Common Targeting Methods in NanoRumble
Hi, I was wondering how many bots in the NanoRumble use linear or HOT targeting. I don't think SquirmyToad has enough space to have a multimode movement, so if most robots use HOT and linear targeting, I may use stop and go. Else, if most robots use pattern matching, I'll keep the random movement. Thanks!
HOT and linear targeting is very common among nano bots. Only a few (top) nano bots use pattern matching. If you can accept getting beaten by them, use stop and go.
I do not write nano bots but what I would suggest is to add a random factor to the direction of stop'n go after a loss(???) so you will have more consistent score against linear and head-on targeting bots while keeping pattern-matching on check.
I'll see if I have space =) Do you think I should do the random factor after it dies too much or gets hit too much by bullets? I saw BlackWidow changes after two deaths. Thanks!
It is better if you do it with deaths as it is easier to keep track of. Also, the number of bullets is a little inconsistent as it depends on how well the enemy dodges your shots.
Oh! I just had an idea for a melee robot. It would use fluid movement and stop and go, since most melee bots use HOT, linear, or circular targeting! I think I'll make a nanobot with the stop and go, and then expand it into a micro melee robot.
I think my robot will add the random factor after 3 deaths. I may need to add some energy management to stop it from disabling itself and losing rounds. Thanks!
You can always use RoboRunner or RoboJogger to fine tune those parameters. I have no idea about melee strategies though, the best I could get is 127th place which I am sure WhiteFang which is a 1v1 only robot would easily pass.
You do not have permission to edit this page, for the following reasons:
You can view and copy the source of this page.
Return to Thread:Talk:SquirmyToad/Common Targeting Methods in NanoRumble/reply (9).
- I would say top 10 as you are already close to top 20.
- I just realized how much I forgot about Robocode API. I had to take a look at WF's radar code to code one that works =)
- Anyways, get prepared for Partial(Lame name, I know) entering the minirumble.
I was unable to find enough space to add stop and go, but I was able to fit in the Musashi trick! It uses the Musashi trick until it dies 3 times, and then goes to random movement. Also, the random movement has been tweaked so that it changes direction more often to throw off some of the linear targeters. Thanks for the help!