I don't think it's particularly taxing to have a starter pitch into the seventh if he's doing well. The eighth? It's been known to happen. Does it increase the risk of injury. Sure. It's generally a marginal, but it's a real one and pitches accumulate. But eventually you have to accept that prevention that causes you to regularly lose winnable games is not worth it. And in the end, having good relievers pitch four or five days a week instead of two or three is also taxing and increases injury risk too. Maybe moreso. Less recovery time and whatnot. And lesser pitchers end up in high-leverage situations. They just do. Over and over. Last night and the night before and the night before that. Benjamin Grimm wrote: The way I look at it, if you use five pitchers in a game instead of two, you have three more guys who might potentially be having a bad day. And that one bad inning that you get from that pitcher could cost you the game. If a pitcher is doing well, and hasn't thrown too many pitches, keep him in. Ride him as long as you can. It seems a simple principle. If you lift someone who is succeeding, you continually place another priority ahead of succeeding. Eventually, and with increasingly regularity, you will fail.