| Elefanto Switzerland. September 17 2011 20:43. Posts 3553 | Profile # |
On September 17 2011 20:41 Evangelist wrote:
Building in cycles does not lower the average since it is the production rate versus the amount of resources coming in that determines the average. Terran structures also tend to work in grouped cycles, so the resource distribution for them is virtually identical - it is actually more efficient to do it this way, since you keep a single internal clock and a mental count.
The macro of Protoss and Terran is actually completely identical - it just feels different. Protoss will wait for full warp ins, but not just for a sudden burst of units. It's because it is most time efficient. A terran will similarly macro in the same way, again because it is time efficient.
Peaks and troughs in a profile will be represented by the different buildings having different timings.
You can still queue units up, meaning you won't miss your cycles if you're fast enough on your keyboard.
|
| | wat |  |
|
|
| Evangelist September 17 2011 20:48. Posts 1169 | Profile Blog # |
On September 17 2011 20:43 Elefanto wrote:
Show nested quote +On September 17 2011 20:41 Evangelist wrote:
Building in cycles does not lower the average since it is the production rate versus the amount of resources coming in that determines the average. Terran structures also tend to work in grouped cycles, so the resource distribution for them is virtually identical - it is actually more efficient to do it this way, since you keep a single internal clock and a mental count.
The macro of Protoss and Terran is actually completely identical - it just feels different. Protoss will wait for full warp ins, but not just for a sudden burst of units. It's because it is most time efficient. A terran will similarly macro in the same way, again because it is time efficient.
Peaks and troughs in a profile will be represented by the different buildings having different timings.
You can still queue units up, meaning you won't miss your cycles if you're fast enough on your keyboard.
But it also means you've using up resources that should be used expanding, upgrading, and so on. That might lower your average Bronze Leaguer's unspent, but at Plat/Diamond or above? Similarly, if someone is queueing up constantly, it tends to be a result of frustration - they aren't building as much as they should, which usually means dead time for the building.
Similarly, Protoss can also queue at their major buildings - a pair of Colossus are worth 20 Marines for the purposes of that tab, so they can also artificially lower it as well. |
|
|
| vonterribad Australia. September 17 2011 20:59. Posts 50 | Profile # |
On September 17 2011 20:26 arbitrageur wrote:
Show nested quote +On September 17 2011 20:03 Elefanto wrote:
Nice work.
About the SQ:
But i think it's really only an indication for a Terran / Zerg, because Protoss macro with
Gateways work different.
You can't queue up units, or constantly build. You build in cycles, often leading your resources to go high. So your unspent resources are often higher than Terran / Zergs that can effectively macro their shit away with keyboard usage.
This is a fantastic point. Also I think terran is so easy to keep your money low compared with zerg in ZvT/TvZ.. some spots in zvt late game you don't want to actually spend any money because you're waiting for your tech transition so you can easily get thousands of minerals and it's not actually bad play to do so. but it's hard for terran to get thousands of minerals unless they're playing bad.
As somone that switched from zerg to terran I disagre (maybe I was a great zerg and terrible terran). However as Terran your income can be very 'spikey' and I also found due to the rigid nature of terran you really need all your unit producing structures up to anticipate more income in time to deal with the influx of minerals you will be getting. Zerg is much more fluid in this regard as long as you are hitting injects and have the larvea support.
I guess it varies from player to player but dumping minerals always felt easier to me as zerg ? (however as Terran I never overqueue which could be a factor). |
|
|
| komokun France. September 17 2011 21:10. Posts 166 | Profile # |
| Thank you for the astounding work |
| |
|
| methematics United States. September 17 2011 21:17. Posts 379 | Profile # |
fucking epic analysis.
whatthefat = god |
|
|
| Flowsick September 17 2011 21:29. Posts 159 | Profile # |
| Wow this is absolutely fantastic! Great work! |
| |
|
| Analysis Netherlands. September 17 2011 21:29. Posts 15 | Profile # |
Awesome work here. Some how it makes me think that the topic starter is a scientist IRL. Since the author is from america I'm guessing that the analysis has been done on the NA server? That might be a influencing factor, since it is very well possible that different servers have different levels of play. In other words, the SQ-scale that you made might be to low for players on the korean server. As example: I'm currently a plat user (I know, I suck  ) but I still have a SQ of 65, which should be according to you, a diamond player. Besides, I'm a really bad plat-player, I usually see opponents that play gold..
(i'm on the EU server btw) Anyway, my two cents: Very well done, the numbers might change though, depending on which server you are playing.. |
| |
|
| Dyme Germany. September 17 2011 21:30. Posts 424 | Profile # |
http://imgur.com/KUZNZ
Shouldn't Protoss have more workers than Terran, at least early game?
