balanced randomness in trigger firing?

**marcvs agrippa** · January 08, 2009, 09:55 AM

Hello all,
afaik all triggers are processed in order of appearance is that right?

Given you have two triggers affecting two different traits which exclude each other (by AntiTrait or by Condition). Both Triggers have a 50% chance to fire...

Code:

Trigger a_1
  Condition Trait B < 1
  Affects A 1 Chance 50
Trigger b_1
  Condition Trait A < 1
  Affects B 1 Chance 50

I found trait A is more Common than trait B. I guess it is because a_1 is tested first and b_1 is tested against what's left, having only a 50 percent chance of firing itself... so the chances of the second trigger are in fact decreased. (remember Monty Hall?

)
Over multiple turns, the first trigger will fire way more often than the second one.

So how do you create a balanced randomness in the firing of triggers, when the (very determined

) order of appearance is a key factor?
How about decreasing the chance of firing, doubling the number of triggers and write them down in alternating order:

Code:

Trigger a_1
  Condition Trait B < 1
  Affects A 1 Chance 25
Trigger b_1
  Condition Trait A < 1
  Affects B 1 Chance 25
Trigger b_2
  Condition Trait A < 1
  Affects B 1 Chance 25
Trigger a_2
  Condition Trait B < 1
  Affects A 1 Chance 25

Would that increase the variety of A and B?
Or would that only make B more common?

thanks for any help...

**Aradan** · January 08, 2009, 10:41 AM

Try this:

Code:

Trigger a_1
  Affects A 1 Chance 50
Trigger b_1
  Condition Trait A < 1
  Affects B 1 Chance 100

**marcvs agrippa** · January 08, 2009, 11:08 AM

Originally Posted by Aradan

Try this:

Code:

Trigger a_1
  Affects A 1 Chance 50
Trigger b_1
  Condition Trait A < 1
  Affects B 1 Chance 100

That would be an "if ... else ..." condition, dropping the third opportunity: that nothing happens.
What I am trying to achieve is an equal chance for two traits to appear or not... reset every turn.
But the latter triggers have no fair chance in this race...

**Aradan** · January 08, 2009, 11:34 AM

Then why not use your original trigger? If the game code that handles these calculations works properly, then each check is independent of the previous (a new random number is generated at every check), so both traits have the same chances.

**Squid** · January 08, 2009, 11:57 AM

Do you want each to happen an equal number of times, where one of the two must happen? If so Aradan's suggestion is the way to go. If you want three mutually exclusive outcomes, A or B or nothing, each occuring with equal probability you need to do it as follows:

trigger 1
affects A 33%

trigger 2
if A = 0
affects B 50%

**marcvs agrippa** · January 08, 2009, 01:55 PM

thank you both for your help!

Originally Posted by SquidSK

If you want three mutually exclusive outcomes, A or B or nothing, each occuring with equal probability [...]

Yes, that's what I had in mind. I'll try out your solutions and report back with the results...

thanks again.

**Squid** · January 08, 2009, 02:47 PM

I can tell you the results, it works, as I use the same thing for traits, it works on conditional probability. A happens 1/3 of the time, that I think is pretty obvious. Of the other 2/3rds of the time B happens half the time and 1/2 * 2/3 = 1/3, which leaves the remaining third for nothing. This can be extended up to however many exclusive outcomes you want, but the conditions get longer and longer. The most I've done it with is 7 traits, where one should be picked randomly but each should be picked equally.