Posted by:Starparty The Gas Issue - part I by Starparty
This is an modifyed article of the one i posted on teamliquid some time ago explaining how the gasnodes work in StarCraft:
I see alot of new maps and mapmakers here who show nice maps with great designs and funny ideas. However, there is one quite anoying pattern that goes trough all of them. Im not saying it's wrong, im saying its not preferable. Im talking about the gas placement.
When Blizzard made this game some things we later on would have appreciated (like up left/right ramps etc) were forgotten, or not made by intetion - what do i know. 1 thing they could have corrected by now, since the game has reached proffesional gaming status, is to fix that anoying gas issue.
This is common belief:
The problem is as following:
"People tend to use the placement since it is symetrical. Unfortunately the game doesn't work that way..."
I was talking to hefty the other day as he seemed to do the same thing, and he said that the WCG maps had the above shown placement [picture #1]. People tend to use the placement since it is symetrical. Unfortunately the game doesn't work that way... Personally i think its quite weird that wcg maps use that placement since a player can be really set back in certain MU's if he doesn't know he has to use 4 workers on the gas...
Tests on a map where 1 Comand Centers had 2 refineries on each side (left and right) shows that that, in 15 minutes, the left refinery mined 4512 gas while the right refinery mined 4080 gas (Thanks to Tuy for taking his time making the tests). This is quite alot, specially considering mirror match ups. In a zerg vs Zerg, one player may have 1 or 2 mutalisk more in the first air battle, meaning not that he will have 2 more when the fight is over, but several more. In a Terran vs Terran, One player has 2 more tanks, and the game could be game over right there.
I am pointing this out since i seem kinda nagging to be commenting on the issue every single time a new map is released... This is what i learned from the mapcompetition this winter, and by experience i know it works this way. Consider this next time you release your new map and my reviews may be less harsh
Hi folks, this is the second part of "The Gas Issue".
I tested the gas thing as well and found something very interesting, but first of, all information about the test itself:
Testruns: 3 times Testtime: 3:28 minutes each Testrace: Terran Testmap: (on demand) Tester: Myself, each time on a different position Workers: 11 workers from startup | 2 on Gas* | 9 on Minerals
(* only 2 workers at the gas because the CPU players only sent 2 workers to the geysir)
(the single mineralblock represents the gas position)
1. As we already knew, you gather the most minerals when the geysir is on 9 or 12 o'clock from the HQ. 2. You gather the most gas when the geysir is on top at 12. 3. Except these positions most of the others, even the diagonally, are almost balanced. 4. The most surprising result was the gas position below the HQ, on 6. It was by far the worst position for the gas! 5. I watched the workers a lot. When the gas was on 6 o'clock position, they moved like they would fly through a crater or a hole or something.
-> they flew slow from the HQ towards the Gas
-> then they speeded up like flying into a hole
-> then they slowed down like flying out of a hole
-> then they flew slow in front of the geysir again
-> as they returned, the excact same thing happened
My final words are, STOP setting the geysir below the HQ on every map because the gas difference with only 2 workers in 3:28 are almost 200 gas!
Anyway, what you do now with this test result is your thing. Just keep it in mind.
Good job now i can go back and say imba at every game i played that i had 6 o clock gas ;]
cant see image
btw, img works fine? F5 ? ^^
isnt 4 workers required for maxium gas mining at those "good" spot also..
i dont see why pro's should send 4 works to gas when its above otherwise
test, wtf why is everything fucked up?
I fixed your text ;) Don't forget to have the " on both sides of your link.
Imo the Test with two shows the difference in distance very well, but you don't know how the data would change with 3/4 workers.
Did all players have the same race? There are slight differences there too!!
jip, all T.
Hmm... i test it with the other positions by myself and 3 workers on gas now. just wait a bit. ^^
hmm.. i think i will come up with a whole testresult with all races in a "real" buildorder to check the difference. I'll write a newspost if i'm done :)
nice test.. though it seems kinda useless since u had only 2 workers at the gas
can u put 3 peons in there next time ?
The computer seems to mine quite uneven too. perhaps try the test manually :P
I HEARD A QUESTION!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
so the best position is 12
Does the exact strategy of the test matter that much, having two scvs, all with T? He made his point that if all are mining equally, despite how, there are disadvantageous gas locations. THose locations are apparent in the test. The point was made quite clearly. I don't think more tests need to be made, for this shows advantageous gas positions and un-advantageos ones. This data can readily be used, without having to retest with 3 workers. Zerg and Protoss shouldn't variate much.
Oh, and there are strategic places to put a supply depot/spawning pool/forge that will speed up the gas (extraction?) process. You place the building in a position to block any route other than strait to the cc/hatchery/nexus. Try it out. It may make the results different, as these placements are not available with all of the Geyser placements.
I know this is old, but.....
The gas, when at the corners, appears to be very close to balanced.