Friday, October 04, 2013

Did Mattingly Blow The Bottom of the 7th Inning?


It's not easy to manage a major league baseball team.  Lots of eyeballs looking over your shoulders, lots of second guessers and definitely a lot at stake.  But there was some controversy or at the very least some second guessing as to how Dodgers manager Don Mattingly handled the bottom of the 7th inning pitching matchups.

The Dodgers were trailing 2-1 in the bottom of the 7th inning.  Chris Withrow was pitching with two outs and runners on second and third.  He had just struck out Elliot Johnson for the innings second out.  The Braves sent up rookie outfielder (left handed bat) Jose Constanza who had really made the Braves post-season roster as nothing more than a pinch runner and an emergency pinch hitter.

Dodgers manager Don Mattingly decided to pull Withrow and go with the platoon matchup and bring in LHP Paco Rodriguez.  Of course Mattingly failed to think ahead to the Braves next move, or at least think clearly ahead and the Braves manager sent up the right handed hitting Reed Johnson to bat for Constanza.  On deck was the leadoff hitter Jayson Heyward.  Yes, he is left handed... but he also is pretty good.  At this point do you go against the platoon matchup but with the pinch hitter penalty (yes players hit significantly worse as pinch hitters than they do as starters) and pitch to Reed Johnson with first base open or do you walk Johnson and go lefty vs lefty against Jayson Heyward.  Mattingly chose to give the intentional walk to Johnson and face Jayson Heyward with Paco Rodriguez with the bases loaded.

So here are the three scenarios we need to look at in terms of which scenario has the highest win probability for the Dodgers (or lowest for the Braves).

1. Chris Withrow pitches to Jose Constanza with two outs and runners at second and third.
2. Paco Rodriguez pitches to Reed Johnson with two outs and runners at second and third.
3. Paco Rodriguez intentionally walks Reed Johnson and pitches to Jayson Heyward with two outs and the bases loaded.

                      Mattingly's Brain Table (100K Games Simulated)
FavoredATL %PitcherBasesOutsScenario
ATL84.607%C.Withrow2nd, 3rd2Pitch to Costanza
ATL84.277%P.Rodriguez2nd, 3rd2Pitch to Reed Johnson
ATL85.842%P.RodriguezLoaded2Walk R.Johnson, pitch to Heyward

Well, it appears that the best decision was to have Paco Rodriguez pitch to Reed Johnson, followed by having Chris Withrow pitch to Jose Constanza and lastly to give the intentional walk to Reed Johnson and have Paco Rodriguez pitch to Jayson Heyward with the bases loaded.  But did it cost the Dodgers enough win probability to really have any heartburn over it?  Only if you think that 1.24% win probability is enough to complain over.

8 comments:

The Dude Abides said...

Howell to McCann, xei.

MP said...

I lost all (remaining) respect for Mattingly after watching him burn Kershaw for 125 pitches in a blowout and then 25 more for Jansen in the same blowout. The very last thing you want to do in a blowout is give the opponent some free batting practice against your young (rarely faced) closer.

Not to mention he looks utterly confused and in coin-flipping mode every time they show him in the dugout.

MP said...

And, yes, while I don't have a sim to back it up, I'd be shocked if walking the bases loaded was ever a good move. I had the over and was very happy to see it happen.

Xeifrank said...

Yes, walking the bases loaded with two outs is not a smart move. There might be the occasional occurrence where it is a coin flip strategy but rare.

MP said...

Totally off-topic, but do you have an estimate for how much batting last is worth in terms of win pct.? That is, how much of std. HFA is due having last ups? I've never seen anyone quantify this, and figured you might the person to ask..

Xeifrank said...

HFA is worth around 4.5% at the start of a game. There are some situations that it is a little more or less.

Example:
65% favorite at home, but 56% favorite on the road.

Is that what you meant?

MP said...

Can you tell how much of the 4.5% is due to batting last, vs. other variables? In other sports, HFA is going to be due to travel, environment, officiating, etc. In baseball there is the added variable of hitting last. Just wanted to see if you can tell what hitting last is worth (separate from the rest of HFA)? I imagine in your sim, you would just put in two identical teams and make no HFA adjustment, and then see how often the teams that bats second wins.

Xeifrank said...

Good idea, I like it. I am not coded up for this at the moment, so it will probably be an off-season project.