|
On July 26 2014 07:34 Hider wrote:Show nested quote +On July 26 2014 07:26 pure.Wasted wrote:On July 26 2014 06:20 Hider wrote:On July 26 2014 06:16 Lunareste wrote:On July 26 2014 06:14 Cheren wrote: IMO the stalker was not designed with Kespa level blink micro in mind and is currently too strong because of that. Blizzard should think about having it do 10 flat damage since Immortals are already good against armored units and come out shortly after stalkers anyway. No, we shouldn't nerf the one mechanic that Protoss has that rewards the user for better micro. Instead, we should re-design more units to do the same for all 3 races. That's true, but only when you realistically can add countermicro to the unit. So for instance if one player can micro unit X really well and become a lot more cost-effective. Then the enemy player should be able to remicro agianst the micro of unit X in order to neutralize the micro as well (assuming he is good). The issue with blink-stalkers is that the micro is one-sided. If the protoss has really good micro he becomes a lot more cost-effective, however it's almost impossible for the enemy to have any countermicro plays against it (at least in the earlier midgame). Bad argument. Same is true of Stim, should we remove Stim because it makes bio too strong when played by Kespa players? If Stalkers are too strong for a unit that maneuverable, they could be made weaker. (I'd like to see a version with higher DPS and lower shields) On top of that, the way Stalkers are used is very different from the way micro-intensive Terran units are used. Stalkers don't do a lot of "high risk high reward" type pokes, like you don't see four groups of Stalkers attacking key positions all over the map at the same time. Protoss midgame doesn't work like that (thx Colossus). With a different midgame, we could absolutely see Stalkers used differently, and have a higher degree of risk attached to their use. No. Everything that occurs after you activate stim creates a ton of new microinteractions. Like bio splitting vs Banelings/storm becomes practical after you have activated stim. When blink is being used, no remicro becomes possible. That's why in my opinion, blink should be more about multitask/harass-oriented situations than about something that has the potential to outright kill the enemy.
There are tons of counter-micro possibilities to Blink. Making sure your units continue attacking the damaged Stalker so it can't regenerate shields, making sure your units DON'T continue attacking the damaged Stalker so they don't get kited, those are just the obvious two. Blink also allows Stalkers to have a "new microinteraction" with WMs by dodging their shots.
You're confusing us not seeing that counter-micro because the current meta doesn't encourage it, with the spell fundamentally not having any. The ability has tons of potential. Turtle 20 minutes into Colossus deathblob simply shuts it the fuck down.
|
Making sure your units continue attacking the damaged Stalker so it can't regenerate shields, making sure your units DON'T continue attacking the damaged Stalker so they don't get kited, those are just the obvious two
Sry, but I think only someone who hasn't played against blink stalkers would recommend target-firing against them.
The issue with target firing is that if the enemy blinks away his stalkers, then your units will follow them blindly for a short while, which bascailly creates an equilibrium where the optimal thing for the enemy to do is to a-move.
This differs from how target firing works vs all other units since they don't instantly move from one place to another. But instead, when they move, they are still vulnerable to target-firing for a short while. This means that the penalty for target-firing against an enemy that "remicro's" isn't as high which actual creates an interaction where targetfiring and repositioning units against target-firing is optimal (unlike the blink-interaction).
Blink also allows Stalkers to have a "new microinteraction" with WMs by dodging their shots.
True, but noone here is talking about the counter-micro possilbiites for the protoss-player. All I am saying that during battles, there is no actual countermicro the enemy can do against it.
