|
On April 08 2013 11:13 BronzeKnee wrote:Show nested quote +On April 08 2013 11:06 opterown wrote:On April 08 2013 11:04 BronzeKnee wrote:On April 08 2013 11:00 opterown wrote:On April 08 2013 10:57 dsjoerg wrote:On April 08 2013 10:48 Emzeeshady wrote:
Also, we already knew Widow Mines are really good vs Zerg. This isn't all that informative. Well there ya go. OP is either wrong or obvious. Or both!  In other news, here's the >10 mutas stats: well with the muta analysis, have we not reached a similar conclusion for ZvP? 49% with less than 10 mutas, 63% with more than 10 mutas for master level. the GM level is even greater (but with less sample size, of course). so we should also discuss how mutas are very strong in ZvP, stronger than widow mines are in TvZ nice stats, btw :D commend you on that! So, you just admitted that his analysis of Widow Mines was correct. And this also explains why every Zerg build Mutalisks in ZvP. However, the reasoning is pretty simple. In WoL, Storm would eat away at Mutalisk packs, while in HotS a single storm isn't crippling because they can move off and regen. Thus Protoss players are forced more and more into Phoenixes, which is similar to when a Protoss player forces a Zerg into Corrupters, you can switch quickly into ground units and just win. when did i say it was correct? i said: if you claim that widow mines are strong against zerg, i expect a similar claim for zvp. no such claim has come out. the fact that mutas, another tier 2 unit has even a greater differential in stats should be a bigger concern. therefore by following your line of reasoning, we need to nerf mutas before we nerf widow mines :D i commend his work and effort, not his analysis or conclusion I'm going try and explain this as simply as possible. Reading the OP, we find that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics. That is the analysis and the conclusion.
No, you're trying to extrapolate the data when the data is shaky. Here's a better idea of what the data tells us:
"In this particular sample, those Terrans who made more than 10 Widow Mines (for whatever reason, including surviving long enough to do so, facing army compositions which may have been weak against Widow Mines, etc.) won 59% of their games as opposed to Terrans who made less than 10 Widow Mines and won 49% of their games."
|
On April 08 2013 11:10 sibs wrote:Show nested quote +This was true for 2010 and 2011 but after all the terran early game nerfs and the queen buff on top of it terran had the worst early game rushes which was proven statistically. is this a troll post? Zerg has by far the worst early game cheeses, from 2010 to 2013, I don't see how this is even arguable, protoss doesn't have much either, I think terran has had best early game shenanigans throughout all the patches.
No. Post Queen Patch, terrans cheeses like 11/11 were totally useless. MKP is the evidence haha.
Anyway, I think these "statistics" are pretty useless. 10 widow mines in all the game are nothing lol. Even when I play marines/marauders/hellbats, I build like 20 mines in a 30 min games, to defend myself or dropping them early, but clearly they are not the key of my victories. 10 is such a subjective choise... If we look in your statistics in another way, without mine Terrans have 40%, so it would implies that without them, Zerg are OP ? Stupid way of thinking.
|
i can safely say coming from a zerg player that widow mines are a pain in the fugging arse to deal with, if i were terran id be building plenty of them too.
|
On April 08 2013 11:13 opterown wrote:Show nested quote +On April 08 2013 11:11 m0ck wrote:On April 08 2013 11:07 opterown wrote:On April 08 2013 11:04 m0ck wrote:On April 08 2013 11:03 opterown wrote:On April 08 2013 10:42 BronzeKnee wrote:On April 08 2013 10:38 Entirety wrote:
- People who make more Widow Mines had the time/resources to do so.
When a Terran dies to 6pool, that is added to the category of "did not produce 10 Widow Mines" and lowers the win rate. Just the fact that the Terran even produced 10 Widow Mines means that the Terran survived really early game cheese.
