Darwinbots Forum
Bots and Simulations => Bestiary => Mutations => Topic started by: Botsareus on April 19, 2016, 04:20:02 PM
-
I am about to make some changes and run this experiment again. My 3 month version is currently stuck. That is why I am making major changes and trying stuff. Funny how more coherent mutations and more intuitive evolution logic seems to produce results more slowly. The one posted here simply maxed out the oscillation reduction before I had enough of it. Just look at the log:
Evolving robot never changed, increasing difficulty. 04-10-2016 14:37:05
Evolving robot lost, decreasing difficulty. 04-10-2016 14:54:43
Evolving robot changed, testing robot. 04-10-2016 15:07:29
Evolving robot lost the test, increasing difficulty. 04-10-2016 16:07:53
Evolving robot changed, testing robot. 04-10-2016 16:41:26
Evolving robot won all tests, setting up stage 1 04-10-2016 17:53:41
Evolving robot lost, decreasing difficulty. 04-10-2016 18:56:14
Evolving robot lost, decreasing difficulty. 04-10-2016 19:36:22
Evolving robot lost, decreasing difficulty. 04-10-2016 19:41:24
Evolving robot changed, testing robot. 04-10-2016 20:14:40
Robot is currently under test against stage 0 04-10-2016 21:23:16
Evolving robot won all tests, setting up stage 2 04-10-2016 21:45:28
Evolving robot lost, decreasing difficulty. 04-10-2016 21:52:16
Evolving robot lost, decreasing difficulty. 04-10-2016 22:18:23
Evolving robot lost, decreasing difficulty. 04-10-2016 22:42:51
Evolving robot lost, decreasing difficulty. 04-10-2016 23:10:54
Evolving robot lost, decreasing difficulty. 04-11-2016 00:29:24
Evolving robot lost, decreasing difficulty. 04-11-2016 01:06:29
Evolving robot lost, decreasing difficulty. 04-11-2016 01:47:54
Evolving robot lost, decreasing difficulty. 04-11-2016 02:25:45
Evolving robot lost, decreasing difficulty. 04-11-2016 03:32:09
Evolving robot lost, decreasing difficulty. 04-11-2016 04:43:37
Evolving robot lost, decreasing difficulty. 04-11-2016 05:35:07
Evolving robot changed, testing robot. 04-11-2016 05:52:17
Robot is currently under test against stage 1 04-11-2016 06:14:52
Evolving robot won all tests, setting up stage 3 04-11-2016 06:44:41
Evolving robot lost, decreasing difficulty. 04-11-2016 07:17:20
Evolving robot lost, decreasing difficulty. 04-11-2016 07:19:00
Evolving robot lost, decreasing difficulty. 04-11-2016 08:19:30
Evolving robot lost, decreasing difficulty. 04-11-2016 09:31:35
Evolving robot lost, decreasing difficulty. 04-11-2016 10:07:18
Evolving robot lost, decreasing difficulty. 04-11-2016 11:21:51
Evolving robot lost, decreasing difficulty. 04-11-2016 12:59:12
Evolving robot lost, decreasing difficulty. 04-11-2016 13:55:50
Evolving robot lost, decreasing difficulty. 04-11-2016 15:33:49
Evolving robot lost, decreasing difficulty. 04-18-2016 14:20:52
Evolving robot lost, decreasing difficulty. 04-18-2016 15:49:37
Evolving robot lost, decreasing difficulty. 04-18-2016 17:18:21
Evolving robot lost, decreasing difficulty. 04-18-2016 19:06:41
Evolving robot changed, testing robot. 04-18-2016 21:08:34
Evolving robot lost the test, increasing difficulty. 04-18-2016 21:09:32
Evolving robot changed, testing robot. 04-18-2016 21:22:20
Evolving robot lost the test, increasing difficulty. 04-18-2016 21:24:11
Evolving robot lost, decreasing difficulty. 04-18-2016 21:58:17
Evolving robot lost, decreasing difficulty. 04-18-2016 23:20:27
Evolving robot lost, decreasing difficulty. 04-19-2016 00:53:18
Evolving robot lost, decreasing difficulty. 04-19-2016 01:31:47
Evolving robot lost, decreasing difficulty. 04-19-2016 02:12:36
Evolving robot lost, decreasing difficulty. 04-19-2016 03:00:16
Evolving robot lost, decreasing difficulty. 04-19-2016 03:49:40
Evolving robot lost, decreasing difficulty. 04-19-2016 04:35:52
Evolving robot lost, decreasing difficulty. 04-19-2016 06:19:57
Evolving robot lost, decreasing difficulty. 04-19-2016 08:11:34
Evolving robot lost, decreasing difficulty. 04-19-2016 09:14:19
Evolving robot lost, decreasing difficulty. 04-19-2016 14:02:59
Evolving robot lost, decreasing difficulty. 04-19-2016 14:49:54
Evolving robot lost, decreasing difficulty. 04-19-2016 14:50:18
'Evolved from zerobot in about a week to test my new mutation algorithms
'Works best if you cap body at 2.5K in global setting cheating prevention and
'Enable delta2 mutations in global settings, then disable mutations species tab. (Alternatively you can select disable all mutations via the mutations tab.)