|
On July 26 2014 03:54 Faust852 wrote:And for people saying Z aren't way stronger than T : Show nested quote +On July 24 2014 15:32 Ghanburighan wrote:While we're looking at winrates, here's another Aligulac list: Just looking at winrates, PvT is rather even, and so is PvZ but TvZ has gone down to the dumps again. On the other hand, the population numbers are the worst ever for Terran. It looks like T has a constant of around 100 games every period, but with the added number of games (last period has 1799 games, this one 3866), only Z and P seem to have added more mirrors. So there are 4.8x as many ZvZ as TvT, and 3.8x as many PvP as TvT. This also means that P has once again caught up with Z populations, last period it was 1.3 ZvZ for every 1 PvP, now it's 1.2. On July 10 2014 20:15 Ghanburighan wrote:Here's the latest Aligulac list (114) with pretty new formatting. With regard to P, nothing seems to have changed. Just like the first half of June, P>T by a slight margin, P and Z are roughly even, and there are roughly the same number of PvP MU's in tournaments. Z did worse in this period, while it was at >55% against T last time, it's now even in winrates. More importantly, looking at populations, while there were 5x more ZvZ than TvT, and 2x more ZvZ than PvP, then now there are only roughly 3x more ZvZ than TvT, and a just over a fourth more ZvZ than PvPs. This suggests that Z is doing worse, and it's mainly doing worse against T (note that worse doesn't imply that they're doing bad, this is a comparison with the previous period). Looking more closely at the population numbers, there appear to have been fewer games, the total for 114 is 1835 and for 113 it was 2379. So for the previous 113 list Z MUs made up 72% of all MUs. P MUs made up 55% (note that the overlap is due to the fact that P plays Z...). T MUs made up 36% of all MUs. In this list, 114, Z MUs made up 65% of all MUs. P MUs made up 57%. T MUs made up 42% of all MUs. So Z is down 7%, P is up 2% and T is up 6%. (with rounding) The previous lists can be found below. On June 29 2014 05:42 Ghanburighan wrote:Sorry for the delay, here's Aligulac 113.. The previous list(s) can be found at the end of this post. Looking at the winrates, P has extended its advantage over T, P has also gained some ground back against Z, yet TvZ has strongly turned in Z favour once gain (it's as bad as it was before the hellbat patch in April). Population numbers are also worse. Previously there were 4x more ZvZ games than TvT games, now there are more than 5x. PvP's have not changed in number, so it's mostly just less terrans and more zergs getting further that's creating the problem. All in all, balance-wise this was a very depressing period. On June 12 2014 15:32 Ghanburighan wrote:Time to post the latest Aligulac list. The previous list can be found at the end of this post. Regarding winrates, PvT has fluctuated back from T having a slight advantage to P having a minuscule advantage. In PvZ, P has also improved although it hasn't caught up with Z. On the other hand, T has improved in the TvZ MU (110 had 45%, 111 had 47%) and its even now. In terms of populations measured in numbers of mirror MUs, there's virtually no change compared to the last list, the proportions are very close. This means that there is no repopulation of terrans according to these numbers and there are 4 times fewer TvTs than ZvZs. As T MUs have even winrates, there cannot really be a repopulation with these numbers. Furthermore, a word of caution, I'd say that this was one of the best periods for Terran in a long while, Taeja won Hsc 9 (where Z had a comparatively weaker list of players), Maru is tearing up Code S, and Innovation is kicking as in teamleagues and the Dragon cup. I don't think they contributed overly much to the final winrates (their games are still a small fraction of all the games), but taken together they did contribute significantly. If they don't keep their winning ways going, winrates can plunge below 50% again. And, their wins aren't helping repopulate in any way. On May 29 2014 02:45 Ghanburighan wrote:Uploading the latest Aligulac list. Unfortunately there was a TvZ patch in the middle of the period, so those numbers could be anything now. But it looks like P is doing worse against Z in terms of winrate. But the population ratios haven't changed compared to the last list, though. It's still roughly 1/4 TvT, 2/4 PvP and 1/1 ZvZ. Hold your horses. 45% for 1-2weeks isnt that bad, given the standards Terrans have taught us in 2013. ;-)
More seriously, Im not against the usage of aligulac (feed me as many stats as possible), yet I've become very careful since that post that revealed that a third (?) of its database are Australian tournaments. It's nice to know that they have a very active community, but it's not the same as looking at Europa or Korea or america alone. Since I have been watching mainly Korean stats, I know that they have very balanced winrates in TvZ. I think in EU it is more Z favored (not just WCs but also e.g. DH). Not sure about America.
|
So I hear Commander Widow Mines are back. Has the winrates skewed in Terran favour? Protoss and Zergs crying yet?
User was warned for this post
|
|
On July 26 2014 07:49 Hider wrote:Show nested quote +Making sure your units continue attacking the damaged Stalker so it can't regenerate shields, making sure your units DON'T continue attacking the damaged Stalker so they don't get kited, those are just the obvious two Sry, but I think only someone who hasn't played against blink stalkers would recommend target-firing against them. The issue with target firing is that if the enemy blinks away his stalkers, then your units will follow them blindly for a short while, which bascailly creates an equilibrium where the optimal thing for the enemy to do is to a-move. This differs from how target firing works vs all other units since they don't instantly move from one place to another. But instead, when they move, they are still vulnerable to target-firing for a short while. This means that the penalty for target-firing against an enemy that "remicro's" isn't as high which actual creates an interaction where targetfiring and repositioning units against target-firing is optimal (unlike the blink-interaction).
So there's a danger of being kited? So... exactly what I said in the second half of the sentence you responded to?