Yes, because Terrans in Masters league die to 6 pools all the time... I clearly remember Idra destroying aLive with 6 Pools... Are you joking? If anything what you said raises the win rate! 11/11 is still pretty strong too... if anything I'd argue that if short games favor a high winrate for Terran, and this has been shown statistically many times in WOL. 11/11 is actually quite a weak build, it relies on zerg players going for something ridic like hatch gas 17pool or botching their micro (which terrans are also well capable of botching too). Yeah, it's a weak 50/50% win build. Come on, 2-rax in whatever form is not a weak build.. * and 10 wm = 4/5 mutas in cost it's a lot lower than 50/50 in pro games Well, I would disagree, but we can't really find an answer to that. 11/11 can be used against all three races, it has been successfully employed from the beta and to the very end of WoL, it has been used to decide the very highest level of matches & it will continue to be used in HotS (at least vs Z). If the effectiveness dropped in WoL (and that is up for debate) I would argue that it was due to how often T would use it during the Z>T era. 11/11 almost never works in TvT, and only really works in TvP when the protoss derps a lot. it's used at the highest level since it's such a high pressure environment and micro mistakes make it more coinflippy. i'm fairly sure the 8/8/8 is a more common and better build these days anyway
Coinflippy sounds a lot like 50/50 to me ^^
Regardless of whether it works or not, it was certainly common enough in GSL all the way to the end (but yeah, other proxy-builds were more popular at times).
8-8-8 is newer and more interesting, but terran will really show their power once they start rediscovering the WoL openings, I reckon.
Anywho, this has gotten really off track, my bad.
|
opterown
Australia54784 Posts
On April 08 2013 11:17 BronzeKnee wrote:Show nested quote +On April 08 2013 11:14 opterown wrote:On April 08 2013 11:13 BronzeKnee wrote:On April 08 2013 11:06 opterown wrote:On April 08 2013 11:04 BronzeKnee wrote:On April 08 2013 11:00 opterown wrote:On April 08 2013 10:57 dsjoerg wrote:On April 08 2013 10:48 Emzeeshady wrote:
Also, we already knew Widow Mines are really good vs Zerg. This isn't all that informative. Well there ya go. OP is either wrong or obvious. Or both! In other news, here's the >10 mutas stats: well with the muta analysis, have we not reached a similar conclusion for ZvP? 49% with less than 10 mutas, 63% with more than 10 mutas for master level. the GM level is even greater (but with less sample size, of course). so we should also discuss how mutas are very strong in ZvP, stronger than widow mines are in TvZ nice stats, btw :D commend you on that! So, you just admitted that his analysis of Widow Mines was correct. And this also explains why every Zerg build Mutalisks in ZvP. However, the reasoning is pretty simple. In WoL, Storm would eat away at Mutalisk packs, while in HotS a single storm isn't crippling because they can move off and regen. Thus Protoss players are forced more and more into Phoenixes, which is similar to when a Protoss player forces a Zerg into Corrupters, you can switch quickly into ground units and just win. when did i say it was correct? i said: if you claim that widow mines are strong against zerg, i expect a similar claim for zvp. no such claim has come out. the fact that mutas, another tier 2 unit has even a greater differential in stats should be a bigger concern. therefore by following your line of reasoning, we need to nerf mutas before we nerf widow mines :D i commend his work and effort, not his analysis or conclusion I'm going try and explain this as simply as possible. Reading the OP, we find that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics. That is the analysis and the conclusion. Though the OP states "Of course this doesn't exactly prove that Widow Mines are OP, but at the very least it suggests that getting/preventing mass widow mines is an important part of TvZ" It basically says nothing. So now you state: On April 08 2013 10:30 opterown wrote:
i bet if you compared games with over/under ten mutalisks, or over/under ten high templar, etc you would get similar stats And then someone states such statistics. You commend the work and effort. By commending the work and effort, you are basically saying "hey look you did a good job collecting statistics" and that is opposed to doing a bad job, in which the statistics are basically wrong. Thus, since the analysis and conclusion were based entirely on the collection of statistics and he makes no inference on whether or not Widow Mines are OP, and then you commend the very same process of statistical collection, you have agreed to the conclusion of the OP which is: "that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics." cool, so you get some winrates. i'm not attacking those numbers, i'm saying the conclusions that a lot of people are going to make are invalid. EXACTLY! Thank you! But remember this, those conclusions that argue that Widow Mines are not overpowered now have to be independent of the statistic presented! Otherwise they are invalid!
