TvZ Winrates with Mass Widow Mine - Page 24

On April 23 2013 06:54 TheDwf wrote:

My sincere apologies. You need to be a chef to judge any meal, pro players never make any mistake and cancelling gas + letting your Spine die in front of the Bunker is a marvelous defence to hold 8-8-8!

On April 23 2013 21:53 Topdoller wrote:
Stephano as actually better off studying for his degree, by the time Zerg work out how to defeat all the bullshit timings and all ins from HOTS and LOTV he could then return in 4 years and start again or go on to a real life profession.

He has the money to do that now and he did say he was only interested in the money

On April 23 2013 20:44 TheDwf wrote:

How so? Zerg winning pre-Hive is absolutely not unheard of, and all the games in which Zerg won at Hive means Zerg did survive the Lair period just fine.


No. Not at all. Not only statistics are easily manipulated to make them tell whatever you want, but on top of that once you have said TvZ is 59% on your glorious 59 games sample size (so please note I am being indulgent even lending an ear to this; all I need to do is laugh at your sample size and you would have absolutely nothing to reply) you have said essentially nothing. Or rather, you have just made numerous assumptions without bringing the slightest evidence for any of them:

• You have assumed that in your 59 games, the names involved from each side were always of equal skill. Of course this is false: just because the games happened at the highest level doesn't mean both players are strictly or even roughly identical skill-wise. Or maybe Flash and Bboong are to be considered even simply on the basis they played at Code S?
  
  
• You have assumed that both Terrans and Zergs have equal rates of development regarding the strategies of their respective races. It happens a lot that a race "discovers" something, and it takes time (i. e. weeks, months) for the other race to fully understand the necessary tweaks to be made (in builds, or micro, or game plans, etc.). And I am not even only talking about something like "microing against Mines," I am also talking about builds. You don't play the same way against a 65 SCVs dual reactor fact 3-bases heavy agression style with a very late fourth as you would against a Terran going 70-75 SCVs, lab on second fact and a 13'30 fourth. Where is your evidence that Zergs have already found all the adjustments? If we even assume there is a gap, where is your proof that Zergs don't have any room left for innovation to overcome this gap?
  It took weeks, perhaps even months for Protoss to find the answers to the Roach max style—assuming something is wrong, where is your proof a similar thing is not happening here? Maybe you want a link to the thread about defending Roach max? Guess we would find many, many complaints about the Roach. "How can this thing be balanced? It only costs 75/25." (Sounds familiar?) Nerf, nerf, blah, blah, blah. The result at the end of the WoL? Roach max vs P was naturally extinct, with no balance changes (except the maps). But maybe Blizzard should have patched Roaches and make them cost 100/25 or removed 20 hit points just because Protoss faced a temporary difficulty before finding the appropriate builds?
  
  
• You have assumed that all games are meaningful to determine balance. When HyuN kills 0 SCV with his Roach/Speedlings attack against Bogus on the Akilon Wastes game, he loses not because "Terran is imbalanced," but because his disguised all-in failed badly. Still, you will aggregate his "loss because of a failed all-in" with "losses in an even game" then claim that your 59 games sample size is a "hard proof". When DRG loses 24 drones to YoDa's Hellion drop on Whirlwind, the rest of the game is a formality. When Bboong cancels his gas against Maru's 8-8-8, his odds at winning the game instantly fall to 0%. Still, you will use his blunder to say Terran dominates Zerg, but of course you can see this is not legitimate. It's like saying "look, RorO won against Bomber on Akilon Wastes, Zergs are fine," then you watch the game and you see Bomber destroyed his own OC. If your sample is a tree, you just have to prune some of the branchs because some blunders or situations make them meaningless to determine balance.

You also have to consider maps. Then you have to look at the actual content of the games, which thus supposes some game knowledge to enlighten why A or B lost... And from my biased amateur perspective, I can tell you that many Zerg losses are perfectly inconclusive about TvZ balance. When TRUE fails a Baneling bust against FanTaSy on Star Station, when HyuN blindly drones to 76 whereas GuMiho goes for a 2-base all-in, when HyuN goes for a 3-bases 46 drones Roaches/Banelings all-in against GuMiho and ends up facing Marines/Tanks behind a wall, when HyuN loses 10 Drones to a 4 Hellions raid from Bogus on Bel'shir Vestige, when Bboong loses 17 Drones to a 4 → 6 Hellions raid from Flash on Whirlwind, what can we say about balance? Nothing. Notice how with all of those examples and the 3 from my last bullet point above, I have listed nearly half of the Terran wins in Code S... The same thing can be said about the other side, by the way. When Terran loses because he doesn't have full sight in his base and a Nydus Worm thus goes up unnoticed, you cannot make any comment about balance.

So if I am clueless about what happens in pro games, fine; but then it means you're even less authorized to make comments about balance, not only because you know less about the game but also because your statistics knowledge seem very limited if you think you can draw any conclusion from a sample size in which a single series can make so much difference in percentages (see Ghanburighan's example).

On April 23 2013 20:44 TheDwf wrote:

How so? Zerg winning pre-Hive is absolutely not unheard of, and all the games in which Zerg won at Hive means Zerg did survive the Lair period just fine.