You could counter his retreating Blink by having a couple of flanking Marauders come in from the side to take out the severely wounded Stalkers that try to get away. That's counter-micro. Or if he's not trying to retreat, if he's just trying to maximize damage output by pulling wounded Stalkers out and then ordering them immediately back into the fight, nothing prevents you from selecting a single Marauder in the fight and clicking on that single Stalker. He'll have to have very fast reaction time to prevent the kill.
Do these things happen in the game? Obviously not. Is that a problem with the ability? Not as far as I can see.
Show nested quote +Blink also allows Stalkers to have a "new microinteraction" with WMs by dodging their shots. True, but noone here is talking about the counter-micro possilbiites for the protoss-player. All I am saying that during battles, there is no actual countermicro the enemy can do against it.
When I said that Stim is just like Blink, you responded "Stim creates micro possibilities for Marines to split against Banelings!" If Marine micro possibilities are fair game in defense of Stim, Stalker micro is fair game in defense of Blink.
And if it's not, we're back to Stim not having any counter-micro, either.
|
You could counter his retreating Blink by having a couple of flanking Marauders come in from the side to take out the severely wounded Stalkers that try to get away. That's counter-micro.
Do these things happen in the game? Obviously not. Is that a problem with the ability? Not as far as I can see.
No it's not counter-micro. Why? Because you have to set this up preemptively. Flanks and anything that happens prior to the battle isn't countermicro. E.g. setting up your Chargelots in front of your army before engaging a bio-ball isn't counter-micro either.
Counter-micro instead occurs when player 1 does a specific action during a battle with one of his units and Player 2 can respond to that efficiently by micro'ing his own units.
Further, the question you must answer here isn't whether it's hyptheotically possible or not to add countermicro to blink, but wheter it's realistic to add in LOTV for instance. I have actually worked alot on trying to make this possible, and my conclusion is that you can't do it. Blink simply provides a way too instantbased form of "micro" that breakes the "rules" of normal RTS-based micro.
You would need some absolutely insane other form of mobility to enemy units in order to add realistic countermicro here.
nothing prevents you from selecting a single Marauder in the fight and clicking on that single Stalker. He'll have to have very fast reaction time to prevent the kill.
This is almost never realistic in an actual game. You said your self you don't play the game, and I would recommend you to be careful about making these types of statements.
When I said that Stim is just like Blink, you responded "Stim creates micro possibilities for Marines to split against Banelings!" If Marine micro possibilities are fair game in defense of Stim, Stalker micro is fair game in defense of Blink.
No becasue stim in it self doens't prevent any type of micro. Blink clearly does as it's often efficient to target-fire Stalkers without blink, but inefficient to target them when they have blink.
|
On July 26 2014 08:18 Hider wrote:Show nested quote +You could counter his retreating Blink by having a couple of flanking Marauders come in from the side to take out the severely wounded Stalkers that try to get away. That's counter-micro. Or if he's not trying to retreat, if he's just trying to maximize damage output by pulling wounded Stalkers out and then ordering them immediately back into the fight, nothing prevents you from selecting a single Marauder in the fight and clicking on that single Stalker. He'll have to have very fast reaction time to prevent the kill.
Do these things happen in the game? Obviously not. Is that a problem with the ability? Not as far as I can see. No it's not counter-micro. Why? Because you have to set this up preemptively. Flanks and anything that happens prior to the battle isn't countermicro. E.g. setting up your Chargelots in front of your army before engaging a bio-ball isn't counter-micro either. Counter-micro instead occurs when player 1 does a specific action during a battle with one of his units and Player 2 can respond to that efficiently by micro'ing his own units.
Pre-positioning isn't counter-micro, but what if there's no way to pre-position perfectly because the Protoss could Blink away in any number of directions? Then you have to use your flanking bio to intercept the Stalker, that's counter-micro.
I know my hypotheticals are very hypothetical, but you're the one making the bold claim that the ability is fundamentally flawed. All I have to provide is reasonable doubt.
Further, the question you must answer here isn't whether it's hyptheotically possible or not to add countermicro to blink, but wheter it's realistic to add in LOTV for instance. I have actually worked alot on trying to this, but I don't see anyway. Blink simply provides a way too instantbased form of "micro" that breakes the "rules" of normal RTS-based micro.
OK, this is a separate question. It's very simple. Does Blizzard overhaul Protoss with LOTV and turn them into a skill-demanding race? If they do, then anything's possible. If they don't, then it doesn't really matter what we do or don't do with Blink, because for me the game's pretty much fucked anyway. And I expect there are other viewers who'll jump ship, too.
I only tolerate Protoss now because there is hope for significant change. If there's no hope of change, I'll keep watching if Protoss goes back to only having 6-8 players per RO32. Any more than that and I just can't take the competition seriously.
