THE BOOK--Playing The Percentages In Baseball

This is ARod by LI levels:

LI Level: WPA, WPA/LI, Number of Plays
1.50-10.9: +1.13, +0.19, 39 plays
0.75-1.50: +0.63, +0.79, 70 plays
0.00-0.75: +0.29, +0.86, 44 plays

The WPA/LI is that “perfect neutral” metric.  That’s the one that rebalances the hits, HR, walk, out values based on the game state.  When the bases are loaded, the walk is worth alot more than the bases empty and 2 outs, etc, etc.  So, this metric gives you the correct impact of the event, without the multiplying effect of the leverage.

So, look at the last line.  In low-leverage situations, ARod performs GREAT.  +0.86 wins in the neutral measure of wins, on 44 plays.  That’s +.02 wins per play.  Give him 600 PA, and he’d be +12 wins, one of the best of all-time.  ARod loves those low-pressure situations.  Unfortunately, because it is low-pressure, he only adds +0.29 wins.

Now, the first line, the high-leverage situations.  Here, he doesn’t do as well, being +0.19 wins in neutral terms, or about +.005 wins per play (that’s +3 wins per 600 PA, or roughly a very good hitter).  But, because of the multiplying effect of leverage, he added an enormous +1.13 wins.  That’s simply fanstastic.

The pattern (not that it’s necessarily real) is that ARod performs better when the pressure goes down.  But, he managed to look incredible based on those two HR.

Completely off-topic, but a random thought I just had after reading some of Schilling’s latest entry.  Quoting him:

“One of today’s oddities is that anytime you face Steve Trachsel you have to be ready to sit for a good long while between innings. When he gets runners on the game slows down immensely. It’s a method that works for him but you know going in the game is going to run a bit longer than normal.”

I’ve never seen any quantitative proof that long waits between innings is harmful to the pitcher about to begin the next half-inning, and therefore beneficial to the team about to come to bat.  But just from watching games, it seems to be the case that starters get lit up a little more often when they just sat for 20+ minutes while their teams poured in a few runs.

This could be due to getting rusty on the bench, or even the pitcher relaxing a bit due to likely having a lead when he enters that next half inning.

In any case, if this is true, would it make sense then that slow-working pitchers, like Trachsel, get more run support from their team overall than their fellow starters on the same ballclub?

I don’t know if MLB.com or retrosheet or any of those sites have some way to parse out time between innings, but I think it would make for a fascinating study.  Imagine you have Johan pitching, and in the top of the 7th the Twins start the inning with two walks and a single.  Would it make sense to start warming up Neshek, in case it turns into a 25-30 minute inning?