|
On October 11 2011 21:35 DarQraven wrote:
See, this is why I often disagree with the advice to "only work on your macro, then you'll get to diamond really quickly".
One week ago, I was still Gold. My SQ from that period, averaged over 24 games is 63, which should put me inbetween platinum and diamond. I can't look at the history for season 2 anymore, but I hardly improved my macro lately. In Season 2 I was Silver, and if I had to estimate, my SQ would have been just below the plat level. (I checked unspent resources and income rate then, as well, just divided one by the other to get an estimate of how well I did).
It's anecdotal, sure, but it does show that there is slightly more to getting better at the game than straight macro.
Unless you got that kind of SQ in very long macro games then it doesn't matter. SQ is absolute garbage because it ranks short easy macro games very highly. Every low leaguer can get 80+ SQ by doing a 2 base all in. While a better player might get a lower SQ because he played a long macro game that doesn't use an optimized all in build. It's a very poor way of displaying macro skill.
|
On October 10 2011 06:09 whatthefat wrote:Show nested quote +On October 08 2011 08:30 habermas wrote:I've had a bit of spare time at work so I created a little app to help keep the track on SQ. I made it mainly for personal use, but I thought you may find it useful. It calculates an average SQ based on the data you give to it and also shows some graphs which you can use to see your progress. + Show Spoiler +It uses a CSV file to store the data, so you can open it in Excel if you wish so. You can also create an Excel file and then open it in the app as long as you keep the following format: first row: date, i, u, sqand then following rows with the data. The sq column can be left empty and it will be calculated by the program on the file load. It is kind of a messy code and I've written it under linux, so if you find any major bugs, let me know. Here's a standalone executable with all the dependencies included. Just run sq.exe: http://www.mediafire.com/?t0d504rmt6ma0ycAnd if you're a developer or if you're suspicious to run random .exe files from the Internet here is the source code: http://www.mediafire.com/?3g9fkth38di21a6However to run this one you're going to have Python installed on your system, as well as matplotlib and wx python libraries. This is a great app! I just tried it out for my last week of games, and it's very cool. I'm not sure if the save/load function is working though. I tried entering data, saving, exiting, then re-opening and loading, and it didn't seem to add the previous data. The command window showed this error: + Show Spoiler +Traceback (most recent call last):
File "sq.py", line 198, in onOpen
File "sq.py", line 177, in readFile
File "sq.py", line 154, in setTimeFrame
ValueError: min() arg is an empty sequence Am I doing something wrong?
Yeah, as I said it's a bit of a mess. I'll look into it when I have a time, but I also noticed that the graph won't display unless you save and then load a file. You can check if your data has been saved by opening the file you specified in notepad, it's a simple text format.
On October 11 2011 14:37 Malazin wrote:Thanks for the write up whatthefat, I have found that by focusing on my macro, I've improved my SQ and ranked up to Diamond from Platinum. My SQ has improved from the high 60s to the mid 70s so I think you're about bang on! In addition, I made an application to track your SQ automatically, a bit like the one posted above. However, if you're on 1920x1080, you can auto-track your SQ simply by taking a screen shot of your stats at the end of the game; no tabbing needed. The UI is shown below: ![[image loading]](http://i.imgur.com/ZxMu4.jpg) The reddit thread with all the info can be found here and the download can be found here
How did you get it to parse the text from screenshots? I was thinking about it but couldn't find any modules to do it for me so didn't bother because of the lack of time.
|
On October 14 2011 06:45 Hulkoff wrote:Show nested quote +On October 11 2011 21:35 DarQraven wrote:
See, this is why I often disagree with the advice to "only work on your macro, then you'll get to diamond really quickly".
One week ago, I was still Gold. My SQ from that period, averaged over 24 games is 63, which should put me inbetween platinum and diamond. I can't look at the history for season 2 anymore, but I hardly improved my macro lately. In Season 2 I was Silver, and if I had to estimate, my SQ would have been just below the plat level. (I checked unspent resources and income rate then, as well, just divided one by the other to get an estimate of how well I did).
