Do you macro like a pro? - Page 39
Forum Index > SC2 General |
Bonham
Canada655 Posts
| ||
Teoita
Italy12246 Posts
On September 19 2011 01:04 pyjamads wrote: Nice write up.... But I was thinking, this SQ might be a bit off. I mean say you are a Gold/Plat player and your worker creation skills are lower than the higher leagues, as was proven in your excellent analysis. Then your avg. income will never get as high as the income of higher level players than you, and you won't have trouble keeping your resources low. In that case you would get a high SQ score, even though your macro could be terrible. So that might explain, why a lot of players in Plat / diamond calculate their SQ and get results around Masters level, even though they get out macro'ed by higher level players than them selves. The point is that there is actually more than one way of being bad at macro, first you could make a ton of workers and not be good enough at spending your income, which is what SQ determines, and secondly you could be bad at making workers but still be good at using all you resources, where SQ has kind of a loop hole. In both cases your would probably be using the same amount of resources as everyone else in your league, which is probably why you are there in gold/plat/diamond or what ever. I'm not hating on the post, I think it's great, just think that it has a loop hole, I mean with this calculation, you kind of assume that "constant worker production" is constant and not faulty throughout the game (at least when your not cutting workers on purpose). Say I play a game, with my gold buddies and I only make like 40 workers all game long, doesn't matter I'm on 3 bases, I just don't remember to build them. And because I'm terrible at macro'ing just like anyone else in gold, I can only use what resources my 40 workers gather. Then at the end of my 30 minute game, I might have avg. unspent 447 and income 1415, which is like IdrA, only IdrA's game is 14 minutes long and he probably got harassed a lot in that period. The problem is that then my SQ would look exactly like IdrA's, and i might even have won because of how terrible everyone is in gold, so now I actually think I'm good, and that was not exactly what was intended with SQ. If I've overlooked something that makes my argument look stupid, please go ahead and ridicule this post hf Yeah, that's partially correct. SQ calculates, mostly, how effectively you spend your income, not how you generate that income. You can however also compare how many workers you made in your game and relate that to the GM graph in the post to know wether you macro'd well or not | ||
pyjamads
Denmark33 Posts
On September 19 2011 01:15 Teoita wrote: Yeah, that's partially correct. SQ calculates, mostly, how effectively you spend your income, not how you generate that income. You can however also compare how many workers you made in your game and relate that to the GM graph in the post to know wether you macro'd well or not I agree, what I'm saying is that while you can compare that graph and all that, SQ as a figure doesn't really work across all levels of skill, as it is put out in the OP. I'm not saying that it is not true, just saying that apparently a lot of people have calculated their SQ and found them selves to be better than expected at macro'ing, and that might be because of the problem I mentioned. Just saying that maybe the SQ formula, needs to take the game length into account. | ||
yeastiality
Canada374 Posts
| ||
CynanMachae
Canada1459 Posts
| ||
ShamTao
United States419 Posts
Thanks! | ||
whatthefat
United States918 Posts
| ||
Nazeron
Canada1046 Posts
| ||
Kess
Denmark1 Post
| ||
Nallen
United Kingdom134 Posts
Must spend moar. | ||
Nemireck
Canada1875 Posts
On September 19 2011 02:05 whatthefat wrote: Regarding the effects of game duration, I didn't find any positive or negative correlations between SQ and game duration for any of the leagues. This is likely because longer games tend to also have higher incomes. I've now added this info to the FAQ as a few people have raised this point. You could, however, calculate an expected income at game-time, and factor in that score in place of the specific game-time. The point many people are raising is that lower players will have a higher SQ in longer games, not because they are spending efficiently per se, but because they are not increasing their income at the same pace as someone with better macro. So if you run the data and find that at 17 minutes, a GM has an income of 2000, but a platinum player only has an income of 1500, you generate a score, and include it in the calculation for SQ, where a player with 1500 at 17 minutes will generate a lower SQ, even though they are spending their income just as "efficiently". | ||
whatthefat
United States918 Posts
On September 19 2011 05:00 Nemireck wrote: You could, however, calculate an expected income at game-time, and factor in that score in place of the specific game-time. The point many people are raising is that lower players will have a higher SQ in longer games, not because they are spending efficiently per se, but because they are not increasing their income at the same pace as someone with better macro. So if you run the data and find that at 17 minutes, a GM has an income of 2000, but a platinum player only has an income of 1500, you generate a score, and include it in the calculation for SQ, where a player with 1500 at 17 minutes will generate a lower SQ, even though they are spending their income just as "efficiently". Thanks for the suggestion. I considered doing something along those lines, and looked at correlations between game duration and income. But even at GM level, the correlation between income and game duration is extremely noisy. In general there seems to be much more useful information in spending as a function of income. I believe this is because there are many different viable strategies - there isn't a single optimal expansion timing - however, all players are constrained by the need to spend their available income as well as possible if they are to win. I agree that at the extremes you could perhaps introduce a penalty (e.g., if you still have a one-base income after 20 minutes, then you are probably doing something wrong). But players who do that tend to also spend very poorly - they don't just sit on one base and spend their meager income perfectly (which would be one way of "faking" good macro by the SQ score). | ||
Dbla08
United States211 Posts
| ||
Mordanis
United States893 Posts
| ||
Excalibur_Z
United States12224 Posts
Also I want to make sure people aren't finding their SQs and using that as some "magic number" to define their skill level like they did in that closed "calculate your MMR" fiasco of a thread. Macro is only one component of the game and there's more to it than that. | ||
Schnullerbacke13
Germany1199 Posts
| ||
Schnullerbacke13
Germany1199 Posts
| ||
Lysergic
United States355 Posts
I also did it for Nada's last 20 games at MLG Raleigh, his SQ was 94.3 O_O, he should get his face on the scale past Idra. | ||
Theeakoz
United States1114 Posts
| ||
MiKTeX
United States234 Posts
bronze, silver, gold, platinum, diamond, master, grand master, and idra | ||
| ||