Posted by John Dewan on December 16, 2015
November 24, 2015
It's time again for the Flat Bat Award, which we give annually to the past season's best bunter. The reigning winner of the Flat Bat Award, Leonys Martin, actually won the Award in both 2013 and 2014. However, poor overall offensive production led the Rangers to demote Martin, and with only 310 major league plate appearances, he was limited to just six bunt hits (on 11 attempts) and three sacrifice bunts.
That means that 2015 will feature a new Flat Bat Award winner. And a new method for determining the winner—more on that below. Here are the top 10 players in bunt run value:
|Name||Sac Bunts||Failed Sac Bunts||Bunt Hits||Failed Bunt Hits||Run Value|
Incredibly, the two top bunters, Delino DeShields and Dee Gordon, finished in a tie! DeShields had one more successful sacrifice in two more attempts while Gordon had 4 more bunt hits in 13 more attempts. Those two differences balance each other exactly, and as such, we’re naming DeShields and Gordon co-winners of the 2015 Flat Bat Award.
This year, we changed up our approach for determining the champion. Previously, we looked at both sacrifice and bunt-for-hit attempts, and without using any specific criteria, we weighed every player’s number of attempts at each against his success rates at each to determine a winner. This year, we relied on run value (changes in run expectancy) to make the call, instead.
On bunt hit attempts, the calculation is pretty simple. Over the last three seasons, a successful bunt hit increased a team’s run expectancy by 0.437 runs and a failed bunt-for-hit attempt cost teams an average of 0.201 runs. For the bunt hit portion of the Flat Bat Award, we are just going to apply those run values. For example, Elvis Andrus had nine bunt hits; multiply nine by 0.437 and you get 3.94 runs. Meanwhile, he had 5 failed bunt hits, which when multiplied by -0.201 is about -1.01 runs. Added together, that is 2.93 runs.
The sacrifice attempt piece is more complicated because, on average, even successful sacrifices lower a team’s run expectancy (by about 0.167 runs). It doesn’t make sense to penalize players for successfully sacrificing themselves because that is what the manager asked them to do. But because the mangers are the ones that are making the decision, we can measure a player’s effectiveness with sacrifice attempts with a plus/minus approach.
Here is how that works. Since 2013, players have succeeded on their sacrifice attempts about 68 percent of the time. Meanwhile, the difference in run expectancy between a successful sacrifice attempt (-0.167 runs) and an unsuccessful attempt (-0.340 runs) is about 0.173 runs. Therefore, to compare each player’s sacrifice bunting prowess to an average bunter, we will add (1 – 0.68) * 0.173 = 0.055 runs every time he successfully sacrifices himself and subtract 0.68 * 0.173 = 0.118 runs every time he fails. With that approach, players who succeed more than the average 68 percent of the time will create positive run expectancy and players who succeed less will create negative run expectancy. Returning to the Andrus example, he successfully sacrificed himself eight times (which multiplied by 0.055 is 0.44 runs) and failed one time (which times -0.118 is -0.12 runs). Putting those two together, 0.44 – 0.12 is 0.32 runs from sacrifice bunting, which added to his 2.93 runs from bunt hit attempts is 3.25 total runs from all of his bunting.