'#generation: 41
'#mutations: 2235
*.fixang <
*430 *.eye1width *.trefvelyourdx not
1 -16 *.aimleft ^ *.refbody .reftie store
store
*.shoot *.eye3 add floorstore
<
*.focuseye and
and
xor
angle -5 *.dn pyth *.dx *.tin2 0 *.setboy * -5 stop
cond
*.in7 -3 *.tmemval 11 angle dec
store
*.tin4 *.tin7 *.mydx dupbool
logx *.ypos *.paralyzed *.tielen1 ~ *.memloc -12 add 1 and
start
clearbool
substore
*.vel .shoot dec
*.ypos store
*.refage ^ *.genes .in3 store
*.in8 *.shup *.dx *.vel *31 add *.tie *.trefvelmysx *.memval
'''''''''''''''''''''''' Gene: 1 Ends at position 78 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 2 Begins at position 79 '''''''''''''''''''''''
cond
store
>
*.timer <
*.refup *391 *.refveldn dist *33 addstore
*.myeye *.tout4 *.myvenom *.tin5 .aimleft store
store
837 *.venom rnd *.fixlen .aimleft store
*.refsx 2 3 *.repro 5 *.eye8 *.eye1 *.focuseye -- -9 store
*0 *.vtimer clearbool
!~=
*.tielen2 pow 24 !=
swapbool
*.trefage *.refvelup add .slime store
%=
*.refveldx drop *.shareshell *.sexrepro *.refpoison 7 *.ypos 1 27 27 * *.mkvenom *.trefsx ~ not
>> substore
++ *.refbody *.repro *.eye7width -16 *.shareslime
'''''''''''''''''''''''' Gene: 2 Ends at position 151 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 3 Begins at position 152 '''''''''''''''''''''''
cond
'''''''''''''''''''''''' Gene: 3 Ends at position 152 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 4 Begins at position 153 '''''''''''''''''''''''
cond
div angle *.mrepro <=
*.poisoned div *.aimright 0 angle angle *.hitang pyth *.refup .velsx multstore
*.tout4 store
*.availability 301 *.setboy *.eye5
'''''''''''''''''''''''' Gene: 4 Ends at position 174 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 5 Begins at position 175 '''''''''''''''''''''''
cond
699 | -17 0 ~ .tin8 store
*.mkshell *.refveldx *.eye7width -22 *.eye4width *.myup *.tin2 *.out9 or
store
*.eye3dir *.shell 768 store
-11 0 add and
29 *.aimright 41 store
9 *.eye7width 2 !=
7 9 else
.dn store
mod *.mkvirus *.tin5 11
'''''''''''''''''''''''' Gene: 5 Ends at position 217 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 6 Begins at position 218 '''''''''''''''''''''''
start
*.aimleft store
sub store
%=
!=
1 *.trefaim %=
xor
*.mkvirus *.up *.chlr 12 *.dx -6 dup *.in2 store
rndstore
0 !=
.venom store
.up store
divstore
--
'''''''''''''''''''''''' Gene: 6 Ends at position 246 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 7 Begins at position 247 '''''''''''''''''''''''
start
cos 8 add <
*.tout6 -- *.tout3 -10 mod angle >=
*.out1'''''''''''''''''''''''' Gene: 7 Ends at position 259 '''''''''''''''''''''''
'#hash: Go%}Jum'p7.3^(6y;!Zj
'#tag:zbstage386.txt
-
I am about to make some changes and run this experiment again. My 3 month version is currently stuck. That is why I am making major changes and trying stuff. Funny how more coherent mutations and more intuitive evolution logic seems to produce results more slowly. The one posted here simply maxed out the oscillation reduction before I had enough of it. Just look at the log:
Evolving robot never changed, increasing difficulty. 04-10-2016 14:37:05
Evolving robot lost, decreasing difficulty. 04-10-2016 14:54:43
Evolving robot changed, testing robot. 04-10-2016 15:07:29
