Sabremetrics and the fallacy of one metric

This post is more about giving Yunel the green light on a 3-0 count. But it highlights the problem of putting too much stock in one metric.

The defense for giving Yunel the green light on 3-0 is that the Nats are aggressive and they are 7-14 when they put the ball in play on a 3-0 count. But Yunel was at bat, not the team!

To me this is a classic example of misusing/misinterpreting historical performance data. The fact that the team was 7-14 when putting the ball in play on a 3-0 count is irrelevant.

According to the gameday data that I have downloaded, there have been 38 PAs this year by the Nats where the count was 3-0 and the PA ended on the fourth pitch. Of those, 24 were walks, leaving the 14 PAs in the 7-14 that MW referenced. Here is the detail for those 14 PAs:

  • Jayson 2-3
  • Desmond 1-4 with two walks
  • Lobaton 1-1 with two walks
  • Danny 1-1 with a walk
  • Ramos 2-2

So that’s 11 of the ABs. Here are the other three:

  • Uggla is 0-1
  • Bryce is 0-2 with 6 walks

Bottom line is that there is nothing in the data to suggest letting Yunel swing is a good idea. In the case where the PA ends on the 4th pitch, over 60% of the time you walk on that 4th pitch. I did not bother to look at what happens if the PA goes beyond the 4th pitch.

And yes, I excluded intentional walks in this analysis.

Combined with the fact that the Padres reliever was having problems throwing strikes makes this a really bad decision IMO. And that is exacerbated by who was on-deck.

Jump to Comments Box

This entry was posted in Sabermetrics. Bookmark the permalink.