chance of psionics

[Ichthyic]

I've tested any number of percent values for modifying the chance to test positive for psionics, but nothing seems to affect it.

 

here are the relevant (AFAICT) lines in the modfile:

 

PSI_GIFT_CHANCE=3

PSI_TEST_LIMIT=5

PSI_TRAINING_HOURS=240

PSI_NUM_TRAINING_SLOTS=3

 

I can change training hours and even the number of slots (you can just see it if you add a 4th slot to the test facility), but no matter what values I put for "gift chance", even 100, I still have only found 2 out of around 50 units to test positive. makes me wonder if this value is some kind of pre-set seed created at the beginning of a game?

 

anyone else have success modifying this?

 

edit:

 

someone already figured this out:

 

"; This number is used in a calculation to determine the % chance that a soldier will have the Psi gift. LOWER numbers INCREASE the chance, HIGHER numbers decrease the chance. If you ; want more soldiers to be Psionic make this a 0."

 

here I thought it was a straight percentage. doh. I'll post again if LOWERING this number doesn't work.

Edited October 17, 2012 by Sir_Toejam

[Beknatok]

Be warned, a setting of 0 may not work, depending on how the Unreal Engine handles divide by zero errors. If nothing else, it is safer to use 1 or turn on the hasPsiGift property in the base character/soldier template.

[Ichthyic]

I experimented... a lot...

 

1 basically *almost* guarantees a positive test, 5 *almost* guarantees a negative test. I think those are 1% checks? I have had 1 positive show up out of around 130 tested at 5, and missed 1 positive out of 80 at 1. I'd have to test further to be sure, but it looks like fractions work as well, so that 1.5 tests less positive than 1, and 1.75 tests noticeably less positive than 1.5 even.

 

logarithmic scale maybe?

 

I settled on 2 to be a reasonable number so that I wouldn't have to hire 200 soldiers to find a few positives. gives me about a 10% positive rate. the base of 4 appears to be around a 5% chance.

 

maybe the limit number affects the scale? *shrug*

Edited October 17, 2012 by Sir_Toejam

[Beknatok]

I experimented... a lot...

 

I *think* the way it works is that it takes your will, divides it by the "chance" score (4 is base), and then looks at the resulting will to determine the chance % the unit will get psi.

 

That is correct. The proof is in the algorithm, which can be found in another upk.

 

My point is I am uncertain how the unreal engine handles divide by 0. If it handles it gracefully, then good for you. But if not, then it could give you unusual results.

 

So you said you experimented... did 0 actually work for you or not?

[Ichthyic]

zero doesn't give an error, but appears to turn off any chance of getting psi at all (probably just turns off psi checking period).

[savoy6]

will over 90 seems to have the best shot. but even then not a guarantee. im getting 2 outta 3 if they have a will over 90.

[Ichthyic]

will over 90 seems to have the best shot. but even then not a guarantee. im getting 2 outta 3 if they have a will over 90.

 

yeah, the game rolls percentages differently that one might expect.

 

the formula is correct though, if you want to make sure every soldier get's psi, change the gift value so that it multiplies by ten, instead of divides by 4, by changing it to 0.1