Evolving robot lost the test, increasing difficulty. 04-10-2016 16:07:53
Evolving robot changed, testing robot. 04-10-2016 16:41:26
Evolving robot won all tests, setting up stage 1 04-10-2016 17:53:41
Evolving robot lost, decreasing difficulty. 04-10-2016 18:56:14
Evolving robot lost, decreasing difficulty. 04-10-2016 19:36:22
Evolving robot lost, decreasing difficulty. 04-10-2016 19:41:24
Evolving robot changed, testing robot. 04-10-2016 20:14:40
Robot is currently under test against stage 0 04-10-2016 21:23:16
Evolving robot won all tests, setting up stage 2 04-10-2016 21:45:28
Evolving robot lost, decreasing difficulty. 04-10-2016 21:52:16
Evolving robot lost, decreasing difficulty. 04-10-2016 22:18:23
Evolving robot lost, decreasing difficulty. 04-10-2016 22:42:51
Evolving robot lost, decreasing difficulty. 04-10-2016 23:10:54
Evolving robot lost, decreasing difficulty. 04-11-2016 00:29:24
Evolving robot lost, decreasing difficulty. 04-11-2016 01:06:29
Evolving robot lost, decreasing difficulty. 04-11-2016 01:47:54
Evolving robot lost, decreasing difficulty. 04-11-2016 02:25:45
Evolving robot lost, decreasing difficulty. 04-11-2016 03:32:09
Evolving robot lost, decreasing difficulty. 04-11-2016 04:43:37
Evolving robot lost, decreasing difficulty. 04-11-2016 05:35:07
Evolving robot changed, testing robot. 04-11-2016 05:52:17
Robot is currently under test against stage 1 04-11-2016 06:14:52
Evolving robot won all tests, setting up stage 3 04-11-2016 06:44:41
Evolving robot lost, decreasing difficulty. 04-11-2016 07:17:20
Evolving robot lost, decreasing difficulty. 04-11-2016 07:19:00
Evolving robot lost, decreasing difficulty. 04-11-2016 08:19:30
Evolving robot lost, decreasing difficulty. 04-11-2016 09:31:35
Evolving robot lost, decreasing difficulty. 04-11-2016 10:07:18
Evolving robot lost, decreasing difficulty. 04-11-2016 11:21:51
Evolving robot lost, decreasing difficulty. 04-11-2016 12:59:12
Evolving robot lost, decreasing difficulty. 04-11-2016 13:55:50
Evolving robot lost, decreasing difficulty. 04-11-2016 15:33:49
Evolving robot lost, decreasing difficulty. 04-18-2016 14:20:52
Evolving robot lost, decreasing difficulty. 04-18-2016 15:49:37
Evolving robot lost, decreasing difficulty. 04-18-2016 17:18:21
Evolving robot lost, decreasing difficulty. 04-18-2016 19:06:41
Evolving robot changed, testing robot. 04-18-2016 21:08:34
Evolving robot lost the test, increasing difficulty. 04-18-2016 21:09:32
Evolving robot changed, testing robot. 04-18-2016 21:22:20
Evolving robot lost the test, increasing difficulty. 04-18-2016 21:24:11
Evolving robot lost, decreasing difficulty. 04-18-2016 21:58:17
Evolving robot lost, decreasing difficulty. 04-18-2016 23:20:27
Evolving robot lost, decreasing difficulty. 04-19-2016 00:53:18
Evolving robot lost, decreasing difficulty. 04-19-2016 01:31:47
Evolving robot lost, decreasing difficulty. 04-19-2016 02:12:36
Evolving robot lost, decreasing difficulty. 04-19-2016 03:00:16
Evolving robot lost, decreasing difficulty. 04-19-2016 03:49:40
Evolving robot lost, decreasing difficulty. 04-19-2016 04:35:52
Evolving robot lost, decreasing difficulty. 04-19-2016 06:19:57
Evolving robot lost, decreasing difficulty. 04-19-2016 08:11:34
