Author Topic: A question about the age costs  (Read 5397 times)

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
A question about the age costs
« on: May 09, 2006, 03:01:34 PM »
I'm wondering how the age costs work exactly. For example: I have a bot that is 131913 cycles old. I've been running this sim without any age costs selected, but for 3000 cycles ago I decided to put in some aging costs. I set the 'age cost to' 0.25 and 'Begin upon reaching' to 20000 cycles. Will this start counting from the moment I set it and click change in the GUI? If that's the case then this bot should start ageing at 148913 cycles. Also I'm wondering what will happen if I save and quit, and then load the sim again. Will the ageing continue to count where it left off or will it start from the beginning again? Also I'm wondering what the 'Increase cost log(age)' mean exactly. I presume it is some kind of acceleration on the cost value over time. If so, how much?  
The internet is corrupt and controlled by criminally minded people.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
A question about the age costs
« Reply #1 on: May 09, 2006, 03:15:01 PM »
The start aging at X means the robot's age, not necessarily when you select the aging cost.

IE: set it to 20000 cycles and bots will begin aging when they are 20000 cycles old.

The increase cost log(age) does indeed indicate that the costs will increase as the bot ages.

More specific details I don't know.

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
A question about the age costs
« Reply #2 on: May 09, 2006, 03:20:21 PM »
I think that is a little strange, because I can't see my old bots get affected by age yet.
The internet is corrupt and controlled by criminally minded people.

Offline Griz

  • Bot Overlord
  • ****
  • Posts: 608
    • View Profile
A question about the age costs
« Reply #3 on: May 09, 2006, 07:19:46 PM »
Quote from: Testlund
I think that is a little strange, because I can't see my old bots get affected by age yet.
crank the cost up high and I bet you will.
it takes balancing the cost with a bots 'intake' of energy ...
you know, input/output ...
in the case of veggies, the nrg/cycle can't be too high, or you will never see
the age cost as it will be insignificant in comparison.
maybe Eric can give us some nominal starting numbers and explain
how they ramp up if using the log function.
« Last Edit: May 09, 2006, 07:21:12 PM by Griz »
不知
~griz~
[/color]
   "The selection of Random Numbers is too important to be left to Chance"
The Mooj  a friend to all humanity
[/color]

Offline Welwordion

  • Bot Destroyer
  • ***
  • Posts: 325
    • View Profile
A question about the age costs
« Reply #4 on: May 09, 2006, 07:40:20 PM »
Actually this can be used quite interesting in Evo sims, make all bots vegs with around 3-11 nrg per cycle
then upon reaching a certain age make the nrg cost 4-11 (+1 to what you gain) so you have -1 netto that way
your bots wont die out and have enough energy to reproduce on the other hand methusalems that do not do anything will be killed and bots who learned to hunt will survive even after losing their veg bonus (this is also quite useful to limit the number of bots and to avoid Alzheimr as all bots can convert waste into energy).

Offline EricL

  • Administrator
  • Bot God
  • *****
  • Posts: 2266
    • View Profile
A question about the age costs
« Reply #5 on: May 09, 2006, 08:23:28 PM »
Quote from: Numsgil
The start aging at X means the robot's age, not necessarily when you select the aging cost.

IE: set it to 20000 cycles and bots will begin aging when they are 20000 cycles old.

The increase cost log(age) does indeed indicate that the costs will increase as the bot ages.

More specific details I don't know.

This is correct.  For example, if you set the age cost to 5 nrg/cycle and configure it to begin at 1000 cycles, then once bots reach 1000 cycles of age, the system will begin charging them 5 nrgs/cycle from that point forward.  If you choose to increase the cost log(age) then the cost charged to the bot will be 5 * log(age) or at 1000 cycles,  15 nrg/cycle.  At 10k cycles, this will have increaed to 20 nrg/cycle.  At 100k cycles, to 25 nrgs/cycle.  Note that log(age) is the log of the age of the bot since it came into being, NOT the log of the number of cycles since the cost began being charged.
Many beers....

