You don't need to use a special signal combination. if each bot just has a copy of their teammate's code they can check for every move whether each of the other bots is behaving as their teammate should. By turn 10 you should have a lock on the correct bot.
It would also cause a huge performance load, because you have to check all possible moves for each single bot to see if it was like you expect and you still need to keep tracks of bots
I'd be a little bit afraid about this situation: your two bots divide a lot, say 60 wins among themselves, but there will be someone else who beats them both because he has, let's say 35 wins.
Yes, that's a possibility. You are relying on the fact that there is not a single other strong bot. But also consider the average size of a win. Most victories are by a fairly slim margin. Usually the 3rd strongest bot + .85% of the 4th strongest bot are still going to be able tot ake 1st place.
Where would you draw the line though Redwolf? If I make a bot that is capable of collecting coins and identifying copies of itself and then feeding itself to the strongest instance of itself at the end of the round can I submit that bot 10 times?
There's a lot of potential in this competition for adding bots that will influence the effectiveness of existing bots in various ways. For instance the more bots you add following any single strategy the less effective that strategy becomes. I think someone had suggested adding territorial bots to try and reduce the effectiveness of Big King Little Hill at some point which I think almost certainly would have been quite effective. This is likely to be an issue in future challenges as well.
I don't think anyone went out of their way to "sabotage" other people's bots in this challenge, but the rules currently don't preclude this possibility.
Well, but if I'm to run a challenge, then I'll have a look at how the others have done it. (And I don't mean to badmouth Redwolf, he's done a great job. I just mean that it could be an even better job.)
Anyways I still think that this is a good experience, for the next challenge he will have lots of rule ideas from our suggestions and experience of this challenge
Well, a bad thing is that your rules were really messy, and ever changing. Also you didn't have a controller made before announcing the challenge, which resulted in the infamous dozens of controllers floating all around.
But the main thing is that a lot of people had a good fun with the challenge.
I must admit that when I was posting my only one code-golf challenge, I didn't put it in the sandbox at all. I just had a look at how the well-upvoted challenges looked like and wrote something in a similar vein.