and whenever did i say the statistics were wrong? i claimed that
a) analysis of over/under 10 of other units would get similar stats
b) you cannot conclude anything significant from this
so i don't quite get why you are specifically talking to me about this
|
On April 08 2013 11:17 Entirety wrote:Show nested quote +On April 08 2013 11:13 BronzeKnee wrote:On April 08 2013 11:06 opterown wrote:On April 08 2013 11:04 BronzeKnee wrote:On April 08 2013 11:00 opterown wrote:On April 08 2013 10:57 dsjoerg wrote:On April 08 2013 10:48 Emzeeshady wrote:
Also, we already knew Widow Mines are really good vs Zerg. This isn't all that informative. Well there ya go. OP is either wrong or obvious. Or both! In other news, here's the >10 mutas stats: well with the muta analysis, have we not reached a similar conclusion for ZvP? 49% with less than 10 mutas, 63% with more than 10 mutas for master level. the GM level is even greater (but with less sample size, of course). so we should also discuss how mutas are very strong in ZvP, stronger than widow mines are in TvZ nice stats, btw :D commend you on that! So, you just admitted that his analysis of Widow Mines was correct. And this also explains why every Zerg build Mutalisks in ZvP. However, the reasoning is pretty simple. In WoL, Storm would eat away at Mutalisk packs, while in HotS a single storm isn't crippling because they can move off and regen. Thus Protoss players are forced more and more into Phoenixes, which is similar to when a Protoss player forces a Zerg into Corrupters, you can switch quickly into ground units and just win. when did i say it was correct? i said: if you claim that widow mines are strong against zerg, i expect a similar claim for zvp. no such claim has come out. the fact that mutas, another tier 2 unit has even a greater differential in stats should be a bigger concern. therefore by following your line of reasoning, we need to nerf mutas before we nerf widow mines :D i commend his work and effort, not his analysis or conclusion I'm going try and explain this as simply as possible. Reading the OP, we find that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics. That is the analysis and the conclusion. No, you're trying to extrapolate the data when the data is shaky. Here's a better idea of what the data tells us: "In this particular sample, those Terrans who made more than 10 Widow Mines (for whatever reason, including surviving long enough to do so, facing army compositions which may have been weak against Widow Mines, etc.) won 59% of their games as opposed to Terrans who made less than 10 Widow Mines and won 49% of their games."
What you said is exactly the same as what I said.
You said "in this particular sample" I said: "based on GGtracker statistics" (GGtracker statistics is the sample)
You said 59% and 49%, I said 10% higher.
They mean exactly the same thing.
|
The guy who talked about 6 pools was actually right. If you take into account only games where Terran produced more than 10 widow mines, that's already every game where Terran survived for more than 10 minutes, i.e. not losing to early pools, speedlings all ins, banelings busts, roach all ins, burrow shenanigans and I don't know what else.
I may be wrong on that matter, but at pro level I saw many more Zerg early killing blow attempts (because they don't want to face T in lategame I don't know) than Terran's, who are more comfortable being greedy with triple CCs nowadays, which is a build that kill no one before 12 minutes.
If you compute the overall winrate, you get a mere 52%, which doesn't point at all towards any imbalance.
Edit: What I want to say is that the number of widow mines produced is just a very very bad and shady measure of in game time elapsed. In WoL, if you would have said "Zerg wins 70% of their game if they spread more than 20 creep tumors", that would probably have been true, yet an equally stupid and obscure way to tell that Zerg pwns the lategame.
|
On April 08 2013 11:11 opterown wrote:Show nested quote +On April 08 2013 11:10 sibs wrote:This was true for 2010 and 2011 but after all the terran early game nerfs and the queen buff on top of it terran had the worst early game rushes which was proven statistically. is this a troll post? Zerg has by far the worst early game cheeses, from 2010 to 2013, I don't see how this is even arguable, protoss doesn't have much either, I think terran has had best early game shenanigans throughout all the patches. i dno, post-patch, any combination of 2-base roach, ling, bane pressure against greedy terrans does a heck of a lot more damage than terran all-ins against greedy zergs who defend with 6 queens.
I don't think opening with 6 queens should be considered as greedy...
|
On April 08 2013 11:21 ETisME wrote:Show nested quote +On April 08 2013 11:11 opterown wrote:On April 08 2013 11:10 sibs wrote:This was true for 2010 and 2011 but after all the terran early game nerfs and the queen buff on top of it terran had the worst early game rushes which was proven statistically. is this a troll post? Zerg has by far the worst early game cheeses, from 2010 to 2013, I don't see how this is even arguable, protoss doesn't have much either, I think terran has had best early game shenanigans throughout all the patches. i dno, post-patch, any combination of 2-base roach, ling, bane pressure against greedy terrans does a heck of a lot more damage than terran all-ins against greedy zergs who defend with 6 queens. I don't think opening with 6 queens should be considered as greedy...
