The Sports PageWe Bet Runs Go Up After a Pitching Change. They Do — Before It.
Do hitters feast after a pitching change? Our editor bet they do, at least in the 4th, 5th and 6th innings, before the best relievers come in. In the 8th and 9th, he guessed, the bullpen wins.
We checked every play from five seasons, 2021 through 2025. In the 6th inning, an inning with a pitching change averages 1.3 runs. An inning without one averages 0.28. That is nearly five times as many.
Most people would read that as proof that changes go badly. Here is the surprise: watch what happens after the new pitcher arrives. He walks into runners and outs he did not create. That spot is worth 0.67 runs on average. He gives up 0.66. The big inning was already happening. Managers pull pitchers who are getting hit. The runs cause the change, not the other way around.
By the end of this you will be able to spot this trap anywhere: the "effect" that is really the cause.
Half the Runs Came Before He Got There
Take that 6th inning apart. Of the 1.3 runs, about 0.64 had already scored before the new pitcher threw a pitch. The rest is what any average pitcher would give up with those runners on and those outs.
That holds in every inning from the 4th through the 9th. After the change, runs land within 2% of what the spot was worth. Early in the game, the 1st through the 3rd, they run about 13% higher. Those are mostly starters being knocked out, and there are only about 1,200 of them, so do not lean on it.
How do we know what a spot is "worth"? We counted. Every time there is a runner on first with one out, say, we tallied how many more runs scored that inning. Do that for all 24 combinations of runners and outs and you have the price of every situation. It is the right yardstick, because it asks what this inning was going to produce anyway.
The new pitcher did not cause the big inning. The big inning caused the new pitcher.
The Fair Test: A Fresh Inning
The cleanest comparison is a change between innings. Nobody is on base, and nothing forced the move. In the 4th, 5th and 6th, a new pitcher allows 0.49 to 0.52 runs. The pitcher who stays allows 0.51. No difference at all. So if you groan when your manager walks to the mound in the 6th, you can relax; the new man is no worse than the old one. And hitters do not hit better against him: in the 6th, about .24 against the new arm, .25 against the old one.
Late in the game the editor is half right, and the other half goes the wrong way for hitters. In the 7th and the 9th, a fresh pitcher allows about one run in ten fewer. In the 9th, hitters bat about .22 against him and .25 against a pitcher who stays in. Hold that loosely, though. The same late edge did not show up in 2015 to 2019.
What to take home
More firefighters show up at bigger fires. Nobody thinks the firefighters made the fire bigger. A pitching change is a firefighter. So when you see two things show up together, ask which one came first. Then compare the result to what the situation was already worth, not to an ordinary day. A hospital looks dangerous next to a gym for the same reason: sick people go there.
Notice how close this came to being a headline the other way. "Innings with a pitching change score five times as many runs" is true, and it is the kind of graphic you will see on a broadcast. It just answers a question nobody asked.
How this was measured, and what it cannot do
Every regular-season play from 2021 to 2025, from Retrosheet. A mid-inning change compares the runs scored from the change to the end of the inning with the expected runs for the runners and outs at that moment, built from the same five seasons. A start-of-inning change compares the new pitcher's inning with innings where the same pitcher came back out. The 9th inning uses top halves only, because walk-offs cut the bottom halves short. Position players pitching are left out. As a check, our parsed run totals matched MLB's official team totals in all eight seasons we tested.
The yardstick is an average pitcher, not the one who was pulled. Nobody can see what would have happened if the manager had left him in. And late in a game, "the same pitcher continuing" is often a starter cruising or a reliever mopping up a blowout, which makes that group a little odd. Adjusting for the batting order and the score barely moves the 9th-inning result.
The information used here was obtained free of charge from and is copyrighted by Retrosheet. Interested parties may contact Retrosheet at www.retrosheet.org.
Argued about something you could not settle? Send it to us · what makes a good question