It's anecdotal, sure, but it does show that there is slightly more to getting better at the game than straight macro. Unless you got that kind of SQ in very long macro games then it doesn't matter. SQ is absolute garbage because it ranks short easy macro games very highly. Every low leaguer can get 80+ SQ by doing a 2 base all in. While a better player might get a lower SQ because he played a long macro game that doesn't use an optimized all in build. It's a very poor way of displaying macro skill.
SQ takes average income into account, and it shows no significant positive or negative correlation with game duration in any league. Your assertion that "every low leaguer can get 80+ SQ by doing a 2 base all in" is just completely untrue. As is your assertion that "a better player might get a lower SQ because he played a long macro game that doesn't use an optimized all in build". Neither of these are borne out by the data; unless you have collected a similarly large dataset, I don't know what you are basing these claims on.
All SQ does is compare your spending efficiency in a game with that of all other players' games with the same average resource collection rate. It doesn't matter whether the game is a short all-in or a long macro game, because it normalizes for that. In other words, the average difference in SQ between Bronze and Diamond is the same regardless of game duration or income.
|
On October 14 2011 05:15 AmericanUmlaut wrote:
Just wanted to say thanks, whatthefat. I grabbed one of the spreadsheets and started keeping track of my SQ every game, and just realizing what I was doing differently in the games where it was high has made me play better to a remarkable degree. Good work, sir!
I'm really glad to hear that! I have found it to be a useful tool for self-improvement myself, so it's nice to hear that it's also useful to others in the community. I'm hoping some of these tools for automated collection will make it even easier for people to collect their own data. Ideally, Blizzard will add the score screen data to the replay file!
|
wow thanks for the calculator!
|
People in lower leagues must realize that high SQ is NOT equivalent to good macro . . . it is a very small though important part of good macro: spending your money.
Good macro includes good use of macro mechanics, good worker production, good powering, and good grasp of economic timings (not getting supply blocked); none of which are reflected in SQ.
|
On October 14 2011 14:50 chenchen wrote:
People in lower leagues must realize that high SQ is NOT equivalent to good macro . . . it is a very small though important part of good macro: spending your money.
Good macro includes good use of macro mechanics, good worker production, good powering, and good grasp of economic timings (not getting supply blocked); none of which are reflected in SQ.
Absolutely. That's hopefully clear to anyone who reads the whole OP, since it is stated explicitly. Having said that, so long as you're not doing silly things to game the SQ, I think it is a useful quantitative metric to track whether your macro is improving (in fact, it seems to be the only such metric).
|
this is so sexy. lost me at around SQ, but then again i havent done math in years and i kind of skimmed it XD. GJ tho, very interesting stuff
|
On October 14 2011 14:50 chenchen wrote:
People in lower leagues must realize that high SQ is NOT equivalent to good macro . . . it is a very small though important part of good macro: spending your money.
Good macro includes good use of macro mechanics, good worker production, good powering, and good grasp of economic timings (not getting supply blocked); none of which are reflected in SQ.
I don't think anyone is arguing that SQ is the only useful metric for measuring macro, only that it's a useful piece of information that correlates well with overall performance and is easy to calculate for every game you play.
I don't understand how you think that not getting supply blocked isn't reflected in SQ, though; there's usually a spike in unspent resources associated with getting supply blocked.
|
On October 14 2011 15:03 AmericanUmlaut wrote:Show nested quote +On October 14 2011 14:50 chenchen wrote:
People in lower leagues must realize that high SQ is NOT equivalent to good macro . . . it is a very small though important part of good macro: spending your money.
Good macro includes good use of macro mechanics, good worker production, good powering, and good grasp of economic timings (not getting supply blocked); none of which are reflected in SQ. I don't think anyone is arguing that SQ is the only useful metric for measuring macro, only that it's a useful piece of information that correlates well with overall performance and is easy to calculate for every game you play. I don't understand how you think that not getting supply blocked isn't reflected in SQ, though; there's usually a spike in unspent resources associated with getting supply blocked.