There's a difference between defend and open
|
Well 10 Marines is far less than 10 Widow Mines in terms of cost. And I bet if you take games where less than 10 Marines are built and compare them to games where more than 10 Marines are built, then you'd get pretty drastic results too.
What the heck, why not
![[image loading]](http://i.imgur.com/tUnq4OV.png)
In Masters TvT, dont make mass marine ?!
I agree with those who say that selection bias is a real thing. However selection bias doesn't negate the observations I made in OP.
|
GGtrackers is not the kind of sites where people put the replays on ? Considering it, people are more prone to put replay of their victories, as people plays more with WM now, it's logical that the winrate ratio is higher.
|
On April 08 2013 11:21 BronzeKnee wrote:Show nested quote +On April 08 2013 11:17 Entirety wrote:On April 08 2013 11:13 BronzeKnee wrote:On April 08 2013 11:06 opterown wrote:On April 08 2013 11:04 BronzeKnee wrote:On April 08 2013 11:00 opterown wrote:On April 08 2013 10:57 dsjoerg wrote:On April 08 2013 10:48 Emzeeshady wrote:
Also, we already knew Widow Mines are really good vs Zerg. This isn't all that informative. Well there ya go. OP is either wrong or obvious. Or both! In other news, here's the >10 mutas stats: well with the muta analysis, have we not reached a similar conclusion for ZvP? 49% with less than 10 mutas, 63% with more than 10 mutas for master level. the GM level is even greater (but with less sample size, of course). so we should also discuss how mutas are very strong in ZvP, stronger than widow mines are in TvZ nice stats, btw :D commend you on that! So, you just admitted that his analysis of Widow Mines was correct. And this also explains why every Zerg build Mutalisks in ZvP. However, the reasoning is pretty simple. In WoL, Storm would eat away at Mutalisk packs, while in HotS a single storm isn't crippling because they can move off and regen. Thus Protoss players are forced more and more into Phoenixes, which is similar to when a Protoss player forces a Zerg into Corrupters, you can switch quickly into ground units and just win. when did i say it was correct? i said: if you claim that widow mines are strong against zerg, i expect a similar claim for zvp. no such claim has come out. the fact that mutas, another tier 2 unit has even a greater differential in stats should be a bigger concern. therefore by following your line of reasoning, we need to nerf mutas before we nerf widow mines :D i commend his work and effort, not his analysis or conclusion I'm going try and explain this as simply as possible. Reading the OP, we find that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics. That is the analysis and the conclusion. No, you're trying to extrapolate the data when the data is shaky. Here's a better idea of what the data tells us: "In this particular sample, those Terrans who made more than 10 Widow Mines (for whatever reason, including surviving long enough to do so, facing army compositions which may have been weak against Widow Mines, etc.) won 59% of their games as opposed to Terrans who made less than 10 Widow Mines and won 49% of their games." What you said is exactly the same as what I said.You said "in this particular sample" I said: "based on GGtracker statistics" (GGtracker statistics is the sample) You said 59% and 49%, I said 10% higher. Sigh.
It seems that you don't understand the difference. Let me explain:
You said "When Terrans make 10 Widow Mines..." This is clearly extrapolating it to every other game. Essentially, it's saying whenever any Terran builds 10 Widow Mines, that Terran will achieve a 10% increase in his win rate. What I said is basically that the sample is not representative of the population.
When you extrapolate the data to the population, a 59% to 49% difference is not actually 10% - 10% is merely the average difference in win rate with an error bound applied to it.
|
For the next analysis, can we get how many times terran wins when making marines vs not making marines?
|
Quit arguing over silly stuff ~_~
You both (Bronze + Opter) agree about what the statistics signify (not much), not need to get all into semantics and whatnot.
