Not really about the disparity. When you split data into two, there's likely going to be a disparity, sometimes it will have meaning, sometimes it's just random variance (in this case, I think it's mostly about Severino going cold in a period of time where Sanchez happened to be injured). If it were the other way around, people motivated to believe Sanchez is the problem would absolutely point to it. Just thought I'd try to illustrate how braindead (yet predictable) it is to try to blame Sanchez for Severino's struggles when he was activated in the middle of them.