Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Using genetic algorithms to find Starcraft 2 build orders

lbrandy.com

Post reply on HN
Print view View original

41–50 of 79 posts

Re: Using genetic algorithms to find Starcraft 2 build orders

#41

Post by Artagra »

alnayyir wrote: post #37
jtchang wrote: post #21

I am a diamond level terran on ladder and have faced this rush build order before. To be honest it is a very strong build but most players will scout this very early. I would be more interested in searching the space for hard counters to build orders with certain match ups. Certainly it gets way more complex (and starcraft is).

Yeah I was going to say, I'm a silver level and this strat does not work that well even in that tier.

I just played 17 games using this strat, in the Gold League 1v1 and Bronze league 2v2. And some of the 1v1's where against Diamond league players. 12 wins, 5 losses. Trust me, it works.

1. Against Toss it's almost always a win, especially if you go all in (attack using queen and half your drones). Unless he's got a closed wall with cannons, he can't stop you. In my experience, no-one does this in just about any league.

2. Against Terran, it's very succesful, if not quite as succesful as against Protoss. If they put up a bunker with marauders, or they micro 2 or 3 marauders behind a wall, you need to fall back. But in general, it works amazingly.

3. Against Zerg it works very well, as long as they don't six pool - if they do six pool and you haven't scouted it then you are in trouble. But in general, it worked well.

Bottom line? Although it seems like it's similar to the traditional zerg rushes, it's actually a very impressive rush. I would say that well executed it's got a 90% success rate on maps with short to medium rush distances. So definitely give it a try :)

Re: Using genetic algorithms to find Starcraft 2 build orders

#42

Post by 3pt14159 »

jacoblyles wrote: post #33
Psyonic wrote: post #31

Earlier quoted context omitted.

I've had that happen more times than I can count. It's hard, because all the terran has to do is hit stim on his MMM ball and A-move. As zerg you have to pull off some Robert E. Lee shit to have a chance.

My biggest problem is pushing into an army backed up by a planetary fortress. I think I need to wait for the Terran to push out more and then destroy his army in the field or drop into his main. Pushing the front of a Terran base is suicide.

if you see that go hard into broodlords. A PF in a home base is a giant waste of money if not used, especially considering the additional building armor upgrade and range upgrade he spent on it and the lost time-income of no mules. You will easily out macro him.

Re: Using genetic algorithms to find Starcraft 2 build orders

#43

Post by coryl »

Artagra wrote: post #41
alnayyir wrote: post #37

Earlier quoted context omitted.

Yeah I was going to say, I'm a silver level and this strat does not work that well even in that tier.

I just played 17 games using this strat, in the Gold League 1v1 and Bronze league 2v2. And some of the 1v1's where against Diamond league players. 12 wins, 5 losses. Trust me, it works. 1. Against Toss it's almost always a win, especially if you go all in (attack using queen and half your drones). Unless he's got a closed wall with cannons, he can't stop you. In my experience, no-one does this in just about any leagu…

1. Wouldn't a sentry forcefield ramp block off do just well? Stalkers have superior micro.

Re: Using genetic algorithms to find Starcraft 2 build orders

#44

Post by richcollins »

codexon wrote: post #40

This is the type of non-obvious optimization that genetic algorithms excel at. This is not true. GA search is a hill climbing algorithm which is prone to finding local maxima. The less obvious a strategy is, the more likely it is that GA will miss it.

GAs are less likely than other learning techniques to get stuck at a local maxima because mutations and crossover can cause to them to search entirely new areas in the space (unlike ANNs for example).

Re: Using genetic algorithms to find Starcraft 2 build orders

#45

Post by NyxWulf »

3pt14159 wrote: post #39
alnayyir wrote: post #37

Earlier quoted context omitted.

Yeah I was going to say, I'm a silver level and this strat does not work that well even in that tier.

Are you sure you are timing it correctly, using it on the right map, and microing ok? It should decimate most Protoss players unless they scout, discover, forge up, delay possible cannons until last min, dual exp, and quick air or go all in sentries.

> It should decimate most Protoss players unless they scout..

It should (annhilate|destroy|overwhelm|) most Protoss players unless they scout...

FTFY

That usage of decimate is as grating to me as it is common. :/ I don't blame you, I'm just shaking my fists at the heavens in impotent rage.

Re: Using genetic algorithms to find Starcraft 2 build orders

#46

Post by sliverstorm »

eru wrote: post #36
seigenblues wrote: post #34

Earlier quoted context omitted.

Totally -- not to mention surviving long enough to let compound interest naturally magnify that difference. As for the tactics or "logistics" -- if that means clicking 500 times/minute, then i really don't see the appeal :)

StarCraft 2 has tastefully upgraded the UI so that you can do lots of things with less clicks than before in SC 1. (Of course, that means at the higher end, people will just do more now than they did. But it's quite a nice feeling on the beginner level, too. E.g. you can get a worker to work right after it's build. And there's also a button for finding idle workers.)

