Code center > Suggestions

League stats tests

<< < (4/5) > >>

Numsgil:
My only concern is removing potentially complex strategies from being valid.  Imagine a bot species that gangs up in the first 100 cycles specifically so it can destroy the opposition before it gains a foot hold.

Or imagine the difference in strategy between PY's praetor bot and Ymir.  Usually it ends up being Ymir vs. about 300 Praetor.  Yet (at least when Ymir was made) Ymir would always win.

I would cut veg's starting nrg to, say, 1/4 of what it is now, and place down 4 times as many vegs.  Then, when vegs repop, repop them with the current nrg settings.  I think that should help alleviate any inequitable starting conditions slightly.

Jez:
I'm still gonna (mostly) side with Nums, IMO;

1. Standardise starting postions
Leads to bot designers exploiting this fact.

2. Increase bot population at start
There is still the same amount of energy being introduced to the system so I imagine it would weigh the scales in favour of the conservative bots. (Not 100% sure this is right or it's the wrong way to do)

3. Flush of veg followed by standardisation
Quick repro bots can take advantage of this, also bot designers.

4. More veg with less energy
More uniform enviroment, it is the randomness of the current enviroment that causes these problems, I vote for this. (Also harder for symbots to hog all the veg)

Not saying that I won't argue with, but I will accept the majority opinion on this. Reducing the 3% random 'bot wins because of start position' would be good.

abyaly:

--- Quote from: Numsgil ---This is a "feature".  Basically, the "wrapping" from left to right or up to down is really just a teleporter.  And to prevent problems from having multibots straddle these borders, the entire multibot is teleported.
--- End quote ---
What problems is this preventing? Because multibots tend to act screwy along the edges the way it is now. It would also be nice to have the option of a multibot that wraps around the entire world.

Endy:
I think it's more of a programming issue than anything else. I'd imagine it would be hard to make the ties angle correctly. Not sure if the new version does it but there used to be alot of bugs associated with it. Most of the time I'll run non-torroidal to prevent them.

abyaly:
On a related note, are leages going to move to 2.4x or has it been decided this wont happen?

Navigation

[0] Message Index

[#] Next page

[*] Previous page