You would need some absolutely insane other form of mobility to enemy units in order to add realistic countermicro here.
Stim+Medivac, creep+fast lings, counter-Blink. I feel that the tools are theoretically there, but I don't see why we couldn't add more.
|
I know my hypotheticals are very hypothetical, but you're the one making the bold claim that the ability is fundamentally flawed. All I have to provide is reasonable doubt
I don't think I am the only one saying that there isn't really much micro you can do vs blink. I think you will get that from almost every T/Z out there as well.
Notice, however, I didn't say fundamentally flawed. Also, I do acknowledge that many protoss players like this units as it is very micro-intensive. So it's not flawed in every way.
Instead, I proposed a solution to make blink-stalkers more centered around the area where the enemy can respond to it through skills. That's mainly when blink-stalkers are used as pressure/harass-oriented/mutltiask-based unit. In this case, a good terran/zerg can multitask better to defend against the blink stalkers.
Then I suggested to make Stalkers less important during battles/timing attacks and instead increase the Immortals role (plus tweak the Immortal slightly to make it function well as a "core-unit"). The difference betwen the blink-stalker and the Immortal is that all we need to do make the Immortal a unit that can be micro'ed and micro'ed against is through stat-tweaking. Blink-stalkers on the other hand, it's almost impossible to create "practical" counter-micro by just stats-tweaking. You would basically need a totally new game for that to happen.
I guess that's kinda what the Onegoal-mod attempted to do by putting Immortal at WG-tech and Sentry at Robo-tech and then making the Stalker more of a harass-unit. However, I think that's just an unncesary overcomplication as it likely will results in multiple unintended consequences. Rather, I think a more simple solution is to make Robotics Tech cheaper in order to open up for multiple Robotics being used in the midgame. This way, it becomes much easier for the toss to get more Immortals into his composition which (too some extent) can free up the Stalker-role.
Stim+Medivac, creep+fast lings, counter-Blink. I feel that the tools are theoretically there, but I don't see why we couldn't add more.
Not-counter micro. Just very mobile units without any specific interaction related to blink.
|
On July 26 2014 08:48 Hider wrote:Show nested quote +I know my hypotheticals are very hypothetical, but you're the one making the bold claim that the ability is fundamentally flawed. All I have to provide is reasonable doubt I don't think I am the only one saying that there isn't really much micro you can do vs blink. I think you will get that from almost every T/Z out there as well. Notice, however, I didn't say fundamentally flawed. Also, I do acknowledge that many protoss players like this units as it is very micro-intensive. So it's not flawed in every way.
I see what you're saying, but the truth is Banelings are the only unit that is used differently against Stimmed Marines compared to regular Marines. In every other situation, including all TvPs, Stim just makes Marines better in every way while providing no concrete way for the opponent to counter-micro.
I don't think counter micro as such is absolutely necessary. It's an awesome unexpected bonus (no one saw Marine/Bane interaction coming) but not a requirement. What's important is that micro isn't negated. Stuff like FG and TW are huge offenders.
Instead, I proposed a solution to make blink-stalkers more centered around the area where the enemy can respond to it through skills. That's mainly when blink-stalkers are used as pressure/harass-oriented/mutltiask-based unit. In this case, a good terran/zerg can multitask better to defend against the blink stalkers.
Then I suggested to make Stalkers less important during battles/timing attacks and instead increase the Immortals role (plus tweak the Immortal slightly to make it function well as a "core-unit"). The difference betwen the blink-stalker and the Immortal is that all we need to do make the Immortal a unit that can be micro'ed and micro'ed against is through stat-tweaking. Blink-stalkers on the other hand, it's almost impossible to create "practical" counter-micro by just stats-tweaking. You would basically need a totally new game for that to happen.
I guess that's kinda what the Onegoal-mod attempted to do by putting Immortal at WG-tech and Sentry at Robo-tech and then making the Stalker more of a harass-unit. However, I think that's just an unncesary overcomplication as it likely will results in multiple unintended consequences. Rather, I think a more simple solution is to make Robotics Tech cheaper in order to open up for multiple Robotics being used in the midgame. This way, it becomes much easier for the toss to get more Immortals into his composition which (too some extent) can free up the Stalker-role.
What's your solution to make Blink Stalkers more centered around an area?
The amount of changes Protoss needs is going to cause so many unnecessary complications anyway that nothing should be off the table.
|
I see what you're saying, but the truth is Banelings are the only unit that is used differently against Stimmed Marines compared to regular Marines. In every other situation, including all TvPs, Stim just makes Marines better in every way while providing no concrete way for the opponent to counter-micro.