Honestly, the graphics may be better, there may be new units, and balance may have improved or whatever else, but the UI improvements are the thing I love the most.

Re: Using genetic algorithms to find Starcraft 2 build orders

#47

Post by Artagra »

coryl wrote: post #43
Artagra wrote: post #41

Earlier quoted context omitted.

I just played 17 games using this strat, in the Gold League 1v1 and Bronze league 2v2. And some of the 1v1's where against Diamond league players. 12 wins, 5 losses. Trust me, it works. 1. Against Toss it's almost always a win, especially if you go all in (attack using queen and half your drones). Unless he's got a closed wall with cannons, he can't stop you. In my experience, no-one does this in just about any leagu…

1. Wouldn't a sentry forcefield ramp block off do just well? Stalkers have superior micro.

Usually when the first three arrive, they don't have a sentry yet (The first three are about 15 to 20 seconds ahead of the next 4). If they do have a sentry, and they block, then you fall back and then attack with all 7. If they don't use it, you can usually use those three to start killing their units, and the next 4 come in and finish the job. If they are rushing an immortal (high level players often do this vs Zerg) then you can usually snipe the pylon. If there is more than one pylon powering the building, then it's tougher.

In 1v1, a variation on this strat is to go all-in - send your queen after you spawn larvae, and then when the second 4 come out send most of your drones. With 7 roaches + drones + queen, it doesn't matter if he gets out the Immortal, or infact what he goes for. Only thing I've seen is a hard cannon wall, which then commits him to a single base Void rush - which is easily countered.

So yes, this strategy is very unbalanced. Played 22 games tonight, 7 losses. 2 of those losses where on 2v2. Last 10 games I played? 9 wins, 1 loss.

I hope Blizzard fixes this soon :( it's going to make Terran and Toss very susceptable early game.

Re: Using genetic algorithms to find Starcraft 2 build orders

#48

Post by jacoblyles »

3pt14159 wrote: post #42
jacoblyles wrote: post #33

Earlier quoted context omitted.

My biggest problem is pushing into an army backed up by a planetary fortress. I think I need to wait for the Terran to push out more and then destroy his army in the field or drop into his main. Pushing the front of a Terran base is suicide.

if you see that go hard into broodlords. A PF in a home base is a giant waste of money if not used, especially considering the additional building armor upgrade and range upgrade he spent on it and the lost time-income of no mules. You will easily out macro him.

Normally it's at his natural, like the open forward expansion on Delta Quadrant. I lost a game just today where I had a huge macro advantage over the Terran because of that PF. I even had Broodlords but I wasn't good enough to micro them against the SCVs repairing the fortress. I ended up destroying it with banelings, but by that point it had eaten up many thousands of minerals worth of my army.

Re: Using genetic algorithms to find Starcraft 2 build orders

#49

Post by jacoblyles »

Chris_Morrell wrote: post #28
NickPollard wrote: post #16

Earlier quoted context omitted.

By 'High-level play' he is most likely referring to tournaments and the e-sports scene, which is almost entirely 1v1 and plays very differently from 4v4.

I see. In that case it applies even more. It might be considered a cheese rush but an all-in with your drones would be very hard to counter unless they were able to hold the choke. It would have a good chance of ending the game in under 8 minutes versus drawing it out another 5-7 minutes.

It seems like cheese has failed more often than not so far in the GSL. At best it's a 50-50 proposition.

Tester deftly deflected cheese attempts from Hyperdub and NexFreeSaga before crushing the remains of their army in season 1. On the other hand, fan favorite TLO was eliminated by some proxy barracks from the cheese-loving Hyperdub and GSL season 1 winner aFruitDealer showed that he wasn't above a cheese game or two with a ballsy six-pool on a four spawn map against Inca.

Cheese fails every time if it's spotted against a high-level player, which is why it's cheese. These guys have definitely practiced against 6-pool, proxy gate, proxy rax, 7 roach rush a thousand times. There's no way they would take a cheesy loss with $80,000 on the line or give up the easy counter and auto-win.

Re: Using genetic algorithms to find Starcraft 2 build orders

#50

Post by henning »

codexon wrote: post #40

This is the type of non-obvious optimization that genetic algorithms excel at. This is not true. GA search is a hill climbing algorithm which is prone to finding local maxima. The less obvious a strategy is, the more likely it is that GA will miss it.

> GA search is a hill climbing algorithm which is prone to finding local maxima

GAs are not pure hill climbers. Variation operators introduce new solutions into the population at random. As long as premature convergence is prevented, GAs will probably explore the search space well. There's always a tradeoff between exploration and exploitation.

> The less obvious a strategy is, the more likely it is that GA will miss it.

This is a tautology that states "the harder a problem it is, the harder it is."

Post reply on HN