Evolving robot lost, decreasing difficulty. 04-19-2016 09:14:19
Evolving robot lost, decreasing difficulty. 04-19-2016 14:02:59
Evolving robot lost, decreasing difficulty. 04-19-2016 14:49:54
Evolving robot lost, decreasing difficulty. 04-19-2016 14:50:18
'Evolved from zerobot in about a week to test my new mutation algorithms
'Works best if you cap body at 2.5K in global setting cheating prevention and
'Enable delta2 mutations in global settings, then disable mutations species tab. (Alternatively you can select disable all mutations via the mutations tab.)
'#generation: 41
'#mutations: 2235
*.fixang <
*430 *.eye1width *.trefvelyourdx not
1 -16 *.aimleft ^ *.refbody .reftie store
store
*.shoot *.eye3 add floorstore
<
*.focuseye and
and
xor
angle -5 *.dn pyth *.dx *.tin2 0 *.setboy * -5 stop
cond
*.in7 -3 *.tmemval 11 angle dec
store
*.tin4 *.tin7 *.mydx dupbool
logx *.ypos *.paralyzed *.tielen1 ~ *.memloc -12 add 1 and
start
clearbool
substore
*.vel .shoot dec
*.ypos store
*.refage ^ *.genes .in3 store
*.in8 *.shup *.dx *.vel *31 add *.tie *.trefvelmysx *.memval
'''''''''''''''''''''''' Gene: 1 Ends at position 78 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 2 Begins at position 79 '''''''''''''''''''''''
cond
store
>
*.timer <
*.refup *391 *.refveldn dist *33 addstore
*.myeye *.tout4 *.myvenom *.tin5 .aimleft store
store
837 *.venom rnd *.fixlen .aimleft store
*.refsx 2 3 *.repro 5 *.eye8 *.eye1 *.focuseye -- -9 store
*0 *.vtimer clearbool
!~=
*.tielen2 pow 24 !=
swapbool
*.trefage *.refvelup add .slime store
%=
*.refveldx drop *.shareshell *.sexrepro *.refpoison 7 *.ypos 1 27 27 * *.mkvenom *.trefsx ~ not
>> substore
++ *.refbody *.repro *.eye7width -16 *.shareslime
'''''''''''''''''''''''' Gene: 2 Ends at position 151 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 3 Begins at position 152 '''''''''''''''''''''''
cond
'''''''''''''''''''''''' Gene: 3 Ends at position 152 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 4 Begins at position 153 '''''''''''''''''''''''
cond
div angle *.mrepro <=
*.poisoned div *.aimright 0 angle angle *.hitang pyth *.refup .velsx multstore
*.tout4 store
*.availability 301 *.setboy *.eye5
'''''''''''''''''''''''' Gene: 4 Ends at position 174 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 5 Begins at position 175 '''''''''''''''''''''''
cond
699 | -17 0 ~ .tin8 store
*.mkshell *.refveldx *.eye7width -22 *.eye4width *.myup *.tin2 *.out9 or
store
*.eye3dir *.shell 768 store
-11 0 add and
29 *.aimright 41 store
9 *.eye7width 2 !=
7 9 else
.dn store
mod *.mkvirus *.tin5 11
'''''''''''''''''''''''' Gene: 5 Ends at position 217 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 6 Begins at position 218 '''''''''''''''''''''''
start
*.aimleft store
sub store
%=
!=
1 *.trefaim %=
xor
*.mkvirus *.up *.chlr 12 *.dx -6 dup *.in2 store
rndstore
0 !=
.venom store
.up store
divstore
--
'''''''''''''''''''''''' Gene: 6 Ends at position 246 '''''''''''''''''''''''
'''''''''''''''''''''''' Gene: 7 Begins at position 247 '''''''''''''''''''''''
start
cos 8 add <
*.tout6 -- *.tout3 -10 mod angle >=
*.out1'''''''''''''''''''''''' Gene: 7 Ends at position 259 '''''''''''''''''''''''
'#hash: Go%}Jum'p7.3^(6y;!Zj
'#tag:zbstage386.txt
Interesting I might test it out.