Well this isn't related to stim itself though. It's not like there is this interaction which is pretty decent between unstimmed Marines and other units, but get's destroyed when stim is added. At least it seems realistic (with solid design) to create interactions between stimmed bio units and enemy units.
What's your solution to make Blink Stalkers more centered around an area?
I meant that Blink Stalkers would be more used as a form of harassment/pressure/multitaskrelated situations, instead of being an important unit during a straight up comebat.
The amount of changes Protoss needs is going to cause so many unnecessary complications anyway that nothing should be off the table.
Dno. With my suggested approach, the WG-issue is also "fixed" as the defenders advantage is increased when Stalkers become less important and Immortals more important. I am not suggesting as big changes here as other ppl who want to completely redeisgn everything about protoss. Rather, this is what I would do here:
- Small nerf to Stalkers in straight up combat, some kind of (small) mobility compensation - Robo cost reduced to 150/50. - Immortal stats changed to so it fits into gameplay dynamics as a core-unit.
|
Actually there is counter micro against blink. Watch pro stream, they often move their units ahead to snipe the injuries stalker even before they blink back and focus fire once they blinked. You can see this in zvp more often
|
On July 26 2014 09:54 ETisME wrote: Actually there is counter micro against blink. Watch pro stream, they often move their units ahead to snipe the injuries stalker even before they blink back and focus fire once they blinked. You can see this in zvp more often
With Hydras, it's still possible in some situations yes. But for that to be possible, unit counts and composition mix needs to be just about right. My point is that this type of micro would still be there if blink didn't exist, and in many situations you cannot micro against blinks (try doing this with Roaches for instance). Blinks in those situations removes the micro that would otherwise exist. So this isn't a counter-micromicro that is added specifically against blink, rather you are just saying that in some situations blink doesn't prevent opponent from microing.
|
Roach ling as well. You see this kind of micro usually against blink.stalker all in. You bracket a small group of your roach from your big pack and move them forward and focus fire that stalker after it blinked
I don't quite understand your argument, because isn't a stim pretty much similar? How do you counter micro stim except using fungal? How do you out micro magic boxed muta with thors?
|
On July 26 2014 07:55 Big J wrote:Show nested quote +On July 26 2014 03:54 Faust852 wrote:And for people saying Z aren't way stronger than T : On July 24 2014 15:32 Ghanburighan wrote:While we're looking at winrates, here's another Aligulac list: Just looking at winrates, PvT is rather even, and so is PvZ but TvZ has gone down to the dumps again. On the other hand, the population numbers are the worst ever for Terran. It looks like T has a constant of around 100 games every period, but with the added number of games (last period has 1799 games, this one 3866), only Z and P seem to have added more mirrors. So there are 4.8x as many ZvZ as TvT, and 3.8x as many PvP as TvT. This also means that P has once again caught up with Z populations, last period it was 1.3 ZvZ for every 1 PvP, now it's 1.2. On July 10 2014 20:15 Ghanburighan wrote:Here's the latest Aligulac list (114) with pretty new formatting. With regard to P, nothing seems to have changed. Just like the first half of June, P>T by a slight margin, P and Z are roughly even, and there are roughly the same number of PvP MU's in tournaments. Z did worse in this period, while it was at >55% against T last time, it's now even in winrates. More importantly, looking at populations, while there were 5x more ZvZ than TvT, and 2x more ZvZ than PvP, then now there are only roughly 3x more ZvZ than TvT, and a just over a fourth more ZvZ than PvPs. This suggests that Z is doing worse, and it's mainly doing worse against T (note that worse doesn't imply that they're doing bad, this is a comparison with the previous period). Looking more closely at the population numbers, there appear to have been fewer games, the total for 114 is 1835 and for 113 it was 2379. So for the previous 113 list Z MUs made up 72% of all MUs. P MUs made up 55% (note that the overlap is due to the fact that P plays Z...). T MUs made up 36% of all MUs. In this list, 114, Z MUs made up 65% of all MUs. P MUs made up 57%. T MUs made up 42% of all MUs. So Z is down 7%, P is up 2% and T is up 6%. (with rounding) The previous lists can be found below. On June 29 2014 05:42 Ghanburighan wrote:Sorry for the delay, here's Aligulac 113.. The previous list(s) can be found at the end of this post. Looking at the winrates, P has extended its advantage over T, P has also gained some ground back against Z, yet TvZ has strongly turned in Z favour once gain (it's as bad as it was before the hellbat patch in April). Population numbers are also worse. Previously there were 4x more ZvZ games than TvT games, now there are more than 5x. PvP's have not changed in number, so it's mostly just