On Sacrifice Bunts

by Ballard Erik on Apr 16, 2010 1:00 PM PDT

Hey everyone! Devoted reader but first time fan poster. Not enough of us non/seldom commenting readers say it, but your community has changed the way I thought about the game I love to play and coach. Thank you.

After a discussion on bunting round the old office cooler (by office cooler I mean rusty sink), I began to wonder about bunting as it relates to win expectancy. I have played a lot of baseball, and coached at the high school level and was never as bunt happy as many of my counterparts. I always felt that in may cases, the sacrifice of an out is not worth the expected reward that accompanies the slight increase in the expectancy that the run crosses the plate. In certain situations (depending on pitcher, hitter, inning) I would put on the bunt sign. As a player, I always hated it. When I watch the mariners, I hate Rob Johnson. Sorry Rob, sure you are a nice guy.

So here is where I ask for assistance. I scoured the internet for information on bunting (also marking the first time I used bing for anything - so if you googled something does that mean you binged something?) I realized that I had a community of intelligent people that knew more about the subject than I would ever know at my fingertips. I know there are so many variables to the answer to my questions, but I will post them anyway:

1) Does win expectancy go up with the sacrifice of an out? If so, when (one runners on, two runners on, bases full, one out, two out)

2) At what point is a sacrifice better than a hitters wOBA? In other words, when is the known commodity of a sacrifice more valuable then the expected result of hitter based on statistics?

3) Does the value of a bunt change given the relative time it happens in a game (early innings vs. late innings - I assume 90 ft is more important in the later innings).

Any assistance to these answers are appreciated. Again, thank you all for the reception I have earned as a casual poster and in advance for your comments on my fanpost.

"The only person who should not be sacrifice bunting is Albert Pujols" - some "small ball " guy I work with. By small ball, I mean the baseball theory - I don't really know about his actual balls.

Tweet Comment 17 comments | 0 recs |

Do you like this story?

Comments

This is a complicated question with no easy answer

but to help, I’ll provide two links:

http://www.tangotiger.net/RE9902score.html

http://assets.sbnation.com/assets/395/WinExp.xls

You’ll want to drop a bunt every now and then for game theory reasons, but there are very, very few occasions during which a bunt is the right thing to do, and they generally come when you’re playing for one run and have a bad hitter at the plate.

by Jeff Sullivan on Apr 16, 2010 1:18 PM PDT reply actions

Tie game, leadoff double, bottom ninth

Pretty sure that the bunt is always the right play there.

by Graham MacAree on Apr 16, 2010 1:27 PM PDT up reply actions

Hopefully Eric Byrnes isn't the next batter up after the bunter.

"LARGENT! UNLOADED! and he got the ball!

by seattle_since_81 on Apr 16, 2010 1:52 PM PDT up reply actions

First and second with no out and Jack Wilson at the plate would seem an almost guarantee as well.

by abender20 on Apr 16, 2010 2:53 PM PDT up reply actions

Especially against Chad Gaudin.

by Matt Erickson on Apr 16, 2010 3:55 PM PDT up reply actions

I'm pretty sure that's not true and even pitchers should be swinging in this situation.

@ERiC_UW

by lailaihei on Apr 17, 2010 10:22 PM PDT up reply actions

Pretty positive pitchers should not be swinging in that situation

Not sure about Jack WIlson

by Dewey N on Apr 18, 2010 6:14 PM PDT up reply actions

I know it is a complicated question

The chart helps. I get the obvious ones like what Graham posted below. I saw something where Tango said the thing about bunting now and again to keep defenses honest. I appreciate the help!

by Ballard Erik on Apr 16, 2010 2:46 PM PDT up reply actions

I should say charts.

These rock!

by Ballard Erik on Apr 16, 2010 2:49 PM PDT up reply actions

Regarding those RE tables...

…I was thinking about this yesterday because those are the tables I always use to look up RE, but 1999 – 02 was right in the heart of the steroids era. I’m wondering if the numbers in the table haven’t changed at this point as the nature of offense has changed.

In other words, I guess I’m asking if recently (over the last three or so years) the run environment isn’t significantly lower than it was from 99 – 02.

by Andersean on Apr 16, 2010 4:02 PM PDT up reply actions 1 recs

Its tricky to figure out because if you just use WE calculations they will almost always tell you not to sacrifice

and its tricky to adjust for your specific lineup because you have to sit down for every situation and go through the math. I think in most situations the sacrifice is overused but for brutal hitters (Jack Wilson or Rob Johnson) it probably makes sense if the WE says its close to a neutral outcome.

You also should consider the run environment. Sacrifices make a lot more sense as the run environment decreases because one run becomes much more valuable. That means in a M’s vs A’s game with Felix and Brett Anderson pitching in Safeco field a sacrifice probably makes more sense. Same thinking goes for different levels of baseball so you have to consider the average run environment.

by Edgar for Pres on Apr 16, 2010 3:34 PM PDT reply actions

Noted.

Thanks for the input.

by Ballard Erik on Apr 16, 2010 3:38 PM PDT up reply actions

A formula

One could create a formula to answer the question “what percentage of the time should we bunt?” given certain quantifiable properties of the game situation. Say a formula that takes as input the inning, the score difference, the general skill of the pitcher, and that of the hitter. Something like a 0 to 5 integer scale for hitter and pitcher skills would suffice.

You could have something like
(Inning + Pitcher quality – hitter quality – difference in score)/14. The closer the number is to one the more favorable the situation is to bunting, the closer to zero the less favorable. If the numerator is negative don’t bother calculating it. The situation is not favorable for bunting. I am sure this could be worked out with some coefficients to yield a reasonable indicator of how frequently to bunt.

I don’t propose this as something that would be terribly useful on the field. It is a game-theoretic toy.

"The holy grail is to spend less time making the picture than it takes people to look at it." -Banksy

by two_hands on Apr 16, 2010 5:33 PM PDT reply actions