No. Not at all. Not only statistics are easily manipulated to make them tell whatever you want, but on top of that once you have said TvZ is 59% on your glorious 59 games sample size (so please note I am being indulgent even lending an ear to this; all I need to do is laugh at your sample size and you would have absolutely nothing to reply) you have said essentially nothing. Or rather, you have just made numerous assumptions without bringing the slightest evidence for any of them:

• You have assumed that in your 59 games, the names involved from each side were always of equal skill. Of course this is false: just because the games happened at the highest level doesn't mean both players are strictly or even roughly identical skill-wise. Or maybe Flash and Bboong are to be considered even simply on the basis they played at Code S?
  
  
• You have assumed that both Terrans and Zergs have equal rates of development regarding the strategies of their respective races. It happens a lot that a race "discovers" something, and it takes time (i. e. weeks, months) for the other race to fully understand the necessary tweaks to be made (in builds, or micro, or game plans, etc.). And I am not even only talking about something like "microing against Mines," I am also talking about builds. You don't play the same way against a 65 SCVs dual reactor fact 3-bases heavy agression style with a very late fourth as you would against a Terran going 70-75 SCVs, lab on second fact and a 13'30 fourth. Where is your evidence that Zergs have already found all the adjustments? If we even assume there is a gap, where is your proof that Zergs don't have any room left for innovation to overcome this gap?
  It took weeks, perhaps even months for Protoss to find the answers to the Roach max style—assuming something is wrong, where is your proof a similar thing is not happening here? Maybe you want a link to the thread about defending Roach max? Guess we would find many, many complaints about the Roach. "How can this thing be balanced? It only costs 75/25." (Sounds familiar?) Nerf, nerf, blah, blah, blah. The result at the end of the WoL? Roach max vs P was naturally extinct, with no balance changes (except the maps). But maybe Blizzard should have patched Roaches and make them cost 100/25 or removed 20 hit points just because Protoss faced a temporary difficulty before finding the appropriate builds?
  
  
• You have assumed that all games are meaningful to determine balance. When HyuN kills 0 SCV with his Roach/Speedlings attack against Bogus on the Akilon Wastes game, he loses not because "Terran is imbalanced," but because his disguised all-in failed badly. Still, you will aggregate his "loss because of a failed all-in" with "losses in an even game" then claim that your 59 games sample size is a "hard proof". When DRG loses 24 drones to YoDa's Hellion drop on Whirlwind, the rest of the game is a formality. When Bboong cancels his gas against Maru's 8-8-8, his odds at winning the game instantly fall to 0%. Still, you will use his blunder to say Terran dominates Zerg, but of course you can see this is not legitimate. It's like saying "look, RorO won against Bomber on Akilon Wastes, Zergs are fine," then you watch the game and you see Bomber destroyed his own OC. If your sample is a tree, you just have to prune some of the branchs because some blunders or situations make them meaningless to determine balance.

You also have to consider maps. Then you have to look at the actual content of the games, which thus supposes some game knowledge to enlighten why A or B lost... And from my biased amateur perspective, I can tell you that many Zerg losses are perfectly inconclusive about TvZ balance. When TRUE fails a Baneling bust against FanTaSy on Star Station, when HyuN blindly drones to 76 whereas GuMiho goes for a 2-base all-in, when HyuN goes for a 3-bases 46 drones Roaches/Banelings all-in against GuMiho and ends up facing Marines/Tanks behind a wall, when HyuN loses 10 Drones to a 4 Hellions raid from Bogus on Bel'shir Vestige, when Bboong loses 17 Drones to a 4 → 6 Hellions raid from Flash on Whirlwind, what can we say about balance? Nothing. Notice how with all of those examples and the 3 from my last bullet point above, I have listed nearly half of the Terran wins in Code S... The same thing can be said about the other side, by the way. When Terran loses because he doesn't have full sight in his base and a Nydus Worm thus goes up unnoticed, you cannot make any comment about balance.

So if I am clueless about what happens in pro games, fine; but then it means you're even less authorized to make comments about balance, not only because you know less about the game but also because your statistics knowledge seem very limited if you think you can draw any conclusion from a sample size in which a single series can make so much difference in percentages (see Ghanburighan's example).

On April 24 2013 05:51 Ghanburighan wrote:


If you are not going to read what people respond to you, why are you here discussing it. Go post a blog, or tweet about it. This is meant to be a discussion, not Emzeeshady's monologue hour.

On April 23 2013 08:58 Markwerf wrote:
Of course there were plenty of other changes like hydra's, vipers, hellbats and swarmhosts but they don't really see have much use in the new 'standard' TvZ.
Especially prehive zerg is just screwed. You can easily go as far as saying that the infestor nerf outweighs the muta buff while terran only got buffs midgame. Ultralisks lategame make up quite a bit because they are far better but I think if the terrans play MMMM aggressively well you can't really make it to lategame without being far behind. Even when getting there the classic broodlords aren't even that scary anymore since terran can easily have decently upgraded vikings and ravens to counter that.