With the final pitch of this past world series in the book, we now have two iconic series-ending takes in recent memory. Miguel Cabrera was frozen by a Sergio Romo fastball when he was perhaps thinking slider, and Carlos Beltran famously flinched at an Adam Wainwright curveball in game seven of the 2006 NLCS.
Of course that’s just two data points, connected only tenuously by situation (last out) and outcome (strike three taken), but it is enough to spawn a digression. Even if it would be kind of crazy to find out that batters swing less often as your average game progresses based on this starting point, crazier things have been born of less consequential moments.
With the help of sabermagician Jeff Zimmerman, I set out to answer this question. Seems easy enough. How often do batters swing, by inning? Let’s limit it to at-bats against starters, considering the difference between a starter making his way through the lineup again in the sixth inning and a fresh reliever in the same inning.
[Updated Thursday Night]
Um. So we’re done here? The “n” in even the smallest bucket is over 1800. And there go the swing percentages, increasing steadily as the game goes on. Easy enough to see. Case closed?
Except that it’s entirely possible that the pitchers are throwing more strikes as the game goes on. Let’s check how often the starters chuck it inside the zone, by inning:
Okay so batters are swinging more against starters as the game goes on, but starters are throwing the ball in the zone just a fraction less. That strange blip in the ninth inning of both tables reminds us, though, that there’s a survivor bias in here. A starter that’s still going in the ninth inning is probably having a great game. What happens when we add the relievers back in?
Everything goes out the window. If batters were swinging more as the game went on, they checked that tendency at the door when a reliever came in the game. If the starters were chucking it in the zone a little less as the game went on, the relievers came in and started afresh with a zone percentage that looked more like it belonged in the first inning.
If there is an interesting string to follow here in the future, it is the slight chance that starters who remain in the game begin playing some sort of cat-and-mouse game, where they throw the ball in the zone less and the batters swing more at the pitches that are left in the zone. Could there be a general relationship between batters faced, and zone and swing percentages? We’ll keep looking.