Offline EricL

  • Administrator
  • Bot God
  • *****
  • Posts: 2266
    • View Profile
A question about the age costs
« Reply #6 on: May 09, 2006, 08:35:54 PM »
Quote from: Welwordion
(this is also quite useful to limit the number of bots
Exactly.  If you find a sim had a population explosion and want to drop the numbers down, just set the the cost to 100 or something and step a few cycyles then set it back.  It's an easy way to kill off the weak (I.e. low energy) bots.  

One of the things I would like to implement are a bunch of built-in script-like settings which one could set to automatically adjust such things as age costs and the autotrophic energy coming into the sim so as to automatically try to keep the bot population numbers within a specified range.  As bots became better adapted, the system coudl increase the age costs (or potentially other costs) to keep their numbers under control.   Similarly, if the population dropped too much, the system could reduce costs for a while to avoid everythign going extinct.   I'm also toying with the idea of environmental variation such as occasional single-cycle burst costs which could simulate particularily harsh winters...
Many beers....

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
A question about the age costs
« Reply #7 on: May 09, 2006, 10:08:16 PM »
No wonder I don't see any effect of the age cost then. I have it too low. On the General tab I have the energy set to 12 NRG PER Kilobody point, wich means I should set the ageing cost to 13, if I understand correctly.

It's night in the sim now, so I have to wait until the sun goes up before making the change, or the sun will go up as soon as I click change in the GUI, wich whould disturb the day/night cycle rythm. I think it's a little annoying that the sun resets whenever I click change in the GUI or when I load up a sim but... We've allready been through this. I can live with it.
The internet is corrupt and controlled by criminally minded people.

Offline EricL

  • Administrator
  • Bot God
  • *****
  • Posts: 2266
    • View Profile
A question about the age costs
« Reply #8 on: May 09, 2006, 11:54:51 PM »
Quote from: Testlund
I think it's a little annoying that the sun resets whenever I click change in the GUI or when I load up a sim but... We've allready been through this. I can live with it.
No, no, no, I really appreciate you bringing these things up.  It's pretty easy to fix little things like this, but I can only fix them if I know about them.  I'll put this one on my list.  Keep finding those bugs!
Many beers....

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
A question about the age costs
« Reply #9 on: May 10, 2006, 08:30:24 AM »
Thanks, Eric. Well, it looks like I set the ageing costs too high after all. All my veggies have been exterminated. I will reload a previously saved sim with a much lower cost, maybe 1 instead of 13. Maybe it is very difficult to find out the right value to keep it balanced.
The internet is corrupt and controlled by criminally minded people.

Offline Welwordion

  • Bot Destroyer
  • ***
  • Posts: 325
    • View Profile
A question about the age costs
« Reply #10 on: May 10, 2006, 10:49:17 AM »
^^ I rather think the problem is that your nrg is per kilobody (per 1000 body) and age cost is a constant per cycle such small bots (and ones that reproduce more often) would be at disadvantage.
« Last Edit: May 10, 2006, 10:52:48 AM by Welwordion »

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
A question about the age costs
« Reply #11 on: May 10, 2006, 01:49:51 PM »
That is what it looks like in my sim right now. I have a bunch of veggies that are between 20000 and 30000 cycles old. Those gain energy. Then I have a lot of offspring that are only a few 100 cycles old. Those LOSE energy. But it still doesn't make any sense to me. The ageing function should only look at how old each bot is before starting to take energy from them. Young bots shouldn't be affected at all. I can't comprehend what the energy settings on the general tab means. I only know that 12 per kilobody point works best for my veggies.
The internet is corrupt and controlled by criminally minded people.

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
A question about the age costs
« Reply #12 on: May 10, 2006, 02:06:17 PM »
I have just found out that it has nothing to do with what I have set the ageing costs to. Even if I have nothing typed in these fields young bots still lose energy while the old don't. The reason must be something else. The ageing works as it should.
The internet is corrupt and controlled by criminally minded people.

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
A question about the age costs
« Reply #13 on: May 10, 2006, 02:24:04 PM »
Well, this was weird. I have changed it so the veggies are given 3 nrg per veggie per cycle instead. Now both young and old gains energy. I have the ageing cost set to 0.25 at 20000 cycles, so those that are older than 25000 cycles or so has started to slowly lose energy. Nice.  
The internet is corrupt and controlled by criminally minded people.