EPTThe Zergling/Marine Efficiency Experiment - Page 2
| Forum Index > Brood War Strategy |
micronesia
United States24676 Posts
| ||
|
TheOvermind77
United States923 Posts
On May 11 2007 20:48 micronesia wrote: I found this a fun read (being a man of science) and interesting in general, but didn't provide much new insight. Maybe you can work out some minor kinks in what you did and then do something else using the adapted method. What would be really cool would be if you could reasonably disprove some kind of a mechanical misconception in bw. Heh, I know, it really only went to show things that many people already knew. Still, I found the numbers interesting to crunch. Finding a 'mechanical misconception' that can be disproved isn't something easy to do... | ||
|
micronesia
United States24676 Posts
On May 11 2007 16:14 TheOvermind77 wrote: 5. The fourth group, the Burrowing group, did not receive such a significant decrease as the number of marines decreased. The efficiency for this group varied quite widely in a dampering simple harmonic oscillator motion. I have yet to explain this. I thought about this while I was in the shower just now (because that's what I think about while I shower). Your classification of this behavior has a 'dampering simple harmonic motion' is premature. I don't believe you have enough data to claim that, even though I'll admit the small bit you do have graphed is strangely like a damped sine wave. I have a feeling if you can carefully increase the number of units, the observed effect will decay away as the interactions become less discretized. If you repeat this part of the experiment you also may get a differently shaped graph. The fact that only the Burrowing group didn't get immediately hurt by an increase in numbers stands to reason, I believe. It is the only of the 4 tests you ran where marine range has little to no benefit, almost regardless of numbers. | ||
|
I-Emerge
United States435 Posts
On May 11 2007 19:16 fusionsdf wrote: Thats a nice strategy, but isnt the terran going to be a bit suspicious of the low ling count? Good observation, There was post I read here that explains the uses of misinformation. The post focused on hiding units in your main to throw off your opponet. In the game I played using barrow I remembered the post's comments and I reset my rally points for all ling producing Hats to a location not normaly scanned in my main. This helped me take advantage because of the lack of Bats created based on misinformation. Also... On May 11 2007 19:18 TheOvermind77 wrote: Low ling count could just mean drone powering. Or lings could be behind sunks or elsewhere on the map. Lots of possibilities! ---I-Emerge | ||
|
TheOvermind77
United States923 Posts
On May 11 2007 21:27 micronesia wrote: I thought about this while I was in the shower just now (because that's what I think about while I shower). Your classification of this behavior has a 'dampering simple harmonic motion' is premature. I don't believe you have enough data to claim that, even though I'll admit the small bit you do have graphed is strangely like a damped sine wave. I have a feeling if you can carefully increase the number of units, the observed effect will decay away as the interactions become less discretized. If you repeat this part of the experiment you also may get a differently shaped graph. The fact that only the Burrowing group didn't get immediately hurt by an increase in numbers stands to reason, I believe. It is the only of the 4 tests you ran where marine range has little to no benefit, almost regardless of numbers. Ok, it is late at night and your post bugged me. I kept thinking about what that graph would look like as it was extended. Go check out the "Extended Study" section at the end of the updated experiment. | ||
|
Equinox_kr
United States7395 Posts
On May 11 2007 16:47 ilovezil wrote: Quite some work you\'ve done there...Yo, pm me and let\'s do some tests with toss upgrades vs terran upgrades etc! This stuff looks mad neat :O. seconded  | ||
|
micronesia
United States24676 Posts
IIRC my conclusion was that zealots>ultras if they have reasonably comparable upgrades, lings>zealots with reasonably comparable upgrades, and ultra+ling >> zealots. When I say comparable upgrades, I mean if zealots don't have at least +2 yet, the ultras don't have any special upgrades at all yet, etc. | ||
|
TheOvermind77
United States923 Posts
On May 12 2007 10:48 micronesia wrote: A few months ago I did something similar regarding pvz. I had armies of zealots fight armies of lings, ultras, and ultra ling, while varying all the upgrades. I didn't publish my results or display my data in a pretty way but it seemed to me to be somewhat useful. Maybe I or someone else should go back through that study again like the way this one was done. IIRC my conclusion was that zealots>ultras if they have reasonably comparable upgrades, lings>zealots with reasonably comparable upgrades, and ultra+ling >> zealots. When I say comparable upgrades, I mean if zealots don't have at least +2 yet, the ultras don't have any special upgrades at all yet, etc. You should get some stats on that. I'm considering taking my spreadsheets and making another experiment...I have a few ideas that may be useful. | ||
|
micronesia
United States24676 Posts
It will be difficult to perform effectively. The goal would be to see what units are more effective given equal money, and a lot of this, in actuality, comes down to the micro. Also, how many zealots have the same cost as an ultralisk? I can count gas and minerals the same (4 zeolots = 1 ultralisk), but that's only partially fair. The way the computer controls the units is key. | ||
|
TheOvermind77
United States923 Posts
| ||
|
Slayer91
Ireland23335 Posts
 | ||
|
TheOvermind77
United States923 Posts
On May 12 2007 13:50 Slayer91 wrote: You too can just work together on it so the comp ai doesn't matter. The only problem is that everything must be done the same every time or else the experiment isn't credible...human error might cause our micro, etc to warp the results. But it is an idea :D. | ||
|
decafchicken
United States20019 Posts
On May 11 2007 18:29 evanthebouncy~ wrote: Ahh I nvr realized it's cheapter marine 50 minerals than zerg's 50 minerals. hmm! It's not. Due to the gas usage of Zerg, they usually have more bases mining, but the gas usage is comparable to the mineral usage. But seeing as you have less gas, you have minerals left over after gas is spent, leaving lots of money to be spent on lings. | ||
| ||
