Conversation

btw, razor #1 would have you choose over A over B as long as the number of assumptions is similar/same
1
"enough" as in there's a really large cluster of observations explained by A, and little enough that needs B
1