|
On May 22 2013 05:00 sitromit wrote:Show nested quote +On May 22 2013 04:53 tenklavir wrote:On May 22 2013 04:50 sitromit wrote:On May 22 2013 04:45 DemigodcelpH wrote:On May 22 2013 04:41 sitromit wrote:
So this article essentially takes Idra and a bunch of random masters players as the high benchmark of skill, and claims based on raw mathematical analysis of their inject rates across random games played out in very different scenarios, that injecting is not very important.
Why I see absolutely no flaw in that aproach at all. It's 44,000 games. Don't be dismissive because you don't personally agree with the results. 44K random games. One may be 5 minutes. The other may be a 40 minute game where the player pulled all his Queens off the hatcheries in the last 15 minutes to transfuse his Ultras. And these are random ladder players, not exactly the high benchmark of skill, and frankly, neither is Idra. You realize random sampling is precisely what you want, right? No, with the number of things that can happen in a game that can affect why a player may or may not be injecting, random is exactly what you DON'T want. Say a player is under attack and transfuses a spine crawler instead of injecting. If you just analyze that mathematically, you come up with the conclusion that that player has less than perfect injects, when in fact he made a calculated sacrifice. It doesn't mean inject is not important, or that it's enough to inject with 60% efficiency and any better makes no difference. The number one rule in doing research that is meaningful in any way, is to control and reduce the number of variables.
So what you're saying is that there are a whole bunch of reasons that you wouldn't want to inject? Or in other words, that inject isn't as high of a priority for high level play as people make it out to be? You do realize that's exactly what the OP is saying right?
|
In a long macro game, inject just becomes much less important. Gas is the number one issue for Zerg to macro and the rest of the larva unused can go into cheap unit lime lings.
That's why we have the fast ultra queen build in zvp. The injects aren't so important and energy for creep and transfused becomes far more important.
Not to mention late game is much less about max but it's about maxing with the right army composition. More larva can't help you get more hive tech units out. Thus larva less important, inject less important etc
|
A lot of people in here who pretend to understand stats but don't, and drawing nonsense (and common sense) conclusions from the data.
Bottom line, this stuff is interesting because it shows a common community perception, that injects separate the men from the boys, is largely imagined.
|
On May 22 2013 03:45 dsjoerg wrote:Many in the community think that consistently keeping your Hatches injected with Larva is an important skill for any Zerg player to have. For example, here are some recent discussion threads on allthingszerg. . . . If the conventional wisdom is right, and consistent injects are an important skill, then we would expect to find that lower-league players would have trouble keeping their hatcheries consistently injected, and that higher-league Zergs would keep their hatcheries injected a notably higher percent of the time.
On May 22 2013 12:20 aksfjh wrote:
A lot of people in here who pretend to understand stats but don't, and drawing nonsense (and common sense) conclusions from the data.
Bottom line, this stuff is interesting because it shows a common community perception, that injects separate the men from the boys, is largely imagined.
The OP analysis is flawed because of the "silver, master? doesn't matter" angle. In a silver vs silver game, a silver player will be able to hit their early/mid game injects. I guarantee that if you put a silver vs a master, that silver is not going to be hitting their injects in the early/mid game. Thus, the perception of "injects separate the men from the boys" IS correct, because the better players will continue to inject even while being harassed and/or when their attention is being taxed in combat/elsewhere. A lesser player will fall apart and forget to inject, which is 100% crucial in the early/mid game.
Injects DO matter. The better players continue to inject vs equally skilled opponents. This data may show something, but it does not show that injects don't matter.
(EDIT: Also, of course, masters+ will be having more hatcheries to inject sooner than a silver player, and the data doesn't seem to take that into account.)
|
idk, Nestea says that if your queens get over 25 energy you lose game. i believe Nestea.
|
The analysis is good, it's always cool to takes huge datas and compute stuff.
But the problem is to know if it's really meaningfull
That means[1] that about 30% of Masters Zerg games have an inject % that’s worse than the average Silver inject %. They did worse than the average Silver game, yet somehow they are in Masters with sub-Silver injects.
For exemple this quote is quite missleading, because no a silver game don't look at all like a master game. If you put the "sub-injects" master player in the "godly-inject" silver player game, you will probably see that he magically hit better injects.
And as zerg is mainly a defensive race, you don't really choice when "action occurs", and don't always have the time to hit your injects. That's why the "parade push" is really annying for a zerg, because it requires constant attention, and makes it harder to hit good injects.
But I think what is to remember from the study is that as you play better, you don't "hit more injects", but more "hit as many injects while doing a lot of other stuff".
I think an interesting study to do now would be "what does the player does while he doesn't inject ?", to see if it's pretty much the same things across all league.
|
I'm better at hitting injects than I am at spending larva. Though 4 larva pop off per inject, I doubt I'm actually -gaining- 4 larva per inject, because I'll leave them idle long enough (especially on 3-4 bases) for the natural spawn of larva to have been delayed enough for the injects to have hardly mattered.
|
On May 22 2013 12:20 aksfjh wrote:
A lot of people in here who pretend to understand stats but don't, and drawing nonsense (and common sense) conclusions from the data.
