THE BOOK--Playing The Percentages In Baseball

Wednesday, December 21, 2011

Hit and Run

Mike does a great job at identifying situations that are disproportionately hit and run situations:

Finally, we arrive at the definition used for a hit-and-run situation in this study: (1) runner on first, bases not loaded, (2) none or one out, (3) a ball-strike count of 0-0, 1-0, 2-0, 1-1, or 2-1, and (4) the team leading or trailing by four runs or less. If the runner went on the pitch and the batter swung in such a situation, I will consider it a likely hit-and-run play.

And gives us these results, among many others:

Teams that attempted the hit-and-run play scored 0.11 runs on the play and 0.69 runs in the remainder of the inning on average, compared to 0.17 runs scored on the play and 0.70 runs in the remainder of the inning by teams that did not attempt the hit-and-run play in hit-and-run situations.

There are a few biases here that need to be controlled, before we can compare the .69 to the .70 (i.e., a wash).  Later in the article, Mike does an excellent job of reviewing most of them, and adjusting for them.  What he does there is exactly the kind of thing a saberist should be doing: identifying reasons for bias, and adjusting for it, as best you can.

He ends up with this huge finding:

Thus, the advantage for attempting a hit-and-run play during 2003-2011 appears to be about .061 runs on average.

That is a much larger number than I expected when I embarked on this research. I have attempted to remove as much of the selection bias as I could reasonably identify. It is possible that I have overlooked some bias or used a mistaken assumption, but every direction from which I came at the analysis pointed to the hit-and-run being a positive offensive play in most circumstances in which it was attempted.

I agree that that number seems simply too high.  Adding .06 runs on one PA is the equivalent of one of the best hitters in baseball.  Basically, it’s too big to be taken as the final number.

One bias that he noted early on, but that it doesn’t look like he adjusted for, is that a hit and run is in neutral or moderate hitter’s counts.  So, we expect more runs on that basis.  This will probably account for .03 or .04 runs of bias.  But, Mike says:

The ball-strike count also plays a role. The more favorable the count is to the hitter, the less likely that the batter will be forced to swing at a pitch he does not like. On the other hand, the same is true if the batter is not protecting the runner, and in that case, he may be more selective and take more powerful swings.

But his chart ends up with actually a net benefit for the hit and run.  Which is confusing me.

In any case, it’s a great piece of research, and hopefully will inspire more people to take up the cause.  You really couldn’t ask for any more for an initial piece of work.  It was a real pleasure to read.