less terrans and more zergs getting further that's creating the problem. All in all, balance-wise this was a very depressing period. On June 12 2014 15:32 Ghanburighan wrote:Time to post the latest Aligulac list. The previous list can be found at the end of this post. Regarding winrates, PvT has fluctuated back from T having a slight advantage to P having a minuscule advantage. In PvZ, P has also improved although it hasn't caught up with Z. On the other hand, T has improved in the TvZ MU (110 had 45%, 111 had 47%) and its even now. In terms of populations measured in numbers of mirror MUs, there's virtually no change compared to the last list, the proportions are very close. This means that there is no repopulation of terrans according to these numbers and there are 4 times fewer TvTs than ZvZs. As T MUs have even winrates, there cannot really be a repopulation with these numbers. Furthermore, a word of caution, I'd say that this was one of the best periods for Terran in a long while, Taeja won Hsc 9 (where Z had a comparatively weaker list of players), Maru is tearing up Code S, and Innovation is kicking as in teamleagues and the Dragon cup. I don't think they contributed overly much to the final winrates (their games are still a small fraction of all the games), but taken together they did contribute significantly. If they don't keep their winning ways going, winrates can plunge below 50% again. And, their wins aren't helping repopulate in any way. On May 29 2014 02:45 Ghanburighan wrote:Uploading the latest Aligulac list. Unfortunately there was a TvZ patch in the middle of the period, so those numbers could be anything now. But it looks like P is doing worse against Z in terms of winrate. But the population ratios haven't changed compared to the last list, though. It's still roughly 1/4 TvT, 2/4 PvP and 1/1 ZvZ. Hold your horses. 45% for 1-2weeks isnt that bad, given the standards Terrans have taught us in 2013. ;-) More seriously, Im not against the usage of aligulac (feed me as many stats as possible), yet I've become very careful since that post that revealed that a third (?) of its database are Australian tournaments. It's nice to know that they have a very active community, but it's not the same as looking at Europa or Korea or america alone. Since I have been watching mainly Korean stats, I know that they have very balanced winrates in TvZ. I think in EU it is more Z favored (not just WCs but also e.g. DH). Not sure about America.
There's two points you're neglecting here. First, not every period is contaminated by Aussies, at least it hasn't been showed. The time it was contaminated was also a period with very few games, unlike this one. Secondly, the population numbers need to be taken into account as well. If you look at GSL numbers without population numbers, you're measuring the results of the very best T against all Z. And you know that's not going to give you a reliable picture. The quoted post showed that population numbers are the most skewed ever.
|
On July 26 2014 14:59 Ghanburighan wrote:Show nested quote +On July 26 2014 07:55 Big J wrote:On July 26 2014 03:54 Faust852 wrote:And for people saying Z aren't way stronger than T : On July 24 2014 15:32 Ghanburighan wrote:While we're looking at winrates, here's another Aligulac list: Just looking at winrates, PvT is rather even, and so is PvZ but TvZ has gone down to the dumps again. On the other hand, the population numbers are the worst ever for Terran. It looks like T has a constant of around 100 games every period, but with the added number of games (last period has 1799 games, this one 3866), only Z and P seem to have added more mirrors. So there are 4.8x as many ZvZ as TvT, and 3.8x as many PvP as TvT. This also means that P has once again caught up with Z populations, last period it was 1.3 ZvZ for every 1 PvP, now it's 1.2. On July 10 2014 20:15 Ghanburighan wrote:Here's the latest Aligulac list (114) with pretty new formatting. With regard to P, nothing seems to have changed. Just like the first half of June, P>T by a slight margin, P and Z are roughly even, and there are roughly the same number of PvP MU's in tournaments. Z did worse in this period, while it was at >55% against T last time, it's now even in winrates. More importantly, looking at populations, while there were 5x more ZvZ than TvT, and 2x more ZvZ than PvP, then now there are only roughly 3x more ZvZ than TvT, and a just over a fourth more ZvZ than PvPs. This suggests that Z is doing worse, and it's mainly doing worse against T (note that worse doesn't imply that they're doing bad, this is a comparison with the previous period). Looking more closely at the population numbers, there appear to have been fewer games, the total for 114 is 1835 and for 113 it was 2379. So for the previous 113 list Z MUs made up 72% of all MUs. P MUs made up 55% (note that the overlap is due to the fact that P plays Z...). T MUs made up 36% of all MUs. In this list, 114, Z MUs made up 65% of all MUs. P MUs made up 57%. T MUs made up 42% of all MUs. So Z is down 7%, P is up 2% and T is up 6%. (with rounding) The previous lists can be found below. On June 29 2014 05:42 Ghanburighan wrote:Sorry for the delay, here's Aligulac 113.. The previous list(s) can be found at the end of this