You know what'd be really useful.
A "Code De-Jargon-ifier".
I mean, the code that is evolved is probably more efficient, computing-capacity-wise, and probably less fragile, but It should be possible for a program to re-organise code to make it 'more' readable.
OFFTOPIC: In a PM or something do you think you could teach me this code that breaks the "Cond [conditions] Start [actions] Stop" DNA setup? (I believe It was referred to as inline conditioning?)
Then I might actually be able to understand fruitflies.
-
Speaking of fruitfly, I am actually spending a core to evolve it >:D in survival evo. Just for a kick because I think I fixed what was wrong with my algorithm.
-
Speaking of fruitfly, I am actually spending a core to evolve it >:D in survival evo. Just for a kick because I think I fixed what was wrong with my algorithm.
Hey bot, think you might be able to write me up some code for like, just moving in an indefinitely long seasnake?
I don't need any handling of food or stuff, just an as simple to read bit of code for at the birth event, adding to the snake, and then moving along like that.
-
Just ran the experiment again. It was good for a while but as soon as it got stable it began to royally lose again:
Evolving robot changed, testing robot. 04-23-2016 10:13:46
Evolving robot won all tests, setting up stage 1 04-23-2016 11:28:50
Evolving robot lost, decreasing difficulty. 04-23-2016 11:41:24
Evolving robot lost, decreasing difficulty. 04-23-2016 12:18:49
Evolving robot changed, testing robot. 04-23-2016 12:23:19
Robot is currently under test against stage 0 04-23-2016 13:12:37
Evolving robot won all tests, setting up stage 2 04-23-2016 14:00:07
Evolving robot lost, decreasing difficulty. 04-23-2016 14:24:16
Evolving robot lost, decreasing difficulty. 04-23-2016 14:41:07
Evolving robot lost, decreasing difficulty. 04-23-2016 15:16:17
Evolving robot changed, testing robot. 04-23-2016 15:44:48
Evolving robot lost the test, increasing difficulty. 04-23-2016 16:26:26
Evolving robot changed, testing robot. 04-23-2016 16:59:48
Evolving robot lost the test, increasing difficulty. 04-23-2016 17:30:58
Evolving robot lost, decreasing difficulty. 04-23-2016 18:02:56
Evolving robot lost, decreasing difficulty. 04-23-2016 18:42:12
Evolving robot lost, decreasing difficulty. 04-23-2016 18:55:59
Evolving robot lost, decreasing difficulty. 04-23-2016 19:00:31
Evolving robot changed, testing robot. 04-23-2016 19:19:27
Robot is currently under test against stage 1 04-23-2016 19:54:48
Evolving robot lost the test, increasing difficulty. 04-23-2016 20:22:48
Evolving robot lost, decreasing difficulty. 04-23-2016 20:26:39
Evolving robot lost, decreasing difficulty. 04-23-2016 20:39:04
Evolving robot changed, testing robot. 04-23-2016 21:55:32
Robot is currently under test against stage 1 04-23-2016 22:25:11
Evolving robot won all tests, setting up stage 3 04-23-2016 22:56:02
Evolving robot lost, decreasing difficulty. 04-23-2016 23:34:42
Evolving robot changed, testing robot. 04-23-2016 23:40:43
Robot is currently under test against stage 2 04-24-2016 00:12:58
Robot is currently under test against stage 0 04-24-2016 00:55:35
Evolving robot won all tests, setting up stage 4 04-24-2016 01:08:41
Evolving robot lost, decreasing difficulty. 04-24-2016 01:45:26
Evolving robot lost, decreasing difficulty. 04-24-2016 02:33:19
Evolving robot lost, decreasing difficulty. 04-24-2016 03:03:06
Evolving robot lost, decreasing difficulty. 04-24-2016 03:18:10
Evolving robot lost, decreasing difficulty. 04-24-2016 03:26:51
Evolving robot lost, decreasing difficulty. 04-24-2016 04:13:57
Evolving robot lost, decreasing difficulty. 04-24-2016 04:29:56