Anyway, while I do think WM could use a change, I do not think it is severely overpowered. It needs a lot more time before any large changes to it should be made.
|
On April 08 2013 11:19 opterown wrote:Show nested quote +On April 08 2013 11:17 BronzeKnee wrote:On April 08 2013 11:14 opterown wrote:On April 08 2013 11:13 BronzeKnee wrote:On April 08 2013 11:06 opterown wrote:On April 08 2013 11:04 BronzeKnee wrote:On April 08 2013 11:00 opterown wrote:On April 08 2013 10:57 dsjoerg wrote:On April 08 2013 10:48 Emzeeshady wrote:
Also, we already knew Widow Mines are really good vs Zerg. This isn't all that informative. Well there ya go. OP is either wrong or obvious. Or both! In other news, here's the >10 mutas stats: well with the muta analysis, have we not reached a similar conclusion for ZvP? 49% with less than 10 mutas, 63% with more than 10 mutas for master level. the GM level is even greater (but with less sample size, of course). so we should also discuss how mutas are very strong in ZvP, stronger than widow mines are in TvZ nice stats, btw :D commend you on that! So, you just admitted that his analysis of Widow Mines was correct. And this also explains why every Zerg build Mutalisks in ZvP. However, the reasoning is pretty simple. In WoL, Storm would eat away at Mutalisk packs, while in HotS a single storm isn't crippling because they can move off and regen. Thus Protoss players are forced more and more into Phoenixes, which is similar to when a Protoss player forces a Zerg into Corrupters, you can switch quickly into ground units and just win. when did i say it was correct? i said: if you claim that widow mines are strong against zerg, i expect a similar claim for zvp. no such claim has come out. the fact that mutas, another tier 2 unit has even a greater differential in stats should be a bigger concern. therefore by following your line of reasoning, we need to nerf mutas before we nerf widow mines :D i commend his work and effort, not his analysis or conclusion I'm going try and explain this as simply as possible. Reading the OP, we find that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics. That is the analysis and the conclusion. Though the OP states "Of course this doesn't exactly prove that Widow Mines are OP, but at the very least it suggests that getting/preventing mass widow mines is an important part of TvZ" It basically says nothing. So now you state: On April 08 2013 10:30 opterown wrote:
i bet if you compared games with over/under ten mutalisks, or over/under ten high templar, etc you would get similar stats And then someone states such statistics. You commend the work and effort. By commending the work and effort, you are basically saying "hey look you did a good job collecting statistics" and that is opposed to doing a bad job, in which the statistics are basically wrong. Thus, since the analysis and conclusion were based entirely on the collection of statistics and he makes no inference on whether or not Widow Mines are OP, and then you commend the very same process of statistical collection, you have agreed to the conclusion of the OP which is: "that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics." cool, so you get some winrates. i'm not attacking those numbers, i'm saying the conclusions that a lot of people are going to make are invalid. EXACTLY! Thank you! But remember this, those conclusions that argue that Widow Mines are not overpowered now have to be independent of the statistic presented! Otherwise they are invalid! and whenever did i say the statistics were wrong? i claimed that a) analysis of over/under 10 of other units would get similar stats b) you cannot conclude anything significant from this so i don't quite get why you are specifically talking to me about this
I caught you in an logical inconsistency and I'm pressing it to prove my point, that statistics don't lie, only those who abuse them.
You stated the conclusion of the OP was wrong "when did i say it was correct?... i commend his work and effort, not his analysis or conclusion", but the conclusion was the statistics.
Then you backtracked "i'm not attacking those numbers, i'm saying the conclusions that a lot of people are going to make are invalid"
As I stated, the numbers were the statistics, which was the conclusion. Then you say that other people are going to make invalid conclusions, which is exactly what I'm fighting against.
|
opterown
Australia54784 Posts
On April 08 2013 11:21 ETisME wrote:Show nested quote +On April 08 2013 11:11 opterown wrote:On April 08 2013 11:10 sibs wrote:This was true for 2010 and 2011 but after all the terran early game nerfs and the queen buff on top of it terran had the worst early game rushes which was proven statistically. is this a troll post? Zerg has by far the worst early game cheeses, from 2010 to 2013, I don't see how this is even arguable, protoss doesn't have much either, I think terran has had best early game shenanigans throughout all the patches. i dno, post-patch, any combination of 2-base roach, ling, bane pressure against greedy terrans does a heck of a lot more damage than terran all-ins against greedy zergs who defend with 6 queens. I don't think opening with 6 queens should be considered as greedy...