post. Looking at the winrates, P has extended its advantage over T, P has also gained some ground back against Z, yet TvZ has strongly turned in Z favour once gain (it's as bad as it was before the hellbat patch in April). Population numbers are also worse. Previously there were 4x more ZvZ games than TvT games, now there are more than 5x. PvP's have not changed in number, so it's mostly just less terrans and more zergs getting further that's creating the problem. All in all, balance-wise this was a very depressing period. On June 12 2014 15:32 Ghanburighan wrote:Time to post the latest Aligulac list. The previous list can be found at the end of this post. Regarding winrates, PvT has fluctuated back from T having a slight advantage to P having a minuscule advantage. In PvZ, P has also improved although it hasn't caught up with Z. On the other hand, T has improved in the TvZ MU (110 had 45%, 111 had 47%) and its even now. In terms of populations measured in numbers of mirror MUs, there's virtually no change compared to the last list, the proportions are very close. This means that there is no repopulation of terrans according to these numbers and there are 4 times fewer TvTs than ZvZs. As T MUs have even winrates, there cannot really be a repopulation with these numbers. Furthermore, a word of caution, I'd say that this was one of the best periods for Terran in a long while, Taeja won Hsc 9 (where Z had a comparatively weaker list of players), Maru is tearing up Code S, and Innovation is kicking as in teamleagues and the Dragon cup. I don't think they contributed overly much to the final winrates (their games are still a small fraction of all the games), but taken together they did contribute significantly. If they don't keep their winning ways going, winrates can plunge below 50% again. And, their wins aren't helping repopulate in any way. On May 29 2014 02:45 Ghanburighan wrote:Uploading the latest Aligulac list. Unfortunately there was a TvZ patch in the middle of the period, so those numbers could be anything now. But it looks like P is doing worse against Z in terms of winrate. But the population ratios haven't changed compared to the last list, though. It's still roughly 1/4 TvT, 2/4 PvP and 1/1 ZvZ. Hold your horses. 45% for 1-2weeks isnt that bad, given the standards Terrans have taught us in 2013. ;-) More seriously, Im not against the usage of aligulac (feed me as many stats as possible), yet I've become very careful since that post that revealed that a third (?) of its database are Australian tournaments. It's nice to know that they have a very active community, but it's not the same as looking at Europa or Korea or america alone. Since I have been watching mainly Korean stats, I know that they have very balanced winrates in TvZ. I think in EU it is more Z favored (not just WCs but also e.g. DH). Not sure about America. There's two points you're neglecting here. First, not every period is contaminated by Aussies, at least it hasn't been showed. The time it was contaminated was also a period with very few games, unlike this one. Secondly, the population numbers need to be taken into account as well. If you look at GSL numbers without population numbers, you're measuring the results of the very best T against all Z. And you know that's not going to give you a reliable picture. The quoted post showed that population numbers are the most skewed ever. The thing is that almost every statistics you can get is skewed one way or another. In addition as we have seen people cherry pick "facts" from those stats that support their arguments or use the ever popular: "we can´t count these X players because they are too good" method. Just saying that you can argue over stats forever if you want.
|
On July 26 2014 14:59 Ghanburighan wrote:Show nested quote +On July 26 2014 07:55 Big J wrote:On July 26 2014 03:54 Faust852 wrote:And for people saying Z aren't way stronger than T : On July 24 2014 15:32 Ghanburighan wrote:While we're looking at winrates, here's another Aligulac list: Just looking at winrates, PvT is rather even, and so is PvZ but TvZ has gone down to the dumps again. On the other hand, the population numbers are the worst ever for Terran. It looks like T has a constant of around 100 games every period, but with the added number of games (last period has 1799 games, this one 3866), only Z and P seem to have added more mirrors. So there are 4.8x as many ZvZ as TvT, and 3.8x as many PvP as TvT. This also means that P has once again caught up with Z populations, last period it was 1.3 ZvZ for every 1 PvP, now it's 1.2. On July 10 2014 20:15 Ghanburighan wrote:Here's the latest Aligulac list (114) with pretty new formatting. With regard to P, nothing seems to have changed. Just like the first half of June, P>T by a slight margin, P and Z are roughly even, and there are roughly the same number of PvP MU's in tournaments. Z did worse in this period, while it was at >55% against T last time, it's now even in winrates. More importantly, looking at populations, while there were 5x more ZvZ than TvT, and 2x more ZvZ than PvP, then now there are only roughly 3x more ZvZ than TvT, and a just over a fourth more ZvZ than PvPs. This suggests that Z is doing worse, and it's mainly doing worse against T (note