Not true.
You can just plop down a bunch of production or queue up a bunch of units. None of these are ways to efficiently spend money, but money is being spent nonetheless.
I'm just pointing out how little SQ means in response to many of the lower level players saying "oh my SQ is high, why am I stuck in [non-GM league]"
|
|
|
On October 14 2011 14:19 whatthefat wrote:Show nested quote +On October 14 2011 06:45 Hulkoff wrote:On October 11 2011 21:35 DarQraven wrote:
See, this is why I often disagree with the advice to "only work on your macro, then you'll get to diamond really quickly".
One week ago, I was still Gold. My SQ from that period, averaged over 24 games is 63, which should put me inbetween platinum and diamond. I can't look at the history for season 2 anymore, but I hardly improved my macro lately. In Season 2 I was Silver, and if I had to estimate, my SQ would have been just below the plat level. (I checked unspent resources and income rate then, as well, just divided one by the other to get an estimate of how well I did).
It's anecdotal, sure, but it does show that there is slightly more to getting better at the game than straight macro. Unless you got that kind of SQ in very long macro games then it doesn't matter. SQ is absolute garbage because it ranks short easy macro games very highly. Every low leaguer can get 80+ SQ by doing a 2 base all in. While a better player might get a lower SQ because he played a long macro game that doesn't use an optimized all in build. It's a very poor way of displaying macro skill. SQ takes average income into account, and it shows no significant positive or negative correlation with game duration in any league. Your assertion that "every low leaguer can get 80+ SQ by doing a 2 base all in" is just completely untrue. As is your assertion that "a better player might get a lower SQ because he played a long macro game that doesn't use an optimized all in build". Neither of these are borne out by the data; unless you have collected a similarly large dataset, I don't know what you are basing these claims on. All SQ does is compare your spending efficiency in a game with that of all other players' games with the same average resource collection rate. It doesn't matter whether the game is a short all-in or a long macro game, because it normalizes for that. In other words, the average difference in SQ between Bronze and Diamond is the same regardless of game duration or income.
Statistical smackdown. Gotta love it.
|
69.09 for most recent game.
I'm in silver..
I need to play more, and average this out methinks.
|
Just came back after a month break, mine seems to vary anywhere from 60 to 85 :/
|
I think a big problem with the lower league players in computing their SQ is that, they simply do not make as many workers as higher league players. The difficulty of spending money compounds as your income increases, which is rather obvious to anyone who has played a drawn out game.
|
On October 24 2011 12:37 H2OSno wrote:
I think a big problem with the lower league players in computing their SQ is that, they simply do not make as many workers as higher league players. The difficulty of spending money compounds as your income increases, which is rather obvious to anyone who has played a drawn out game. That is exactly the point of the SQ equation - it takes income into account. SQ measures your spending efficiency relative to how well all other players spend at the same level of income.
To give some concrete examples:
At an income of 600, an SQ of 70 corresponds to average unspent of about 300.
At an income of 1000, an SQ of 70 corresponds to average unspent of about 500.
At an income of 2000, an SQ of 70 corresponds to average unspent of about 2000.
|
I'm way late into this thread and haven't had the chance to read all 49 pages (haha), but I just wanted to say this is a great write-up and with a little formatting touch up you could very easily use this as a Computer Science conference paper and likely receive a publication out of it. Congrats on all of your findings.
|
is there any news about this being integrated into SC2 Gears?
|
is the SQ league classification correct?
i just had a few games and my SQ is around master to grandmaster even though i am in platinum o.o and also my workers were at 100 at the 24min mark .
not sure if this is just a macro dellusion for me.
|
Malaysia5612 Posts
wow u could make a long lecture out of this.
what subject would this call as?
|
|
|
|
|
|
EPT![[image loading]](http://img687.imageshack.us/img687/6148/scr1ux.jpg)
![[image loading]](http://img684.imageshack.us/img684/948/scr2an.jpg)