Evolving robot lost, decreasing difficulty. 04-24-2016 04:44:16
Evolving robot changed, testing robot. 04-24-2016 05:04:34
Evolving robot lost the test, increasing difficulty. 04-24-2016 05:05:07
Evolving robot lost, decreasing difficulty. 04-24-2016 05:56:47
Evolving robot lost, decreasing difficulty. 04-24-2016 06:35:48
Evolving robot lost, decreasing difficulty. 04-24-2016 07:06:20
Evolving robot lost, decreasing difficulty. 04-24-2016 07:22:44
Evolving robot lost, decreasing difficulty. 04-24-2016 07:44:48
Evolving robot lost, decreasing difficulty. 04-24-2016 08:23:36
Evolving robot lost, decreasing difficulty. 04-24-2016 09:22:06
Evolving robot lost, decreasing difficulty. 04-24-2016 10:48:09
Evolving robot lost, decreasing difficulty. 04-24-2016 11:36:06
Evolving robot lost, decreasing difficulty. 04-24-2016 11:40:22
Evolving robot lost, decreasing difficulty. 04-24-2016 12:24:32
Evolving robot lost, decreasing difficulty. 04-24-2016 13:15:06
Evolving robot lost, decreasing difficulty. 04-24-2016 13:45:21
Evolving robot lost, decreasing difficulty. 04-24-2016 14:03:46
Evolving robot lost, decreasing difficulty. 04-24-2016 14:49:07
Evolving robot lost, decreasing difficulty. 04-24-2016 16:04:39
Evolving robot lost, decreasing difficulty. 04-24-2016 16:44:16
This is rather weird.
edit:
The only difference between this and my 3 month version is I do not fittest or handycap for body. Plus the fix for CopyError and CopyError2 that was minor.
I decided to get rid of body because of big birthas and robots like fruitfly benefit from having as little body as possible.
-
The only difference between this and my 3 month version is I do not fittest or handycap for body. Plus the fix for CopyError and CopyError2 that was minor.
I decided to get rid of body because of big birthas and robots like fruitfly benefit from having as little body as possible.
Tiny changes make big impacts. E.G. Charity, or Trigonometry, or Politics/Law.
Also, When you say you decided to get rid of body....do you mean you've removed body from DB (because it was quite a fun thing to utilize).
Next we might as well remove viruses because some robots benefit from having viruses to shoot at other bots :P
-
No, body is still there. I am just experimenting with not having body in fitness calc. so population is favored more.
I actually do not mind viruses because after talking to a med student and some proofs of concept we figured out that viruses could be beneficial to evolution. They are like a cheap way to do sexrepro. But if you hate them that much you can always run F2 settings. A note that you can not tie feed in F2 though.
-
No, body is still there. I am just experimenting with not having body in fitness calc. so population is favored more.
I actually do not mind viruses because after talking to a med student and some proofs of concept we figured out that viruses could be beneficial to evolution. They are like a cheap way to do sexrepro. But if you hate them that much you can always run F2 settings. A note that you can not tie feed in F2 though.
Surely in fitness calc factoring in body is beneficial, (same for league victory determination), as you could have so to speak a "tank", or really fat, strong bot, which would lose against 10 smaller weaker cancerous bots in the calculations.
-
[Bump]
I need a couple more days to figure it out. Maybe you are right Spike.
edit: A note that leagues are strictly based on population. And that is really the main idea behind this change. Can you show me any 'Tanks' on top of the leader board? :sarcastic:
Small robots working together are always stronger than large robots working alone. When you finally figure out fruitfly code you will understand what I mean.