Bottom line, this stuff is interesting because it shows a common community perception, that injects separate the men from the boys, is largely imagined.
It's mostly just Zerg mains being illegitimately dismissive out of emotional response. We're all open to good arguments, but generally the most people are doing here is throwing out random anecdotes and other irrelevant things.
|
One thing to consider, if someone were near perfect, would it benefit them. It'd be interesting to see someone like a DRG, Life or w/e inject %
|
On May 22 2013 11:21 IcedBacon wrote:
Yeah you really can't get much information from data like this. So many other things to take into consideration such as the number of bases, macro hatches, queens being used for other purposes. A masters player is far more likely to use a transfuse during a rush or a creep tumor early on, etc. But of course people are just going to read the summary and instantly cry about how Zerg takes no skill or something.
this would be a valid point if I could make 6 colossus at a time
|
You're missing a key reason for why Master players have less injects at the beginning of the game: They're using their queens to fend off harassment.
Especially in ZvT, Queens are crucial for spreading creep and defending hellions, but also in ZvP and ZvZ many early games are focused on Queen defence.
If you REALLY want to see the difference in macro mechanics at various levels look at average Queen energy. Right now you're treating laying creep tumours like a macro mistake.
|
On May 22 2013 13:34 _Search_ wrote:
You're missing a key reason for why Master players have less injects at the beginning of the game: They're using their queens to fend off harassment.
Especially in ZvT, Queens are crucial for spreading creep and defending hellions, but also in ZvP and ZvZ many early games are focused on Queen defence.
If you REALLY want to see the difference in macro mechanics at various levels look at average Queen energy. Right now you're treating laying creep tumours like a macro mistake.
Generally speaking, you use extra queens on creep spread, not queens that usually inject. The only exclusion might be ZvP where you spare 1 inject to get a creep tumor down for your fast third.
|
On May 22 2013 13:34 _Search_ wrote:
You're missing a key reason for why Master players have less injects at the beginning of the game: They're using their queens to fend off harassment.
Especially in ZvT, Queens are crucial for spreading creep and defending hellions, but also in ZvP and ZvZ many early games are focused on Queen defence.
If you REALLY want to see the difference in macro mechanics at various levels look at average Queen energy. Right now you're treating laying creep tumours like a macro mistake.
Low level players have to defend hellions too; if anything low level players are even more vulnerable to harassment and will spend more time defending it. Futhermore, the article isn't measuring any kind of mistake or treating exceptions as mistakes as these exceptions occur at every level while being normalized by a sample-size of data that is by all means statistically representative of the entire population.
|
On May 22 2013 13:41 DemigodcelpH wrote:Show nested quote +On May 22 2013 13:34 _Search_ wrote:
You're missing a key reason for why Master players have less injects at the beginning of the game: They're using their queens to fend off harassment.
Especially in ZvT, Queens are crucial for spreading creep and defending hellions, but also in ZvP and ZvZ many early games are focused on Queen defence.
If you REALLY want to see the difference in macro mechanics at various levels look at average Queen energy. Right now you're treating laying creep tumours like a macro mistake. Low level players have to defend hellions too; if anything low level players are even more vulnerable to harassment and will spend more time defending it. Futhermore, the article isn't measuring any kind of mistake or treating exceptions as mistakes as these exceptions occur at every level while being normalized by a sample-size of data that is by all means statistically representative of the entire population.
p-p-p-p-p-pow! Holy Moley Batman! That was so bad, your momma made a mom joke about your momma.
User was warned for this post
|
On May 22 2013 11:04 DemigodcelpH wrote:Show nested quote +On May 22 2013 06:17 Eventine wrote:
I swear, every time someone provides some data point, instead of sparking interesting debate, half the responses are like your data is wrong, my intuition is perfect and therefore i reject your findings. This. I find it amusing how every Zerg (well Zerg mains because I do play Zerg) player suddenly attempted to become a statistics major. Quite convenient, eh? To quote someone else: Show nested quote +A while back, someone collected similar data for workers-produced, time-supply-blocked, and surplus-resources-banked. Their results showed vast differences between the different divisions, implying that the above metrics are a major differentiating factor between the different divisions & skill levels. The empirical approach does work. The above metrics are also subjected to the same "real active game world" scrutiny in the same exact way, and yet they scale non-marginally with skill level.