hatch first into pool into no gas and a third base, then six queens off three base is pretty greedy :D
On April 08 2013 11:25 BronzeKnee wrote:Show nested quote +On April 08 2013 11:19 opterown wrote:On April 08 2013 11:17 BronzeKnee wrote:On April 08 2013 11:14 opterown wrote:On April 08 2013 11:13 BronzeKnee wrote:On April 08 2013 11:06 opterown wrote:On April 08 2013 11:04 BronzeKnee wrote:On April 08 2013 11:00 opterown wrote:On April 08 2013 10:57 dsjoerg wrote:On April 08 2013 10:48 Emzeeshady wrote:
Also, we already knew Widow Mines are really good vs Zerg. This isn't all that informative. Well there ya go. OP is either wrong or obvious. Or both! In other news, here's the >10 mutas stats: well with the muta analysis, have we not reached a similar conclusion for ZvP? 49% with less than 10 mutas, 63% with more than 10 mutas for master level. the GM level is even greater (but with less sample size, of course). so we should also discuss how mutas are very strong in ZvP, stronger than widow mines are in TvZ nice stats, btw :D commend you on that! So, you just admitted that his analysis of Widow Mines was correct. And this also explains why every Zerg build Mutalisks in ZvP. However, the reasoning is pretty simple. In WoL, Storm would eat away at Mutalisk packs, while in HotS a single storm isn't crippling because they can move off and regen. Thus Protoss players are forced more and more into Phoenixes, which is similar to when a Protoss player forces a Zerg into Corrupters, you can switch quickly into ground units and just win. when did i say it was correct? i said: if you claim that widow mines are strong against zerg, i expect a similar claim for zvp. no such claim has come out. the fact that mutas, another tier 2 unit has even a greater differential in stats should be a bigger concern. therefore by following your line of reasoning, we need to nerf mutas before we nerf widow mines :D i commend his work and effort, not his analysis or conclusion I'm going try and explain this as simply as possible. Reading the OP, we find that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics. That is the analysis and the conclusion. Though the OP states "Of course this doesn't exactly prove that Widow Mines are OP, but at the very least it suggests that getting/preventing mass widow mines is an important part of TvZ" It basically says nothing. So now you state: On April 08 2013 10:30 opterown wrote:
i bet if you compared games with over/under ten mutalisks, or over/under ten high templar, etc you would get similar stats And then someone states such statistics. You commend the work and effort. By commending the work and effort, you are basically saying "hey look you did a good job collecting statistics" and that is opposed to doing a bad job, in which the statistics are basically wrong. Thus, since the analysis and conclusion were based entirely on the collection of statistics and he makes no inference on whether or not Widow Mines are OP, and then you commend the very same process of statistical collection, you have agreed to the conclusion of the OP which is: "that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics." cool, so you get some winrates. i'm not attacking those numbers, i'm saying the conclusions that a lot of people are going to make are invalid. EXACTLY! Thank you! But remember this, those conclusions that argue that Widow Mines are not overpowered now have to be independent of the statistic presented! Otherwise they are invalid! and whenever did i say the statistics were wrong? i claimed that a) analysis of over/under 10 of other units would get similar stats b) you cannot conclude anything significant from this so i don't quite get why you are specifically talking to me about this I caught you in an logical inconsistency and I'm pressing it to prove my point, that statistics don't lie, only those who abuse them. You stated the conclusion of the OP was wrong "when did i say it was correct?... i commend his work and effort, not his analysis or conclusion", but the conclusion was the statistics. Now your trying to backtrack.
