Bunt to Win, August ’13 Edition #smrtbaseball #bunttowin

In July, there were a total of 153 non-pitcher, non-squeeze sacrifice bunts (~35/wk). This is the highest total of the year to date, and it makes us sad.

– 8 had a WPA/RE24 > 0 (6 were due to fielding errors, 2 were FC), 145 had a WPA/RE24 <= 0
That means that ~95% of those sacrifice bunts HURT the team, or did nothing to help the team win while giving up an out.
– 22 (14%) were in the first 3 innings of the game (37 in inn. 4-6). That's low, relatively speaking.
– 54 (35%) were by the batter in the 1st or 2nd spot in the batting order.
– 88 (58%) were 1– –/0 ; 5 were 1– –/1 (*)
– 28 (18%) came with the bunting team behind (7) or ahead (21) by 2 or more runs. Ahead by 2 or more runs. 21 times. STOP BUNTING.
– 4 (~3%) came via pinch-hitter. 1 was by a pitcher in the 13th inning, which I guess we can look past. But 3 times a manager said "You, get a bat, go give an out away." Amazing.

(*) – The 5 1- -/1 bunts were more likely bunt for hit attempts, but we and the Play Index aren’t perfect.


– If your defense of a sac bunt includes the point “But the defense might make an error!”, you’re right. If a 3.9% chance is a big part of your defense of the decision, you’re probably in the wrong.
– The Pirates led the month with 11, the Tigers followed with 9. 
– 3 teams had just 2 sac bunt attempt all month (Tampa Bay, the Phillies, and the Rockies).
– #smrtbaseball #bunttowin favorites Bruce Bochy and Dusty Baker called for 4 and 6 saccys, "respectfully".

SMRT-est bunt of the week: 8/9, Minnesota at White Sox. Alejandro de Aza, 2-3 (B10) | 1– –/0. -.22 RE24. There were 3 other sac bunts of -.22 RE24, but this wins out for being in the highest leverage situation. 

The curious case of Dusty Baker & the sacrifice bunt (4/1/13)

In Cincinnati’s opening day game vs. the Angels, tied at 1-1 in the bottom of the 8th inning. Xavier Paul reaches on an error, and Shin-Soo Choo follows with a single to give us 12-/0. Brandon Phillips now at the plate with Joey Votto and Chris Heisey to follow. Dusty Baker calls for Phillips to bunt.

Look, if there’s anything we hate about baseball, it’s sacrifice bunts. And sliding head first into first base. That said, almost every rule has an exception (we say “almost” because only the Sith speak in absolutes). For the head first slide, it’s to avoid an incoming throw that’s high and to the plate side, making the swing/slap tag much harder to apply. For the sacrifice bunt, it’s:

8th/9th inning, in a tied or 1 run game (for the team behind), 12-/0.

In this situation, 1 run is critical. It either ties the game if the team is behind or puts a team ahead with <= 6 outs remaining. Now, there is a tradeoff involved, in that the bunt generally means giving up some of the probability of scoring multiple runs in the inning but INCREASING the probability of scoring 1 run. From the first group of columns of the first two tables here: http://www.tangotiger.net/re24.html 12-/0 results in 1.556 runs, on average. -23/1 results in 1.447 (giving up ~.11 runs). However... 12-/0 will score a run 64.3% of the time. -23/1 will result in a run scored 69.8% of the time (gaining ~5.5%). The primary argument against the bunt is that the Angels will automatically walk Votto to load the bases, taking the bat out of his hands. That gives us 123/1, which has a 67.9% chance of scoring a run. While down from the 69.8% chance with -23/1, it's still higher than the 64.3% we had at 12-/0. To go along with that, both B-R (http://www.baseball-reference.com/boxes/CIN/CIN201304010.shtml#wpa) and FanGraphs (http://www.fangraphs.com/plays.aspx?date=2013-04-01&team=Reds&dh=0) indicate that the bunt added a little less than a 2% increase in win probability. Some additional info to note, which may or may not be significant enough to influence the decision: - The Angels still might have walked Votto if Phillip's at bat resulted in 12-/1 even without trying to sacrifice (*). - Phillips easily has the highest GIDP% (~3%) of the 3. - Heisey actually has a much higher career FB% (44.5%) than Votto (34.4%). Not knowing how to classify FBs that would let a runner tag and score is where I'm not sure this can be considered. All told, this sac bunt is easily defensible. Don't fret too much, though, Baker made a much worse decision in the 13th. Walking Kendrick (-2-/1) to get to a PH (in this case Hank Conger) with 12-/1 made it more likely that the Angels would score AND score more runs in the inning. There's your smrtbaseball.