itch.io

this looks so professionally made, you should charge for this masterpiece!

Erster!

Erster!

First!

Pretty tough to reach 1600 points...??.?? and I reached only ~1550, it's so hard to get there again.

btw anyone know who this Savelas is? I'm pretty sure it's an alternative account from a top player since he beat me very often and to be honest he gave me a few ideas for my ai. But i don't like his name -.-

Yeah it's an interesting idea with the flanking formation, it has some potential to be discovered. But for the moment i prefer luring the center robot into my trap and attack with full power while the enemy spread their shots over my bots.

This challange turned into some kind of war, playing paper, rock, scissors and hoping that you have chosen the right one.

Great thanks. In fact, it's because of the randomized chance of getting hit i think. If both players use similar ai it's all up to the fortune, right? I hope you enjoyed the prioritize map. ;)

But it's true. On maps like 'starving' and 'you need to priorize' the starting formation of the robots prefers a certain strategy.

As for complexity: The more complex it is the more bugs it may have. You should test it as much as possible tweaking it by a little each time. This requires a lot of time so I prefer fast and efficient ais, because if you encounter an ai perfectly working against you, you have to make a new one anyway, and update it from time to time. Currently I use one main ai but with different subtree nodes, which can be easly exchanged. This way I don't have to change each robot if I want another strategy. But for specific maps, if i find an interesting idea, I love to do some high complex ai, just for the matter of joy.

To be honest I started to worry about the information I could give you about my ai, so here's a HQ picture of my all-around ai: http://www.directupload.net/file/d/4437/3yqivthg_jpg.htm

Some great ais you made there Steelwatch.

I love how you can be top 1 and suddenly someone discover how to beat you.

One more additionally thing I would like to mention if you host several games with same map it's recommended to use different ais even tho you have this one and only masterpiece since people adjust their ais if they know what they are dealing with.

I guess Phil got frustrated cause of 200 points of loss haha.

Anyone here up for a 1600 elo challange?

Actually you can see I use long range when my units are recovering.

I finally managed to beat Phil! But as you can see it's devastating on split team, nice matches to see how powerful tha ai can be. I started to make more complex ais now.

I dropped hardly after the changes, but doing better now. Phil is the only player I NEVER won against.

Up for a game anytime.

haha good idea btw

I guess 15% is enough to convince you you faced exactly that ai

I know this code seems so simple but the robots are amazing in the actual fight, now I made new ais based on this one

To be honest this should encourage me to improve it since you know the weak spots now. ^^

btw CHALLANGE ACCEPTED

(Firefox crashes constantly when trying to edit on my tablet)

I needed over 30 failures to develop this baby. You can ignore the far right, it's just for grabbing balls and running away to make space for transporting robots. As you can see I focus on the weakiest one. But the thing is, if I try running for the weakiest I end up being attacked with 90% while I'm still trying to catch this low health dude. So I prioritize low health enemies in sight. My robots still go for the weakiest enemy causing some interesting regroup maneuvers. As for the shield I just take one step back and attack. It doesn't matter if this unit is still being attacked. With a 90% dodge chance you probably will recharge shield anyway and your enemy just lost some important shots.

The game mostly ends in a regrouped army which strongly holds together and the enemy is spread doing only little damage. I win by pushing enemies to the barrier.

Greetings strategists!

I discovered this master piece of great potential yesterday, have done the quick solo missions and stepped into multiplayer. Never got addicted to a game that fast, I found myself with 1500 elo today. But I had not enough so I discover this newly created community as well.

Reading through the posts I found the exact same thoughts when developing my lovly ai. It seems like the current styles of play haven't changed a lot. Everybody (well, except the guys who can't wait until training mode is out, so they try crazy stuff) is trying to kill your babies one by one, going either one step or fully back to recharge (which I don't recommend due to the fact, that you should stay close at all cost) and then collecting the balls at the end.

I would like to share you the build but somehow I have problems with posting the image.

http://www.directupload.net/file/d/4436/7y8j3f66_png.htm