I find it amusing that you assume that people not well trained in statistics. Statistics are not perfect predictors, but they are good tools to understand what is happening. As my prof says, all models are wrong, some are useful.
|
On May 22 2013 14:00 Eventine wrote:Show nested quote +On May 22 2013 11:04 DemigodcelpH wrote:On May 22 2013 06:17 Eventine wrote:
I swear,every time someonew ides some data point, instead of sparking interesting debate, half the responses are like your data is wrong, my intuition is perfect and therefore i reject your findings. This. I find it amusing how every Zerg (well Zerg mains because I do play Zerg) player suddenly attempted to become a statistics major. Quite convenient, eh? To quote someone else: A while back, someone collected similar data for workers-produced, time-supply-blocked, and surplus-resources-banked. Their results showed vast differences between the different divisions, implying that the above metrics are a major differentiating factor between the different divisions & skill levels. The empirical approach does work. The above metrics are also subjected to the same "real active game world" scrutiny in the same exact way, and yet they scale non-marginally with skill level. I find it amusing that you assume that people not well trained in statistics. Statistics are not perfect predictors, but they are good tools to understand what is happening. As my prof says, all models are wrong, some are useful.
Its much safer to assume people don't understand stats, I have a major in econometric and stats is extremely complicated.
A wrong test can lead to a very wrong conclusion and mislead the readers.
For example in this report they mentioned they might do a test on win ratio and inject ratio. This is already gonna be one huge mistake because they will be omitting so many important variables that the importance of inject will be significantly enlarged.
There is a reason why there are so many specialised statistic units out for different majors like econometric and psychometric. If basic level static is enough to cover complicated matters like a starcraft 2 game or real world business simulation, then we won't need advanced stats course for these majors.
|
The conclusion set by this statistical analysis is singular: larva injections become less paramount as the game progresses due to unit-larva costs. The statistics is somewhat sound, although the Silver-Master statement is very misleading and discrediting; I can't quite put my finger on it, but there's some statistic that I think is omitted that is crucial (though I am typing this at an inopportune time, so I could just be loopy). There is nothing in this analysis that speaks to the skill difficulty of managing the zerg macromechanics, or indeed zerg mechanics as a whole, compared to other races. All this analysis does is do away with a myth, a preconceived notion that seems intuitive, but in reality is not; why that is, be it harassment, creep spread or another queen distraction or utility, is a different argument for a different time.
|
On May 22 2013 12:27 Jinky wrote:Show nested quote +On May 22 2013 03:45 dsjoerg wrote:Many in the community think that consistently keeping your Hatches injected with Larva is an important skill for any Zerg player to have. For example, here are some recent discussion threads on allthingszerg. . . . If the conventional wisdom is right, and consistent injects are an important skill, then we would expect to find that lower-league players would have trouble keeping their hatcheries consistently injected, and that higher-league Zergs would keep their hatcheries injected a notably higher percent of the time. Show nested quote +On May 22 2013 12:20 aksfjh wrote:
A lot of people in here who pretend to understand stats but don't, and drawing nonsense (and common sense) conclusions from the data.
Bottom line, this stuff is interesting because it shows a common community perception, that injects separate the men from the boys, is largely imagined. The OP analysis is flawed because of the "silver, master? doesn't matter" angle. In a silver vs silver game, a silver player will be able to hit their early/mid game injects. I guarantee that if you put a silver vs a master, that silver is not going to be hitting their injects in the early/mid game. Thus, the perception of "injects separate the men from the boys" IS correct, because the better players will continue to inject even while being harassed and/or when their attention is being taxed in combat/elsewhere. A lesser player will fall apart and forget to inject, which is 100% crucial in the early/mid game. Injects DO matter. The better players continue to inject vs equally skilled opponents. This data may show something, but it does not show that injects don't matter. (EDIT: Also, of course, masters+ will be having more hatcheries to inject sooner than a silver player, and the data doesn't seem to take that into account.)
You can't guarantee that through the data, just your own hunches. For the data we have, we only know that lower level players hit their injects with roughly the same consistency as higher levels. I'm also not saying that injects don't matter, but that there isn't a clear distinction between them and levels of play.
|
In the later stages of the game, good players will generally be on many bases with macro hatches, but won't have queens at them all because of the supply sink of the queens. This is probably the biggest reason why inject percentage is the way it is for higher level players. It has nothing to do with zerg macro being easy, as stated more than once in this thread. So please, before you try to bash zerg players just because you're angry, think of a better response than "SEE, zerg macro EZ durrrr".
|
On May 22 2013 14:27 knOxStarcraft wrote:
In the later stages of the game, good players will generally be on many bases with macro hatches, but won't have queens at them all because of the supply sink of the queens. This is probably the biggest reason why inject percentage is the way it is for higher level players. It has nothing to do with zerg macro being easy, as stated more than once in this thread. So please, before you try to bash zerg players just because you're angry, think of a better response than "SEE, zerg macro EZ durrrr".
Especially when terran macro is the same lategame.
While it asks a lot early game to have constant scv production and good build orders without supply block, once on 3 bases, a terran can just watch his army and control it while producing 17 marines per 17 marines when he loses parts of his army.
Would love to see barracks usage (if they are doing nothing at certain stage of the game) statistics.
|
|
|
|
|
|
EPT