Chronoboost vs Mule?Last edit: 2011-09-17 21:31:55 |
|
|
| sAsThark France. September 17 2011 21:34. Posts 26 | Profile # | |
| | http://fedoraproject.org/ |  |
|
|
| Komeidon Finland. September 17 2011 21:37. Posts 2 | Profile # |
| The leagues seem to be totally off... I'm a masters terran, and it gave me 59.15 over last 27 games. I won 18 of those, so it should be a decent sample. Min: 27, max: 78. |
|
|
| Kwanny Germany. September 17 2011 21:38. Posts 219 | Profile # |
Well, I think many of you guys fail to see that it doesn't really show how good one's macro is in a general term, but in a very specific.
It explains how good the average spending was in terms of a comparable number from roughly 0 to 100, with any given average income.
You could take 1000 and 2000 income games and compare those. Realistically, the 2000 income guy will always win. And you could explain that the formular allows for a exponentially higher tolerance of unspend ressources in relationship to the amount of average income, the higher the income is, without having a drastic negative effect on the resulting SQ. So basically, it takes the empirical data into account that the more income there is, the sloppier a player is likely to spend his money. And the higher the SQ, the better he has spent his ressources.
The average income is still dependent on even satuation on expos, number of expansion and timings, and number of workers. Even if your SQ is higher and your spending is better doesn't mean that you have overall mined more resources than your enemy. And most likely, the guy who just HAS more, wins more often than not, which is his spent ressources - ressources lost.
The SQ can not account for that, but IN CASE that the macro timings and average income is the same of two players, then a higher SQ would indicate that this player has momentarily that much more spent ressources than his opponent. And still, that number would, with a reasonable difference between SQs, be pretty marginally. And SQ doesn't measure a progress, but a single point in time. Conventiently used at the end of the game for analysis, but really applicable to any point in the game.
- EDIT: I just figured tough theoretically, that a higher SQ player with same income would mean that this player on average has more out on the field at any given time during the game than his opponent, up to the point in time in which the SQ "screenshot" was taken - by how much? Depends on two variables, 1. the level of income they are playing on, and 2. the difference in SQ.
Anyways, there are so many things SQ can't do, and so many more SQ is capable of. Last edit: 2011-09-17 22:03:10 |
| |
|
| firesoulhc Germany. September 17 2011 21:48. Posts 2 | Profile # |
Very nice analysis.
Thank you for pointing out this part of macro! In my opinion it is another way of determining ones differences to other leagues. In no way it includes lack of strategy or use of "special" ones. We can see that by the large distributions of one league over a wide range of SQ.
On the other hand it does fit my personal league placement almost perfectly. Though my SQ has a deviation of about +-20 from game to game. |
|
|
| oMonarca Portugal. September 17 2011 22:10. Posts 11 | Profile # |
| TL needs to implement +1/-1 to posts. Awesome work man! Can't wait to see this passing over to SC2 gears. |
| |
|
| WightyCity Canada. September 17 2011 22:25. Posts 765 | Profile # |
| nice. now i know why im so horrible |
| | 90% watching it 8% talking about it and 2% playing it - sc2 | |
|
|
| Felorati Denmark. September 17 2011 23:01. Posts 3 | Profile # |
| I am so impressed, thanks for the very good read! |
|
|
| vsL United States. September 17 2011 23:02. Posts 20 | Profile Blog # |
| Very impressive. Great post!!! |
|
|
 repsac September 17 2011 23:48. Posts 91 | Profile # |
last 20 ladder games:
78.52
92.04
86.12
77.03
95.04
81.5
87.18
74.83
71.77
76.17
93.24
108.08
73.85
90.22
92.89
75.42
90.3
83.81
65.38
59.89
avg:
82.66
i tend to do poorly when facing quality harass.
the scores in the 70s are generally in tvzs where i'm constantly attacking. i could do a better job of spending my money in those cases.
i do best when i'm able to control the game and get as big as i want. that's where the scores over 90 tend to come in.
i should aim to perform better against early game harass and spend my money better when constantly attacking.
high masters NA terran. |
|
|
| Rising_Phoenix United States. September 17 2011 23:56. Posts 362 | Profile Blog # |
| An interesting and well researched article, however it boils down to one simple hypothesis: Better players macro better and have less unspent resources than worse players. Now if you went into what pro players think or do to become better players I would feel like it is much more useful. |
| | Early to bed and early to rise, makes a man stupid and blind in the eyes |  |
|
|
| InfStopher United States. September 18 2011 00:09. Posts 14 | Profile # |
| wonderfultastic, this post is wonderfultastic. |
|
|
| Boony Australia. September 18 2011 00:22. Posts 87 | Profile # |
Firstly I would like to congratulate you on an awesome study and write up.
Now a couple of questions/suggestions.
You compare game duration and number of workers for each league. I think it would also be interesting to compare game duration and average unspent resources. This would most likely yield similar results to the worker comparison, but would still be interesting.
I also wanted to point out that the game duration will affect the average unspent resources. When a player maxes out they begin to stockpile resources. As I understand how the SQ is calculated, this stockpile will cause the cause a low SQ score. I can not think of a method to avoid this besides removing particularly long games, which is less than perfect. |
|
|
| Prev 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 Next |
|
Store
![[image loading]](http://imgur.com/KUZNZ)
