12 Comments
User's avatar
Kenny's avatar

Is there a possibility of correlation between "Starter pitches deeper into game" and "opposing offense is weaker" (and the inverse of "Starter leaves game early" and "opposing offense is stronger") that helps explain the Bullpen ERAs in each instance?

Expand full comment
Dylan's avatar

Would think there’s at least some signal there

Expand full comment
Daniel Evensen's avatar

Without jumping into the math, I’ll say this: baseball is really the only sport where the most crucial moments usually don’t feature the best players.

Chances are really good that you’ll have some relief pitcher pitching against a pitch hitter in the late innings of a close game.

Sure, there are well-known closers — but they tend to be primadonnas, and, when you realize most of them only work an inning at a time, are likely overrated.

Expand full comment
Gary Maske's avatar

How often do we see a pitcher pitch well into, say, the 6th inning, give up a bloop single and a double, get pulled, and the floodgates open on the fresh reliever? Managers are doing the "safe" thing but perhaps not the smartest thing. Maybe it's the hitters that have awakened? Has anyone tracked BA by the number of the hitter's at-bat?

Expand full comment
Will Titus's avatar

A lot of it is risk-mitigation on the manager's role. If they let a starter go into the 6th/7th and the starter gets shelled, they look bad for not pulling the start earlier. But if they pull the starter in the 5th and the reliever in the 6th does poorly, it's much easier to blame that on the reliever not doing their job. It gets even worse when there's the 2nd order effects for the following day. If the bullpen struggles the next day, people will rarely remember that they have a weakened bullpen because the manager pulled the starter early the previous day (and it's really just marginally worse so there's no guarantee that the bullpen struggled because of that).

In short, the manager has an incentive to make the bullpen look worse than for himself to look worse. And sure, the manager's "true" incentive is to win which shouldn't care, but just look at NFL GMs. They frequently make draft trades that the analytics don't support in order to save their jobs. (good video on the NFL https://www.youtube.com/watch?v=2Kdb2b0em-I)

Expand full comment
Gary Maske's avatar

How often do we see a pitcher pitch well into, say, the 6th inning, give up a bloop single and a double, get pulled, and the floodgates open on the fresh reliever? Managers are doing the "safe" thing but perhaps not the smartest thing. Maybe it's the hitters that have awakened? Has anyone tracked BA by the number of the hitter's at-bat?

Expand full comment
Gary Maske's avatar

How often do we see a pitcher pitch well into, say, the 6th inning, give up a bloop single and a double, get pulled, and the floodgates open on the fresh reliever? Managers are doing the "safe" thing but perhaps not the smartest thing.

Expand full comment
Gary Maske's avatar

How often do we see a pitcher pitch well into, say, the 6th inning, give up a bloop single and a double, get pulled, and the floodgates open on the fresh reliever? Managers are doing the "safe" thing but perhaps not the smartest thing.

Expand full comment
ogc's avatar

Great analysis of this situation!!! When I saw this idea, it sounded good on the surface, but I couldn’t think of what could go wrong, and you brought out a ton of stuff that might go wrong, great job! Thanks!

Not that it’s bad, but it would also end the use of openers as well as any tandem starts.

Expand full comment
Kyle's avatar

You worked your butt off on this! Well done.

Expand full comment
Mark Kolier's avatar

Great detail and well presented!

Expand full comment
Mason Linken's avatar

Great read!

Expand full comment