Bunting is becoming less terrible (but still terrible)

In a great article over the weekend, Derrick Goold addressed how the current, lower run scoring environment might impact when it makes sense to bunt, and when Mike Matheny chooses to bunt. My hope is just that those two things become more closely correlated.

Goold reports that Matheny asked the Cardinals Brain Trust to provide him some research on bunting this off-season and was presented with several reports and presentations. There may have even been a diorama (my own speculation).

For those of us who worship at the Church of Stop Bunting, our Holy Text is Run Expectancy. For every situation in baseball - inning, outs, baserunners, there is a certain percentage chance of scoring a run. A sacrifice bunt is supposed to be a strategic move to increase a team's likelihood of scoring, so when a manager deploys the bunt in a way the Sacred Book tells us actually decreases their chances, people like me swear at their TV set.

But it's important to remember run expectancy is not fixed throughout eras. The odds of a runner on first base with nobody out scoring was very different in 2000, when every 8th-place in the order second baseman hit 20 home runs. As Goold notes, the gaps in run expectancy on potential bunt plays are narrowing, perhaps changing the book on when a sacrifice is in order.

Maybe.

The problem with evaluating when to bunt based only on run expectancy is that it assumes that a sacrifice bunt will always be successful, and often they are not. As I wrote about last fall, lots of things can go wrong when you try to bunt. If the batter misses or fouls a ball (or two) off, they put themselves in a serious pitchers count without taking a good hack of their own. And even if the batter gets the bat on the ball, bunts can be popped up, the defense can still nail the lead runner, etc.

Last season, the Cardinals attempted 97 sacrifice bunts. Only 12 times did the result of the play increase their run expectancy. That's awful.

Another way to look at those 97 sac bunts is Win Probability Added, which also takes into account the score and inning. That's significant because moving a runner up in a close game in the later innings may be more valuable. But even by that measure, only 13 of the 97 sacrifice bunt attempts improved WPA. (Nine more produced a neutral effect.)

Keep in mind, those 97 "sacrifice bunt attempts" capture only plays which ended in a bunt or a strikeout while attempting to bunt. That doesn't even include cases where a batter may have attempted to bunt earlier in the count, then ultimately swung away. Those tend to end badly as well.

Why such a low rate of actually improving run expectancy? There are a few factors.

Some situations would not improve run expectancy even if the bunt was successful. The one that most drives me crazy - on 16 different occasions, the Cardinals tried to bunt a runner from 2nd to 3rd base - and only five of those were with a pitcher batting. A runner scores from 2nd or 3rd on almost any base hit. Yes, a runner on 3rd may score on a passed ball or a deep flyout, but the odds of scoring from 3rd are not nearly better enough to be worth giving up one of your 27 outs.

The other reason for the lack of success is just that BUNTING IS HARD. Some of these were situations where a bunt was a reasonable piece of strategy, but the batter struck out, or popped up, or bunted it right to a fielder who threw out the lead runner.

We don't know what the content of those presentations to Mike Matheny on bunting were. If it were me, it would consist of me writing STOP BUNTING on a whiteboard and then pointing at it emphatically for twenty minutes. It is true that the lower run scoring environment is making the bunt a smarter play than it was a few years ago, but just by a bit, in a few very specific situations.

But even with that change in environment, given the atrocious results of the Cardinals when sacrificing last season, the result this season should be less bunting rather than more.

Correction: An earlier version misstated the extent of the changes in run expectancy in Derrick Goold's piece.