-
[Bump]
I need a couple more days to figure it out. Maybe you are right Spike.
edit: A note that leagues are strictly based on population. And that is really the main idea behind this change. Can you show me any 'Tanks' on top of the leader board? :sarcastic:
Small robots working together are always stronger than large robots working alone. When you finally figure out fruitfly code you will understand what I mean.
Theoretically speaking, if a tank was made strong enough. It could be 'better'. It would still lose though, but its more to do with the bias probably effecting bots ranking in the middle of the leaderboard (say 6th, or 10th place).
It could also mean that a slightly less small bot than fruitflies, might possibly top fruitflies (and it generally adds a new element of mystery to the results, I think everyone knows that fruitflies will always win unless some of the mechanics change).
-
I ended up reverting the fittest code as you suggested. It really has little impact on how much the robots seem to lose in survival mode and if anything fittest can be configured to only favor population so that solves that little problem. However, I am still keeping my fitness function based on energy for now. I might reverting it as well as a last resort. Lets see what happens and sorry for the long wait between version drops if I had more computing power it would go allot faster.
-
I ended up reverting the fittest code as you suggested. It really has little impact on how much the robots seem to lose in survival mode and if anything fittest can be configured to only favor population so that solves that little problem. However, I am still keeping my fitness function based on energy for now. I might reverting it as well as a last resort. Lets see what happens and sorry for the long wait between version drops if I had more computing power it would go allot faster.
Ok. Np.
OFFTOPIC: I'm not sure if I mentioned this in a post, but I was thinking, we could try moving to a new, shinier forum (there has to be some fancier one out there now), and then It'll be lovely and clean for any newcomers. This forum could then be a sort of archive.
-
As expected it is working a bit worse even with the effect of body reduced by 3/10. I am not taking body out completely but I might end up reducing it to 1/10. I can also try the other way and revert both the fittest and the survival fitness to normal body.
-
As expected it is working a bit worse even with the effect of body reduced by 3/10. I am not taking body out completely but I might end up reducing it to 1/10. I can also try the other way and revert both the fittest and the survival fitness to normal body.
What about factoring in kills per bot.
So the fittest function records the kills each bot has, till it dies, which that becomes a fixed figure, the average tells you how successful a species is, for individual bots look at their individual kills?
-
Damn, just tried a zerobot run with my better? mutation algorithms robots are just sitting there without doing anything for a long time. I am seriously considering reverting most of my changes and just keeping the bug fixes.
-
Few, it took a while but I got a robot that is moving and spinning. There is hope after all.
-
woot? A robot won a zb round by converting body into energy since body and energy are favored disproportionately. Problem is it died after converting all of its body into energy.
-
Ended up reverting that bit completely and simply adjusting "Find best settings based on" slider to favor population 75% and body+energy 25%.
-
Ended up reverting that bit completely and simply adjusting "Find best settings based on" slider to favor population 75% and body+energy 25%.
Is it living after victory this time round?
-
I just overwrote the handycap with a maxout(150)
If it does not level off at some point I am considering handycaping for body again.
If that does not work I will try to put the reposition code back.
If that does not work, meh.
-
I just overwrote the handycap with a maxout(150)
If it does not level off at some point I am considering handycaping for body again.
If that does not work I will try to put the reposition code back.
If that does not work, meh.
What about unnecessarily overcomplicating the code to factor in practically anything the bot could rely on to win :3?
-
offtopic: Joe Lopez reminds me of neo sometimes. Just sits late nights infront of a computer waiting for the magic answer to escape reality he does not understand.
-
I just overwrote the handycap with a maxout(150)
If it does not level off at some point I am considering handycaping for body again.
If that does not work I will try to put the reposition code back.
If that does not work, meh.
What about unnecessarily overcomplicating the code to factor in practically anything the bot could rely on to win :3?
Yes, I do think having "save epigenetic memory as a delgene" and "periodically reset epigenetic memory" was a total waste of my time but someone higher up the food chain wanted no metadata.
-
Successful tie feeding it seems is nearly impossible to evolve btw. At some point it might be necessary to have an even more specific mutation algorithm to help robots evolve such complex behavior. :P