THE BOOK--Playing The Percentages In Baseball

Thursday, August 07, 2008

Does it really matter if Juan Pierre plays every game?

I will let you decide.  Analysts (or wannabe analysts, I should probably say) love to hate Juan Pierre.  And they hate him even more when he bats leadoff.

The fact of the matter is that there are many worse ball players.  And the fact of the matter is, just as lineup order does not matter much, it doesn’t really matter whether Pierre bats first or last. (While it would be pretty bad to bat a slow, terrible hitter first, you can easily get away with batting a fast, poor hitter first - and Pierre is by no means that poor of a hitter overall - he is a little worse than an average CF’er.)

I put in Pierre and Kemp, Pierre and Ethier, and Kemp and Ethier (with Kemp playing CF), versus a RHP, into my sim to see how many runs they would score and allow.

With Pierre, the Dodgers score around .15 runs per game less than with Kemp and Ethier and no Pierre.  My sim assumes that Pierre is the better fielder (because I have him with a better projected UZR in CF), which I think he is (despite his poor arm).  (The sim also uses arm ratings, so it “knows” about Pierre’s bad arm.) The sim has the opponent scoring only around .03 runs less with Pierre in the OF.  So the net for Pierre is a loss of .12 runs per game.  If Pierre plays 40 games rather than 10, that costs the Dodgers a little more than a 1/3 of a win for the rest of the season.

As I said, I will leave it to you to decide if that is worth getting all worked up over, as well as what the chances are that a 1/3 of a win will be the difference between them making the post-season or not (I’ll give you a hint - it won’t make much of a difference).

The bottom line is that platooning Pierrre and Kemp in CF is a no brainer, if you are going to use Pierre at all.  I assume that even Torre will do that.

Joe Sheehan, in an article today about Torre’s “obsession” with using Pierre, says:

Dodgers offense by Leadoff Man
Hitter G R R/G
Furcal 31 169 5.45
Kemp 21 100 4.76
Pierre 58 195 3.36
Other 2 4 2.00

If the Dodgers fail to reach the postseason, it will be in part because Furcal got hurt. You can’t just ignore that part of the equation. But it will be just as much because Joe Torre elected to kneecap his offense by putting a bad baseball player in a critical role, and stubbornly sticking with that decision despite what it was doing to his offense. No amount of geniality, experience, speed, or hustle can counter the statistics above. When anybody but Juan Pierre leads off, the Dodgers score 50 percent more runs than they do when Pierre leads off. Consistently.

Those last two sentences, while they may technically be true, are a pathetic way of supporting an argument that Pierre should not be playing.  If you used the ole “runs scored or win/loss record with and without a player” in 25 or 50 games, I assume you could make a false case for or against almost any player if you choose the right time frame (cherry picked it).  And even if you chose a time frame without cherry picking, I assume you could make a false case for dozens of players, just by chance alone. Heck, I have to run 100,000 games in my sim, just to “see” a difference between Pierre and Ethier or Kemp.  Sheehan wants us to believe that 30 or 50 games tells us something reliable!