MS notes

Pop merge/splits

From KRT regarding this tweet: "But the issue is very subtle, and he is one of folks who made the error. ms does not change rates of coalescence for merges." Here, most folks assume that if we have a simple split of a population of size N into two, the rate of coalescent halves (as they assume the two split pops are each N/2). However, this isn't the case — ms models the new pop sizes as N. Apparently there are lots of issues in the literature caused by this mixup.