Ah. Nonetheless given this is analysis over historical data I would label any PR that hasn't been merged in over the last year as effectively spam. The fact that none of their stats seem to include how many PRs were actually merged in is rather concerning.