Bots and Simulations > Bot Tavern

Bot Transition- 2.37.6 to 2.43

<< < (6/7) > >>

Numsgil:
To get to the very top of the leagues, you usually need to invent some brand spanking new strategy that's never been seen before.  It's hard to be the best just by using the more traditional methods.

For example, suppose you made a bot that works as a community to hide all the veggies in the corner of the arena, in an attempt to starve out the other bot.  Or if you made a bot that slowly builds a gigantic multibot web that makes it impossible for the other bot to effectively kill (since you'd be using sharenrg).  Or maybe something like Light's Helios, back before it was nerfed (or rather, the program was unnerfed, but same thing).  It used the max number of ties so you couldn't tie feed from it.

All these sorts of ideas require a great deal of programming acuity, and a great deal of imagination.  But that's about the only way you can get to the top anymore.  Hence why most of my "recent" bots haven't been strong contendors.

Jez:
Ooh, ooh, a challenge.

If I had used the virus beating technique you [Stranger] are using I wonder if Blue could get to the top of F1 as well as F2...

T. Mosq is a well written bot so don't feel bad if you have trouble beating it.

T. Mosq, Blue and Bubbles are the three top recent entries into the league.
You have a choice now; continue tinkering with your bot to see if you can fine tune it a bit more or enter it in the league and write a new bot from scratch using what you have learnt so far to write a better bot.  

EDIT
I thought Callidus was a virus bot but it seems not, I shall change it's bestiary entry if needed and update its title for the next league change.  

Jez:
I have to admit Nums, you are right, it is much easier to come up with a  "brand spanking new strategy that's never been seen before" than write a bot that is an 'old style supremist'. Either that or I get distracted too easily!

Stranger, if you want a 'not before used strategy', for viruses at least, send me a PM. I have thought of one that you probably aren't aware of, especially as  it's not in the wiki sysvars yet. I like the challenge of limiting myself to F2 bots so it's not something I will use.

One of Light's bot's is, only for now, stopping a basic prototype bot of mine climbing to a new first place in F2...

PS
Having more than 300 eye's just stops Blue fooling your bot, it won't fool Blue's ID system...

Endy:
I'd recommend against using any of the refvars as identificiation, it's too easy to flub it with a simple "eye virus" a slight difference of a single command can cause your bots to attack their own.

cond
start
*.eye5
*.thisgene .mkvirus store
.vshoot inc
stop

I've been using a combination of in/out and .memloc/.memval with great success. It's extremly difficult to copy all of these simultaneously, while another bot is attacking.

Oh yeah you might want to switch to using ".shoot dec" vice "-1 .shoot store" and similarly ".fixpos dec" it's a pretty decent nrg saver(cost of .25 versus 1 unit of nrg).

Light:
I've got an experimental bot that uses a different id system that you might be interested in looking at, its not in the bestiary but will add as an attachment to this post

Navigation

[0] Message Index

[#] Next page

[*] Previous page