Author Topic: Lionfish 6+ on IM  (Read 2850 times)

Offline MacadamiaNuts

  • Bot Destroyer
  • ***
  • Posts: 273
    • View Profile
Lionfish 6+ on IM
« on: December 15, 2007, 07:46:55 PM »
I finally got LF evolution & teaching fixed. It didn't successfully cross the internet barrier yet, but here it has run to 30th generation already against stickerbush veggies and I've got some noticeable changes.

971: the mass needed for reproduction was 2 and now it's 3 on average.

972: while vanilla bots turn using all eyes by a factor of 5 (against -4 -3 -2 -1 - 1 2 3 4), half I sampled were using a factor of ~16. The sim is bigger than 32k x 32k.

973: they dropped a bit of max speed. It was 100 and now it's ~85.

974: the % they gave to their offspring was 50 and I had bots that went on both directions, from 44 to 77.

975: no changes on venom: ~100 units max.

976: all but one were producing a bit more poison, from 100 to ~110.

977: no changes on slime: ~300 units max.

978: most of them were producing 120-140 shell units max, when they started with 100 only.

979: no big changes to the body management threshold, started with 5000 and all were ~4955.

980: the extra width for eye1 and eye9 was 30 and now there's values in the range 30 - 5.

981: the amount stored for body feeding was -500 and now it's ~-525.

982: the vshoot was 10, and now most have deactivated virus shooting.  

983: the settling threshold started at 30 and now was all through the range 30-9.

I sampled only bots that were at least 300 cycles old. If the birth tie breaks early they default part of the racial memory, but when they find an older bot -a survivor- they copy its values with some random variation.
« Last Edit: December 15, 2007, 07:54:22 PM by MacadamiaNuts »
Sometimes you win, and sometimes you lose...