no, i never said it was wrong or right, go read the posts. "when did i say it was correct" does not mean "it is wrong"
|
On April 08 2013 11:25 BronzeKnee wrote:Show nested quote +On April 08 2013 11:19 opterown wrote:On April 08 2013 11:17 BronzeKnee wrote:On April 08 2013 11:14 opterown wrote:On April 08 2013 11:13 BronzeKnee wrote:On April 08 2013 11:06 opterown wrote:On April 08 2013 11:04 BronzeKnee wrote:On April 08 2013 11:00 opterown wrote:On April 08 2013 10:57 dsjoerg wrote:On April 08 2013 10:48 Emzeeshady wrote:
Also, we already knew Widow Mines are really good vs Zerg. This isn't all that informative. Well there ya go. OP is either wrong or obvious. Or both! In other news, here's the >10 mutas stats: well with the muta analysis, have we not reached a similar conclusion for ZvP? 49% with less than 10 mutas, 63% with more than 10 mutas for master level. the GM level is even greater (but with less sample size, of course). so we should also discuss how mutas are very strong in ZvP, stronger than widow mines are in TvZ nice stats, btw :D commend you on that! So, you just admitted that his analysis of Widow Mines was correct. And this also explains why every Zerg build Mutalisks in ZvP. However, the reasoning is pretty simple. In WoL, Storm would eat away at Mutalisk packs, while in HotS a single storm isn't crippling because they can move off and regen. Thus Protoss players are forced more and more into Phoenixes, which is similar to when a Protoss player forces a Zerg into Corrupters, you can switch quickly into ground units and just win. when did i say it was correct? i said: if you claim that widow mines are strong against zerg, i expect a similar claim for zvp. no such claim has come out. the fact that mutas, another tier 2 unit has even a greater differential in stats should be a bigger concern. therefore by following your line of reasoning, we need to nerf mutas before we nerf widow mines :D i commend his work and effort, not his analysis or conclusion I'm going try and explain this as simply as possible. Reading the OP, we find that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics. That is the analysis and the conclusion. Though the OP states "Of course this doesn't exactly prove that Widow Mines are OP, but at the very least it suggests that getting/preventing mass widow mines is an important part of TvZ" It basically says nothing. So now you state: On April 08 2013 10:30 opterown wrote:
i bet if you compared games with over/under ten mutalisks, or over/under ten high templar, etc you would get similar stats And then someone states such statistics. You commend the work and effort. By commending the work and effort, you are basically saying "hey look you did a good job collecting statistics" and that is opposed to doing a bad job, in which the statistics are basically wrong. Thus, since the analysis and conclusion were based entirely on the collection of statistics and he makes no inference on whether or not Widow Mines are OP, and then you commend the very same process of statistical collection, you have agreed to the conclusion of the OP which is: "that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics." cool, so you get some winrates. i'm not attacking those numbers, i'm saying the conclusions that a lot of people are going to make are invalid. EXACTLY! Thank you! But remember this, those conclusions that argue that Widow Mines are not overpowered now have to be independent of the statistic presented! Otherwise they are invalid! and whenever did i say the statistics were wrong? i claimed that a) analysis of over/under 10 of other units would get similar stats b) you cannot conclude anything significant from this so i don't quite get why you are specifically talking to me about this I caught you in an logical inconsistency and I'm pressing it to prove my point, that statistics don't lie, only those who abuse them. You stated the conclusion of the OP was wrong "when did i say it was correct?... i commend his work and effort, not his analysis or conclusion", but the conclusion was the statistics. Now your trying to backtrack.
A 4 year old spends 2 days fingerpainting on the ceiling
I can commend his effort fingerpainting-it doesn't mean the painting is beautiful
|
On April 08 2013 11:26 GTPGlitch wrote:Show nested quote +On April 08 2013 11:25 BronzeKnee wrote:On April 08 2013 11:19 opterown wrote:On April 08 2013 11:17 BronzeKnee wrote:On April 08 2013 11:14 opterown wrote:On April 08 2013 11:13 BronzeKnee wrote:On April 08 2013 11:06 opterown wrote:On April 08 2013 11:04 BronzeKnee wrote:On April 08 2013 11:00 opterown wrote:On April 08 2013 10:57 dsjoerg wrote:[quote] Well there ya go. OP is either wrong or obvious. Or both! In other news, here's the >10 mutas stats: well with the muta analysis, have we not reached a similar conclusion for ZvP? 49% with less than 10 mutas, 63% with more than 10 mutas for master level. the GM level is even greater (but with less sample size, of course). so we should also discuss how mutas are very strong in ZvP, stronger than widow mines are in TvZ nice stats, btw :D commend you on that! So, you just admitted that his analysis of Widow Mines was correct. And this also explains why every Zerg build Mutalisks in ZvP. However, the reasoning is pretty simple. In WoL, Storm would eat away at Mutalisk packs, while