Hi! Following up on whatthefat's fantastic article Do You Macro Like A Pro, here are some quick stats about Spending Quotient from 40,000 AM (Americas) games and 80,000 EU (Europe) games: http://ggtracker.com/spending_skill_stats

The chart there is interactive -- you can pick region and race, and see how SQ varies by game length.

The basic findings:

Higher-league players have higher SQ (duh )
SQ declines as games get longer, especially for Zerg
Terrans have higher SQ, all other things being equal
EU SQs are a bit higher than AM, maybe about half a league.

I'm about to go to bed now, so I'll check in tomorrow and answer any questions!

Yoshi Kirishima

United States10382 Posts

December 19 2012 04:19 GMT

Wow this is awesome, will check it out! :D

ff7legend

United States213 Posts

December 19 2012 04:22 GMT

is there any way to do this with at least top korean games to compare their SQ to other regions?

Yoshi Kirishima

United States10382 Posts

December 19 2012 04:32 GMT

So apparently I have diamond SQ

i do stack up quite a lot of resources often... haha

Insomni7

667 Posts

December 19 2012 04:35 GMT

Cool website. I assume it is continuing to update with new games?

CyDe

United States1010 Posts

December 19 2012 04:56 GMT

Doesn't this method of calculation not provide accurate results for Protoss? Because of warp cycles they have to stockpile more resources than the other races, then do "burst spending."

VPVanek

Canada238 Posts

December 19 2012 05:09 GMT

My spending skill is grandmaster apparently.
AWesome! :D

Lonyo

United Kingdom3884 Posts

December 19 2012 05:15 GMT

On December 19 2012 13:56 CyDe wrote:
Doesn't this method of calculation not provide accurate results for Protoss? Because of warp cycles they have to stockpile more resources than the other races, then do "burst spending."

Well comparing divisions will be accurate, since all protosses have the same mechanic/issue.
Comparing P to Z to T is kinda silly due to the mechanics of production.
Since itrelies on average unspent, average for P and Z will be higher than T, in theory, but a GM P will be better than a Dia P, and same Master Z vs silver Z.

Thienan567

United States670 Posts

December 19 2012 05:18 GMT

So does zerg when remaxing. Honestly Terran is the only race with constant production so I think it's to be expected that they have higher SQ over longer games. Overall terran players have better macro.

You know what that means. Terrans have better macro -> more skill -> other races are for noobs :p

Kyrao

United States161 Posts

December 19 2012 05:54 GMT

#10

Looking at the individual player profiles makes me wonder how accurate this is... the first 4 games of mine are like 8-10 months old and are bottomed out in bronze level spending, while the 3 last ones are slightly more recent and have me a GM/master level spending... uhh, ok.

Mia

75 Posts

December 19 2012 06:19 GMT

#11

On December 19 2012 14:18 Thienan567 wrote:

Show nested quote +

once remaxing other races spend resourses faster(insta remax from 60 larva vs couple rounds of rax/factory/starport build time), so actually there sq should be higher in late game..the only time zergs/toss sq is lower is early-mid game, when zerg saves up for flock of mutas, or toss does warpin round. But once maxed out, its the other way around.

oOOoOphidian

United States1402 Posts

December 19 2012 06:22 GMT

#12

This makes sense, as Terran has no burst production and they also can queue up. Also zerg tends to get maxed faster, which causes them to save up resources late game.

Emperor

Norway68 Posts

December 19 2012 08:08 GMT

#13

Seems like a cool way to check your macro compared to others. Right now i apparently got macro about equal to average EU GM. Which seems about right.

Kyrao

United States161 Posts

December 19 2012 08:23 GMT

#14

I don't understand the logic behind this idea that protoss should have a lower sq because of warpins... warping in 10 zealots is the exact same as selecting your barracks and hitting "a" 20 times when it comes to sq. The fact that protoss have an instant return on their investment, in terms of in game units, has no effect on these numbers...

romans

Australia18 Posts

December 19 2012 09:14 GMT

#15

On December 19 2012 17:23 Kyrao wrote:
I don't understand the logic behind this idea that protoss should have a lower sq because of warpins... warping in 10 zealots is the exact same as selecting your barracks and hitting "a" 20 times when it comes to sq. The fact that protoss have an instant return on their investment, in terms of in game units, has no effect on these numbers...

I think you are correct but a terran player can queue up units which instantly spends the money. Also, protoss players can be better served by waiting for every gateway to finish before making another round of units because they have to actually look at a pylon to do it while terrans can produce with a control group

Wraithan

United States18 Posts

December 19 2012 09:19 GMT

#16

It isn't the initial spending that makes SQ go up, it is having the second marine/tank/medivac/etc queued up which spends the money as well (also keeps the build times tighter together) you have to be absolutely on time to keep your warp in tight together, and even then that might be bad because you are saving to prism them, or so that you can warp in a proxy pylon as/right after you engage (depending on free supply) also higher income makes it easier and terran have mules to spike their income back up when they grab their fourth/fifth.

dynwar7

1983 Posts

December 19 2012 09:29 GMT

#17

Interesting information

cvgHuShang

Canada95 Posts

December 19 2012 09:36 GMT

#18

well im double a eu grandmaster... something is obviously wrong lol. (25 minute game too)

Wraithan

United States18 Posts

December 19 2012 12:57 GMT

#19

Interesting things about data sets like this. It is an average, not the range, of people in a given league. Perhaps your macro is pretty awesome but your micro is less awesome. Maybe those games you were starved for resources most of the time (because of people sniping your workers) so it was easy to keep your resources down. It takes a decent amount of data before you can really come to a conclusion.

In looking at a single person's data, I watch them (no game length data in the data I have) fluctuate between low gold in some matches up to high diamond in others and they are platinum.