Your math must be incorrect:

4(2) | 3.05 gas/second
5(2) | 3.33 gas/second
6(2) | 3.94 gas/second

The Increase from 4(2) to 5(2) is .28g/s, and the increase from 5(2) to 6(2) must also be .28g/s, making 6(2) worth 3.61g/s.

The only thing that could change this is if the map / spot you chose had a different distances. I would recommend redoing the data on a map where the gases are equidistant from the Hive/Nexus/CC

DarkChrono

United States16 Posts

April 26 2010 21:36 GMT

#82

I'm laughing so hard at gasexpfunction.png (2nd graph). You interpolate from the data of 1,2,3,4,5,6 workers, and then add the caveat that that everything you've interpolated is meaningless. So essentially what you've done is intentionally introduced error to your data points by artificially fitting it to an inverse exponential, with no benefit. If you think about it, the real function should be like 1/x, but this is still meaningless, because as you discuss this formula only applies on the domain {1,2,3,4,5,6}.

If this seems harsh I apologize, the rest of the article was very interesting and it is great you are taking the time to do this.

e: oops I see this has been said about 5 times before.

Slardar

Canada7596 Posts

April 26 2010 21:52 GMT

#83

The first thing that popped in my head was the effect of geysers returning 0 gas once depleted would have on the game. Perhaps an analysis of this has been discussed earlier and I've overlooked it. Otherwise a very good read for any aspiring SC2 Pro :D

blahman3344

United States2015 Posts

April 26 2010 22:41 GMT

#84

very interesting statistics there.
glad to know that there is no question between 3 or 4 workers on one gas geyser anymore.

with that, we are one step closer in analyzing sc2 =D

stork4ever

United States1036 Posts

April 26 2010 22:58 GMT

#85

whoa, this is why i joined this forum

Onea

United States35 Posts

April 26 2010 23:22 GMT

#86

Thank you for proving what most people "make an ASS out of U and ME"

bongjwa

United States199 Posts

April 27 2010 00:01 GMT

#87

awesome research and write up i really enjoyed reading it! just felt as if it could've been summed up a lot quicker

brocoli

Brazil264 Posts

April 27 2010 00:18 GMT

#88

Good initiative from OP, and nicely exposed, but this article is missing the most useful thing it could hold : analysis of far away geysers vs close geysers.

Also, any good game has research like this. It is what makes playing them more than just a passtime, and what makes playing them fun and rewarding.

aztrorisk

United States896 Posts

April 27 2010 02:03 GMT

#89

HMMM, i guess blizzard makes up for the depleted mining by adding a second geyser. However, this is devolution by not being about to mine 8 gas per mine!!!

alaphonse

Canada5 Posts

April 27 2010 02:05 GMT

#90

so this means that a 5(2) is better than a 6(2) in the start of the game cause this seems a little hard to understand . . .

ironchef

Canada1350 Posts

April 27 2010 02:53 GMT

#91

Very interesting good to see hard data on the 3(2) > 3(1). Also 4(2) > 3(1)+ 1(1) .

The obvious questions I guess is are there any interesting fast 2 refinery builds to take advantage of this. Perhaps to speed up a gas heavy timing attack?

bay

United States20 Posts

April 27 2010 03:07 GMT

#92

Wow- really fantastic article. This is the kind of stuff I have been looking for! Thanks so much

Great job-

SichuanPanda

Canada1542 Posts

April 27 2010 03:10 GMT

#93

There are high-yield geysers in the editor, so once release I'm sure we'll see maps popping up all the time that have high-yield gas, completely changing the pacing of gas.

Wintermute

United States427 Posts

April 27 2010 03:25 GMT

#94

On April 26 2010 23:43 Archerofaiur wrote:
ESCIENCE!

But seriously is there any game on the face of the earth that recieves this kind of dedicated research. Im sure Halo doesnt have statistical analysis on a professional level.

World of Warcraft theorycrafting blows SC theorycrafting away in terms of sheer volume and complexity.

At this point, it's complicated enough that the community uses mods, monte carlo simulations, and any other automation they can to determine optimal gear setups, talent specs, and DPS rotations.

It's amazing the level of dedication people have to figuring out things that aren't real.

iheartpurplez

Canada54 Posts

April 27 2010 03:33 GMT

#95

this was extremely well written !
the content was nice too, not sure how significant it may turn out to be due to all the things you mentioned at the end,

great post nonetheless!

Wi)nD

Canada719 Posts

April 27 2010 03:46 GMT

#96

On April 26 2010 17:36 prOxi.swAMi wrote:
SCIENCE.

Seriously, very interesting. Kudos for going so in-depth into it. Your graphs got me thinking, I think SC2's release will hopefully add some really interesting graphs in their scores screen. The current ones are good but just two is not enough there's room for way more.

CaptainPlatypus

United States852 Posts

April 27 2010 04:29 GMT

#97

So if I have two refineries and 3 SCVs mining each, how long will it take before my government develops alternative energy sources?

Jacopana

El Salvador210 Posts

April 27 2010 04:50 GMT

#98

Great Analisys, so I guess I was right, thanks for the confirmation. GJ

Iri

150 Posts

April 27 2010 06:19 GMT

#99

Is there any possibility of adding scv and drone mining lost to the minerals per gas rate comparison? Obviously a non-issue for protoss, but I would be particularly interested to see if the 3(2) and 4(2) come out as far ahead with mining lost taken into account.

hacpee

United States752 Posts

April 27 2010 06:30 GMT

#100

As for the maths, I think the WoW community does the most analysis out of all the blizzard games. They run simulations and derive equations used by blizzard and run cost/benefit analysis fairly regularly with every new mechanic introduced.