Creatures Community Chat

Sunday, July 3, 2011

Creatures 3: Intelligence Test

So as I said in my last post I wanted to try running some experiments with Creatures 3, mostly because generations progress faster and despite not quite having the same attachments to the Norns, it's great for experimentation.

Well I started a world on my home computer and decided I'd start by making some Treehuggers do an intelligence test.

Generally an intelligence test involves hatching a group of Norns in an isolated area with no food. The creatures then have to progress through lifts, doorways etc to find a room with food in it. Many people, myself included use the Docking Station workshop as the starting point, forcing the Norns to have to travel through the main connection hub and into the meso before they can find a bite to eat.

All my Treehuggers died before they got out of the workshop. I went home over the weekend and booted up Docking Station on my mother's laptop and tried the experiment again, this time being limited to Chichi Norns. Not only did they all survive but they then proceeded to breed like rabbits.

Finding this difference interesting I decided to do a batch intelligence test of all the different species I had, and decided that comparing the normal genomes to the CFE breeds would be interesting. Here are my results:

Toxic CFE: 5/6
Toxic Non-CFE: 5/6

--------------------------

Chichi CFE:5/6
Chichi Non-CFE:6/6 (Very fast)

--------------------------

Treehugger CFE: 0/6
Treehugger Non-CFE:2/6

--------------------------

Civet: CFE: 5/6
Civet None-CFE: 5/6

--------------------------

Hardman CFE: 6/6 (very fast)
Hardman Non-CFE: 6/6 (Very fast)

--------------------------

Magma CFE: 5/6
Magma Non-CFE: 5/6

--------------------------

Bengal CFE: 6/6
Bengal Non-CFE: 6/6

--------------------------

Bondi CFE: 5/6
Bondi Non-CFE: 5/6

--------------------------

Bruin CFE: 6/6
Bruin Non-CFE: 4/6

--------------------------

Banshee Grendel CFE: 5/6
Banshee Grendel Non-CFE: 5/6

--------------------------

Ettin CFE: 6/6 (Very fast)

--------------------------

Astro/Fallow/Harlequin/Siamese/Zebra CFE: 5/5
Astro/Fallow/Harlequin/Siamese/Zebra Non-CFE: 5/5


I hatched each batch in groups of 6. 3 males and 3 females, with the exception of the Astro/Fallow/Harlequin/Siamese/Zebra, as I believe all these breeds are based on the Chichi genome with different sprites.

The numbers show the survivers vs the number of creatures each run. So "Bruin Non-CFE: 4/6" means 4 out of the 6 Norns survived and made it to the meso.

As you can see there is very little difference in the survival rate between CFE and non-CFE based creatures. The final results are 59/60, with non-CFE Norns being slightly more successful.

Obviously this is a rather small sample size and I'd prefer to do this test with several hundred Norns to compare the CFE vs non-CFE (and maybe I will), but initial results indicate that the CFE genome doesn't give a large survival advantage, at least in terms of finding food.

What I found interesting was with almost every group 5 of the creatures would hurry off to the meso very quickly, leaving one behind who did almost nothing. It just sat there looking sad before it died.

The other interesting and somewhat disheartening thing is the Treehuggers are terrible at this test. They clearly did significantly worse than any other breed or species in this test. Currently I'm unsure what makes them so poor at it.

So my first challenge will be getting some Treehuggers to pass this test. I will experiment with different numbers of Norns to see if it's their overcrowdedness that prevents them from completing the test. Then I will try crossbreeding them with a Chichi based genome and see if the offspring can't do better.

I've found this test today to be very interesting, mostly in regards to the CFE/non-CFE comparisons. I'd be interested to see some more tests done to see just how much an improvement this genome really is.

6 comments:

  1. Very interesting test! I agree with you about using C3 for experimentation: I get way too attached to the original Norns to do much more than nurture them through their entire lives! Creatures 2 is also nice for experimenting, but I think I'm just pretty bad at letting nature take control.

    The similarity in numbers between the CFE and non-CFE Norns is very surprising: I would have expected a somewhat significant difference. I'll be looking forward to your next batch of results! Good luck!

    ReplyDelete
  2. I accidently left the Bengels running while I made dinner. I came back to about 15 of them in under an hour (running in fast mode).

    ReplyDelete
  3. Anyone know if creatures get smarter... generation to generation?

    ReplyDelete
  4. Hey WaKKO151,

    It is theoretically possible for the creatures to get smarter, although not a gaurentee.

    Generally from what I've seen they don't get much smarter, just different and better at certain things.

    For example they tend to age faster, breed earlier and sometimes become immortal.

    But you'll never see a Norn writing poetry. Their genetics are too limited to get much smarter. Hopefully you can breed some better eaters though!

    ReplyDelete
  5. @Wakko151 Actually they can't get much smarter - because the standart genome don't allow brainmutations.

    In case you get a genome that allows brain mutations, you could get smarter norns - but they have to get a rought life! Then harder it is, then smarter they must be! - but the chances are much higher, they just end up reproducing faster instead of becoming smarter.

    You would have to make a really strong and harsh selection, and you should have a high population number - around 100 norns - but only 10 should be allowed to reproduce - only the smartest ones! After you done that you need nothing but luck - and many generations f norns....

    ReplyDelete