[Icfp04-discuss] multiple ants in one game

[Clive Gifford]

From: "Eijiro Sumii" <sumii at saul.cis.upenn.edu>

> > Actually, it would be pretty easy to add to my simulator as it doesn't
have
> > any fancy optimisations, etc. But what would the modified martial arts
rule
> > be? Does an ant die when surrounded by 5+ ants of any foe, or must they
all
> > be from the same foe?
>
> Hmm - good point.  I like both, but might prefer the latter to prevent
> "accidents".  What do you think?
>
> Besides, what _color_ do you like for the thrid ants, the fourth,
> etc.?:)

Yeah - I also liked the idea that killing ants should all be from the same
colour so it was a more likely to be a planned event. However, to be a more
accurate representation of the real world, I suspect it should really be any
kind of ant except our own. Note I haven't actually researched this aspect
of ant behaviour though!

Anyway, I've tweaked my simulator along these lines but that has apparently
also introduced some new bugs (in addition to the new ants and old bugs), so
a bit more work is required... :-)

As for COLOURS, I went with Red, Black, White, Green, Blue, and Yellow for
no particular reason. (Note to self: more research needed to verify colours
of common ant species...) Allowing for up to six possible species at once
seemed to be in keeping with hexagonal cells, six markers, perfect numbers,
and so on...

It will be interesting to see how well strategies designed primarily for one
on one competition translate to higher numbers and what difference (if any)
it makes to the relative strengths of the different species. Overcrowding
might be an issue also but changing the size of the world will probably have
an even greater effect.

What was the reasoning behind the 100000 tics per match? I've only
experimented with longer runs a couple of times, but it would be interesting
also to see what strategies might work better over a longer time period.
Tortoise versus the hare and all that.

Clive.