BigPete

Student at the University of Bristol (England) studying Engineering Mathematics.

Bots

PH4448 -- computer science project (submission deadline 8th May so not long to work on it!).

---

Problems

Feel free to help me improve!

Antigravity produces a vector that osscilates when moving arround a minima. Is there a method of "smoothing" this so that I can get a straffing movement over the minima?

• Do you mean like you get stuck vibrating in one spot because it is the "safest"? I found the idea of "pulse" points to help a lot when I was working with AntiGravityMovement - basically, the strength goes up and down over time. In the end, I find MinimumRiskMovement to be a lot easier to work with, but I'm no melee expert and a lot of great melee bots do use AntiGravity. Maybe Florent or ABC could chime in here. -- Voidious

• In X2 I dont realy have this problem since I have forces on lots of moving things (not only bots, but also bullets). It adds enough variation so I dont get stuck too much. The only times I get stuck is when there are 4 or less bots and no one is firing at me, I guess that you could add gravity points with a small random strength around you. -- Florent

Yes I tend to start vibrating at a point. I have forces from bots for distance and for staying tangential to them so it's not as bad as without. Putting grav points on bullets was an idea I had, but currently I am scanning the whole battle field so I'm a bit worred about inaccuracies. -- BigPete

• You could try putting a very small grav point at the last location you visited. I did some experiments with anti grav a while ago and that's what fixed it for me. --wcsv

I attempted to do that, but found that I just went straight forward. Could try it again with an even smaller value.

• Do you always have a force for staying tangential to the bots? I think it would help your movement if you only use thrm for bots likely to be firing at you. -- Florent

I use an inverse-square law to find the size of the tangential force. As a result, any bots that are far away do not contribute much to the force. I was thinking about including the remaining power of the robot.

---

Hi, welcome to the Robowiki! Someone with a thorough knowledge of mathematics must be able to write some pretty competitive bots! Succes.

Two tips: when you enter a bot on the RoboRumble/Participants you must also enter the verions of the robot (as listed in the bot.properties file and separated from the bot-name with a space). Second, links between pages on the Robowiki are given by double '[' and ']' things (i don't know the name). Greetings --Loki

Thanks Loki. Revising for exams at the moment but when that is finished will devote some more time to it. Quite a lot of my degree seems applicable to it so should be interesting to see what I can bring to you guys. (Then again, looking at the performance of some of your bots you guys don't need any help!) -- BigPete

Welcome from me, too. (I also need to be studying for exams right now...) While it's true that a lot of the Robocode concepts are pretty well developed, I think there's always room for improvement. Some areas, like teams and (maybe) PerceptualBots, are still largely unexplored, too. This wiki has a ton of information, and there are quite a few OpenSource bots, if that interests you. Best of luck with your tank(s). -- Voidious

By the way, if you set your name under "Preferences", it will show your name correctly on the Changes page and at the bottom of pages you edit. -- Voidious