11-05-2016, 09:16 PM
Well, that was fun. I finally got the additive portion of the pool working at a (relatively) decent rate. And I reworked the non-binomial distribution to only include dice greater than the target number. Wow, that ran quickly. I'm currently running two instances of additive on the "good" computer. One modified to calculate pools of two to eight dice (which I mistakenly uploaded instead of the 2 to 11) and the other set to process pools of 9 dice. I figured that since one script used only ~50% of the processing, I may as well run two.
But I attached both scripts if you want to take a look. A couple notes about them: Additive gives the number of successes. To get the probabilities, divide by the product of the pool. Additive also gives a number 1 greater than the target number, i.e. a number to match or beat, not just beat. It's left to the student as an exercise to modify the script to match the output type of the other.
Next step is to do some matrix cross-multiplication to come up with a final probability vector. And yes, mathematically, that last sentence was complete gibberish. But bear with me. Say we have a pool {2d4, d8, d10, 2d12} and the target number is 9. What's the probability of failure? Of only one success? Or of three? To start, there are two sub-pools we look at: {2d4, d8} and {d10, 2d12}, one of which has small dice (add to beat the TN) and the other's dice are large. Doing lookups on these two Tables gives us:
9 [12, 12, 10] [1.0125, 0.7875, 0.18749999999999997, 0.012499999999999997]
Something very wrong.
Let me get back to you.
But I attached both scripts if you want to take a look. A couple notes about them: Additive gives the number of successes. To get the probabilities, divide by the product of the pool. Additive also gives a number 1 greater than the target number, i.e. a number to match or beat, not just beat. It's left to the student as an exercise to modify the script to match the output type of the other.
Next step is to do some matrix cross-multiplication to come up with a final probability vector. And yes, mathematically, that last sentence was complete gibberish. But bear with me. Say we have a pool {2d4, d8, d10, 2d12} and the target number is 9. What's the probability of failure? Of only one success? Or of three? To start, there are two sub-pools we look at: {2d4, d8} and {d10, 2d12}, one of which has small dice (add to beat the TN) and the other's dice are large. Doing lookups on these two Tables gives us:
9 [12, 12, 10] [1.0125, 0.7875, 0.18749999999999997, 0.012499999999999997]
Something very wrong.
Let me get back to you.
Getting me free admission into gaming conventions for a decade