that worse doesn't imply that they're doing bad, this is a comparison with the previous period). Looking more closely at the population numbers, there appear to have been fewer games, the total for 114 is 1835 and for 113 it was 2379. So for the previous 113 list Z MUs made up 72% of all MUs. P MUs made up 55% (note that the overlap is due to the fact that P plays Z...). T MUs made up 36% of all MUs. In this list, 114, Z MUs made up 65% of all MUs. P MUs made up 57%. T MUs made up 42% of all MUs. So Z is down 7%, P is up 2% and T is up 6%. (with rounding) The previous lists can be found below. On June 29 2014 05:42 Ghanburighan wrote:Sorry for the delay, here's Aligulac 113.. The previous list(s) can be found at the end of this post. Looking at the winrates, P has extended its advantage over T, P has also gained some ground back against Z, yet TvZ has strongly turned in Z favour once gain (it's as bad as it was before the hellbat patch in April). Population numbers are also worse. Previously there were 4x more ZvZ games than TvT games, now there are more than 5x. PvP's have not changed in number, so it's mostly just less terrans and more zergs getting further that's creating the problem. All in all, balance-wise this was a very depressing period. On June 12 2014 15:32 Ghanburighan wrote:Time to post the latest Aligulac list. The previous list can be found at the end of this post. Regarding winrates, PvT has fluctuated back from T having a slight advantage to P having a minuscule advantage. In PvZ, P has also improved although it hasn't caught up with Z. On the other hand, T has improved in the TvZ MU (110 had 45%, 111 had 47%) and its even now. In terms of populations measured in numbers of mirror MUs, there's virtually no change compared to the last list, the proportions are very close. This means that there is no repopulation of terrans according to these numbers and there are 4 times fewer TvTs than ZvZs. As T MUs have even winrates, there cannot really be a repopulation with these numbers. Furthermore, a word of caution, I'd say that this was one of the best periods for Terran in a long while, Taeja won Hsc 9 (where Z had a comparatively weaker list of players), Maru is tearing up Code S, and Innovation is kicking as in teamleagues and the Dragon cup. I don't think they contributed overly much to the final winrates (their games are still a small fraction of all the games), but taken together they did contribute significantly. If they don't keep their winning ways going, winrates can plunge below 50% again. And, their wins aren't helping repopulate in any way. On May 29 2014 02:45 Ghanburighan wrote:Uploading the latest Aligulac list. Unfortunately there was a TvZ patch in the middle of the period, so those numbers could be anything now. But it looks like P is doing worse against Z in terms of winrate. But the population ratios haven't changed compared to the last list, though. It's still roughly 1/4 TvT, 2/4 PvP and 1/1 ZvZ. Hold your horses. 45% for 1-2weeks isnt that bad, given the standards Terrans have taught us in 2013. ;-) More seriously, Im not against the usage of aligulac (feed me as many stats as possible), yet I've become very careful since that post that revealed that a third (?) of its database are Australian tournaments. It's nice to know that they have a very active community, but it's not the same as looking at Europa or Korea or america alone. Since I have been watching mainly Korean stats, I know that they have very balanced winrates in TvZ. I think in EU it is more Z favored (not just WCs but also e.g. DH). Not sure about America. There's two points you're neglecting here. First, not every period is contaminated by Aussies, at least it hasn't been showed. The time it was contaminated was also a period with very few games, unlike this one. Secondly, the population numbers need to be taken into account as well. If you look at GSL numbers without population numbers, you're measuring the results of the very best T against all Z. And you know that's not going to give you a reliable picture. The quoted post showed that population numbers are the most skewed ever.
Well, the population numbers in Code A, Code A Qualifiers and Proleauge werent that bad. Moreover, I have mabually checked a few days of that periode and there were 1-2 such Australian tournaments (ACL, Seacraft) that did have like 20-0-100 stats in terms of mirrors. I dont know if thats all or if there's more Aussi stuff around for said period, but the fact alone that these kinds of tournaments are mixes with Code A etc. makes me more careful with these stats. Same obviously also goes for all the other weekly cups like go4sc2. That's not saying that aligulac is wrong, rather that I wouldnt give too much on the exact numbers. A movement of a percent or a few hundred mirror matches can easily happen without any implications an balance, just based upon wether a certain amateur tournament is being held an recorded or not.
|
|
I don't understand. Is this live in EU now? I don't see new patch in the Bnet client?
|
On July 26 2014 17:22 mage3 wrote: I don't understand. Is this live in EU now? I don't see new patch in the Bnet client? It's a server patch, not client. They don't need to force an update for changing variable.
|
|
|
|