in HotS a single storm isn't crippling because they can move off and regen. Thus Protoss players are forced more and more into Phoenixes, which is similar to when a Protoss player forces a Zerg into Corrupters, you can switch quickly into ground units and just win. when did i say it was correct? i said: if you claim that widow mines are strong against zerg, i expect a similar claim for zvp. no such claim has come out. the fact that mutas, another tier 2 unit has even a greater differential in stats should be a bigger concern. therefore by following your line of reasoning, we need to nerf mutas before we nerf widow mines :D i commend his work and effort, not his analysis or conclusion I'm going try and explain this as simply as possible. Reading the OP, we find that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics. That is the analysis and the conclusion. Though the OP states "Of course this doesn't exactly prove that Widow Mines are OP, but at the very least it suggests that getting/preventing mass widow mines is an important part of TvZ" It basically says nothing. So now you state: On April 08 2013 10:30 opterown wrote:
i bet if you compared games with over/under ten mutalisks, or over/under ten high templar, etc you would get similar stats And then someone states such statistics. You commend the work and effort. By commending the work and effort, you are basically saying "hey look you did a good job collecting statistics" and that is opposed to doing a bad job, in which the statistics are basically wrong. Thus, since the analysis and conclusion were based entirely on the collection of statistics and he makes no inference on whether or not Widow Mines are OP, and then you commend the very same process of statistical collection, you have agreed to the conclusion of the OP which is: "that people who make 10+ Widow Mines in TvZ have a 10% high winrate based on GGtracker statistics." cool, so you get some winrates. i'm not attacking those numbers, i'm saying the conclusions that a lot of people are going to make are invalid. EXACTLY! Thank you! But remember this, those conclusions that argue that Widow Mines are not overpowered now have to be independent of the statistic presented! Otherwise they are invalid! and whenever did i say the statistics were wrong? i claimed that a) analysis of over/under 10 of other units would get similar stats b) you cannot conclude anything significant from this so i don't quite get why you are specifically talking to me about this I caught you in an logical inconsistency and I'm pressing it to prove my point, that statistics don't lie, only those who abuse them. You stated the conclusion of the OP was wrong "when did i say it was correct?... i commend his work and effort, not his analysis or conclusion", but the conclusion was the statistics. Now your trying to backtrack. A 4 year old spends 2 days fingerpainting on the ceiling I can commend his effort fingerpainting-it doesn't mean the painting is beautiful
Haha, except when you first say someone elses fingerpainting that is exactly the same is beautiful... you obviously didn't read the whole exchange between us.
|
but at the very least it suggests that getting/preventing mass widow mines is an important part of TvZ.
Well I even disagree with that statement so...
"at the very least it suggests that getting/preventing building of a spawning pool is an important part of TvZ."........yeah doesn't seem so smart. I agree with opterown, you can say the same shit for every goddamn unit in the game.
|
On April 08 2013 10:57 dsjoerg wrote:Show nested quote +On April 08 2013 10:48 Emzeeshady wrote:
Also, we already knew Widow Mines are really good vs Zerg. This isn't all that informative. Well there ya go. OP is either wrong or obvious. Or both! In other news, here's the >10 mutas stats: ![[image loading]](http://i.imgur.com/SZR9gQa.png) Most surprising to me here is Master ZvT, where >10 mutas doesn't seem to make much of a winrate difference. To the naysayers, would it be of any use to restrict the analysis to armies over a certain size and at a certain time? I can retrieve TvZ games where T's army size was at least ABC, and then we can split it by <10 WM and >10 WM. I have no stake in the outcome here btw, I play Protoss Just trying to learn the game from a numerical perspective, so as to avoid arguments that ultimately boil down to an Appeal to Authority.
I want to see these numbers for swarm hosts and then I can know which I should be going in ZvP :p
While it is obvious that the terrans are just ultra-defensive because it happened to be about widow mines and ZvT (guilt?), they are right that this doesn't show anything about balance really. But it is actually interesting to see these kind of statistics against many different units. Rather than having similar rates for "all" tier 2 units, in actuality what you will find I think is that some unit compositions are stronger than others. While it is "obvious," as someone said, that this will be the case, what is interesting is to see statistics on what those compositions are. It is kind of useful information, I think moreso if it had some reasoning behind the units and the quantity chosen. For example, in addition to mutas vs swarm hosts, I'd be fascinated to see numbers on roaches vs zerglings, or ultras vs broodlords.
|
|
|
|
|
|
EPT
