I'm going to ignore 8 innings of baseball from last night and focus squarely on the 9th inning. I will also be lifting heavily from Tom Tango's "Win Probability Added" and "Run Expectancy" matrix to illustrate my point. For reference, please visit Fangraphs.com for the WPA from last night, and my RE24 tables are from Chanceis.com (mostly since I couldn't find my copy of "The Book" which also contains these tables).
In the 9th inning, with the Yankees down by 3 runs, the Yankees were immediately returned to life by Jorge Posada. One pitch later, Russell Martin laces a double into the left-center gap giving the Yanks their best opportunity of the night. Brett Gardner, who after seeing the last two pitches travel 435 feet and 399 respectively, opts to take strike one (in standard BG fashion) but is let off the hook in the at bat on an error by Scott Sizemore. Enter Derek Jeter.
At this point in the game, the Yankees have a Win Probability of 35.5% and have an average run expectancy with 1st and 2nd and 0 out to score 1.62 runs in the bottom of the 9th. Jeter receives the sign to bunt, lays it down successfully, and advances the game-tying run into scoring position. Upon completion of the play, the Yankees Win Probability falls to 31.5% and their run expectancy with 2nd and 3rd with 1 out stands at 1.51.
To recap. Girardi called for the bunt, it was executed successfully, and the Yankees saw their odds of winning fall? Why is this?
From the sabermetric view, Girardi made the right call if all he was concerned about was tying the game. With 1st and 2nd with no one out, the Yankees have a 16% chance of scoring 2 runs in the inning (tying the game). In the same situation, they have a 26% chance of scoring more than 2 runs in the inning.
Once Girardi calls for the bunt, the play-state changes to 2nd and 3rd with one out. In this play-state, the Yankees have a 22% chance of scoring 2 runs and tying the game, an improvement in our chances to prolong the game. However, the odds of scoring more than 2 runs and winning the game has fallen from 26% to 18%. The +6% chance to tie is not offset by the -8% chance to win.
From an ordinary baseball perspective (ignoring the math) this should make sense. Sacrifice bunts are best used as a means of scoring necessary runs without the relying on getting a hit, ie bunting the tying run to 3rd with less than 2 outs so you can tie the game on a sacrifice fly.
However, bunting the tying run to second with only one out means you still need to get a hit to win the game. Not only that, a weak single won't even score that runner from second with less than 2 outs, as a runner is not going to tempt getting thrown out at the plate with only one out. By sacrificing the tying run to second base, all the offense has managed to do is give up an out in exchange for a situation that still requires them to get a hit.
Last night, Joe Girardi made hitter who was 3/3 with a walk lay down a bunt to advance the game tying run to second base. This move was only justified by either saying he was worried about Jeter hitting into the double play (which he would never admit out loud), or that all he wanted to do was tie the game to try and win it in extra innings. Next time Joe, don't give up the free out. After all, with only one out, that near grand-slam from Swisher becomes a sacrifice fly to tie the game.
No comments:
Post a Comment