THE BOOK--Playing The Percentages In Baseball

By “situational”, aren’t we often talking about contact ability?  While inning matters, most often situational is about converting base runners in to runs.  In those situations, it would seem the value of a ball in play goes up, as a walk isn’t likely to drive in runners—even if it does up the RE for the inning.

Because Dunn struggles to put balls in play generally, and because pitchers have the option of pitching around him (and are encouraged to do so given Dunn’s power), we would expect a decrease in AVG, a corresponding decrease in SLG and an increase in OBP due to the extra walks more than offsetting the drop in AVG.

We can really see this affect when we compare the different runners on base scenarios, focusing on the runner on 1B.

Bases Empty: .250/.357/.528
Any runners on Base: .243/.407/.507

Runner(s) on, including 1B : .259/.389/.537
Runner(s) on, but not on 1B: .201/.444/.438

We can also see the impact of the number of outs.  With two outs and RISP, Dunn’s AVG drops 30 points (from his career average) and his SLG drops a proportional 73 points.  However, his OBP jumps 58 points to .438. (-.030/.058/-.073)

With just RISP, regardless of the out situation, the effect is somewhat muted (-.022/+.033/-.043), suggesting the additional leverage due to the out situation amplifies the affect.

When the value of getting a hit goes up in proportion to the value of merely getting on base, it appears Adam Dunn gets pitched around. (of course this is a differential diagnosis instead of looking directly at pitch data).  Given Dunn’s skill set, he’s likely to take the walk if given the opportunity.  He’s not likely to find success in expanding his strike zone.  I imagine we’d see similar results for most high SLG, low AVG hitters, with their OBP moving in proportion to their “natural” willingness to take a walk.

By situational, I mean by situational.  I don’t know what the CW on that is.  You have 24 base/out states, each of which requires its own pitching/batting approach.  You throw in the inning/score, and you’ve got thousands of situations.  (You can also include the speed of the runners and the quality and profile of the batters due up.)

Good numbers on 1B being open and Dunn getting his walks.  It would be better if you remove his IBB, since if a pitcher is giving him an IBB, then he’s figuring that the win impact of that IBB is the same as facing Dunn (which would be +.004 wins per PA times the leverage index for that game state).

Yeah, I realize that was just a quick slice of all of the possibilities, but hopefully somewhat instructive on what the distribution might look like across those two dimensions (base/out).  Obviously game score is very influential as well.

Removing the IBB and PA, his 1B empty OBP drops to .373.  It would definitely be interesting to see the leverage distribution for those IBB.  Presumably those are coming in the highest leverage situations, particularly given the unfortunately reality that Dunn has often batted low in the order, with significantly worse hitters behind him.

Without looking at it at all, I would guess that Dunn would be ideal as a #2 hitter.

I had always wanted Dunn to hit 2nd for the Reds. He hit 5th (sometimes 6th!) way too often.

I know lineups don’t add much to wins, but it would have been nice to see the best hitter get an extra 40-45 PAs (or so) per year.

But now he’s in AZ.

Tom, us saber-inclined Reds fans have been pushing that idea for years.  It got a little play under Bob Boone, but not much since.  Major league managers just can’t seem to fathom the idea of putting a 40 HR guy with no speed at the top of their lineup.

His OBP, fly-ball tendencies (avoiding double plays), and low average all support it.  Batting him 5th just exacerbates the disadvantage of his low AVG game.  Meanwhile, guys like Brandon Phillips or Jacque Jones end up batting higher in the order because of their batting average and speed—both of which are better leveraged lower in the order. 

It just doesn’t make any sense when you think about it.

Tango, I don’t know whether Dunn’s performance suffers because of lack of situational hitting, but I don’t see the evidence. Dunn’s Palmer wins on bb-ref 17.2, WPA/LI on fangraphs 17.92. This would be a lot easier if Fangraphs would put a simple lwts on their pages so we could see how the park factoring matched up. Of course Miguel Cabrera is going to put up higher WPA/LI if they have the same slash line - one played mostly in Cincy and the other mostly in Florida. Cabrera has 18.5 Palmer wins and 19.6 WPA/LI in 3585 PA. Bay has 12.0 Wins and 13.7 WPA/LI. Teixeira has 14.9 Wins and 12.0 WPA/LI. Dunn is right in line with that group. I think your comparison just shows that J.D. Drew, perhaps contrary to his reputation, has been a top-notch situational hitter, putting up 17.9 Wins and 24.6 WPA/LI in 4772 PA.

Wouldn’t we expect that of two players that had the same OB/SLG, the one with the higher batting average would be the most efficient hitter of the two?

I don’t know this from any work done on the question, but it is just one of those things that would seem to make sense on the surface.

Nat, this has been discussed many times here and on my main site.  The lower the BA, the more effective you are.  But, just barely.

#9 - is that because of fewer double plays?