Overview

Monte-Carlo simulation is perfectly suited for parallel processing: it needs to run the same block of code a huge number of times. It looks like Rayon would be a great choice for implementing this feature.

Implementation Notes

I think that most of the library utilities should be good to go for threading, we just need find an efficient way to aggregate the scores. My initial thought would be to have one AggregateScore object per thread, and then periodically merge those back into a shared (Arc) AggregateScore that synchronizes the data across threads.