL end up with Protoss hitting 3/3 first because of chronoboost. Come on man. So, if there are only problems coming from Protoss OPness, how do you explain the even winrate? Better, Terran is actually ahead. My only problem with Terran is the strength of their one base play because of minerals oversaturation with mules. This is really problematic, design wise. Why would they be allowed to have 25% more income on one base than Protoss or Zerg (you saturate with roughly 16-20 SCV and a MULE mines like 4 SCVs so it's literally 20-25% more mining)? It's gigantic and emphasizes all-in play, because of the momentum you gain over the over races. The rest I'm ready to acknowledge everything you want, from "chargelots a-move boohoohoo OP, I must stutter step and it's hard " to "mech is so bad, why can't I go mech even though Protoss can't go stargate either :'(". But I'd like the one base terran imbalance (in my humble opinion) to be dealt with, first. Then, nerf Protoss to the ground if you want. I DIDN'T SAY THERE WERE ONLY PROBLEMS. Please ffs just read what I was responding to. He wrote why TvP was changing from P favored to T favored, I said that his reasons didn't hold up. I don't believe in these win rates I think the majority of the system data is inflated due to the nature that TLPD accumulates it's data, the skill difference will be quite large. And in addition look at what Protoss has to offer in Korea among the top tier of players. What have they done recently and what have their Terran equivalents won? It's not due to imbalance it's due to a difference in player skill that has never really been examined and cannot be quantified into little graphs that come up every month.
Oh look, it's the usual "Terran players are just better".
|
On January 02 2012 20:25 DaveVAH wrote:
^^ I Auto-all in vs P every single game after 1.4.2
Either way in korea its 52% T to 48% P
Same. 1/1/1 or marine/SCV all-in every time. My winrate stays above 50% in the matchup but it has nothing to do with playing real games.
On January 02 2012 20:35 Elwar wrote:
Terran started dominating late game TvP with ghost/viking and protoss reintroduced the turtle-to-mothership back to PvZ. Zergs are starting to deal with motherships much better, though I've yet to see a convincing answer to late game terran from any protoss.
This has to be a troll post. Absolutely has to be.
|
On January 03 2012 00:51 ZenithM wrote:Show nested quote +On January 02 2012 23:39 Badfatpanda wrote:On January 02 2012 23:00 ZenithM wrote:On January 02 2012 22:47 Badfatpanda wrote:On January 02 2012 20:15 ToastieNL wrote:
TvP is likely changing because
A- people learned to aim with smaller EMP
B- Skymehci s being developped and P needs to adapt
C- The core problem of the matchup (Colossi strong gateway weak) hasn't been attacked
D- Terran goes 2 engi bay faster than 2 forges for P
What? these are from professional matches. Skymech hasn't been seen being used other than here and there in a random bo3 since beta against protoss. People aren't learning to aim with EMP any more than 3 months ago, they've just been getting more ghosts. The core problem of the matchup is warpgate remax and zealots ability to tank damage for Protoss AOE units so well. And Terran getting ebays earlier than Protoss STILL end up with Protoss hitting 3/3 first because of chronoboost. Come on man. So, if there are only problems coming from Protoss OPness, how do you explain the even winrate? Better, Terran is actually ahead. My only problem with Terran is the strength of their one base play because of minerals oversaturation with mules. This is really problematic, design wise. Why would they be allowed to have 25% more income on one base than Protoss or Zerg (you saturate with roughly 16-20 SCV and a MULE mines like 4 SCVs so it's literally 20-25% more mining)? It's gigantic and emphasizes all-in play, because of the momentum you gain over the over races. The rest I'm ready to acknowledge everything you want, from "chargelots a-move boohoohoo OP, I must stutter step and it's hard " to "mech is so bad, why can't I go mech even though Protoss can't go stargate either :'(". But I'd like the one base terran imbalance (in my humble opinion) to be dealt with, first. Then, nerf Protoss to the ground if you want. I DIDN'T SAY THERE WERE ONLY PROBLEMS. Please ffs just read what I was responding to. He wrote why TvP was changing from P favored to T favored, I said that his reasons didn't hold up. I don't believe in these win rates I think the majority of the system data is inflated due to the nature that TLPD accumulates it's data, the skill difference will be quite large. And in addition look at what Protoss has to offer in Korea among the top tier of players. What have they done recently and what have their Terran equivalents won? It's not due to imbalance it's due to a difference in player skill that has never really been examined and cannot be quantified into little graphs that come up every month. Oh look, it's the usual "Terran players are just better".
every time i see that post i just laugh, "guys terran win rates are just high cause u know terrans are just the best players duah"
|
On January 03 2012 00:56 laharl23 wrote:Show nested quote +On January 03 2012 00:51 ZenithM wrote:On January 02 2012 23:39 Badfatpanda wrote:On January 02 2012 23:00 ZenithM wrote:On January 02 2012 22:47 Badfatpanda wrote:On January 02 2012 20:15 ToastieNL wrote:
TvP is likely changing because
A- people learned to aim with smaller EMP
B- Skymehci s being developped and P needs to adapt
C- The core problem of the matchup (Colossi strong gateway weak) hasn't been attacked
D- Terran goes 2 engi bay faster than 2 forges for P
What? these are from professional matches. Skymech hasn't been seen being used other than here and there in a random bo3 since beta against protoss. People aren't learning to aim with EMP any more than 3 months ago, they've just been getting more ghosts. The core problem of the matchup is warpgate remax and zealots ability to tank damage for Protoss AOE units so well. And Terran getting ebays earlier than Protoss STILL end up with Protoss hitting 3/3 first because of chronoboost. Come on man. So, if there are only problems coming from Protoss OPness, how do you explain the even winrate? Better, Terran is actually ahead. My only problem with Terran is the strength of their one base play because of minerals oversaturation with mules. This is really problematic, design wise. Why would they be allowed to have 25% more income on one base than Protoss or Zerg (you saturate with roughly 16-20 SCV and a MULE mines like 4 SCVs so it's literally 20-25% more mining)? It's gigantic and emphasizes all-in play, because of the momentum you gain over the over races. The rest I'm ready to acknowledge everything you want, from "chargelots a-move boohoohoo OP, I must stutter step and it's hard " to "mech is so bad, why can't I go mech even though Protoss can't go stargate either :'(". But I'd like the one base terran imbalance (in my humble opinion) to be dealt with, first. Then, nerf Protoss to the ground if you want. I DIDN'T SAY THERE WERE ONLY PROBLEMS. Please ffs just read what I was responding to. He wrote why TvP was changing from P favored to T favored, I said that his reasons didn't hold up. I don't believe in these win rates I think the majority of the system data is inflated due to the nature that TLPD accumulates it's data, the skill difference will be quite large. And in addition look at what Protoss has to offer in Korea among the top tier of players. What have they done recently and what have their Terran equivalents won? It's not due to imbalance it's due to a difference in player skill that has never really been examined and cannot be quantified into little graphs that come up every month. Oh look, it's the usual "Terran players are just better". every time i see that post i just laugh, "guys terran win rates are just high cause u know terrans are just the best players duah"
Oh, look its the usual i make mass stalker, my opponent made marauders. Me mad now.
|
Thanks for this - great work the past year.
|
Nice these are some of the most interesting stats in the community.
|
I would really like to know the win ratio of any TvP game that goes over 20-25 minutes.
|
|
|
|
|
|
EPT
