The Sacrifice
Time and time again, if you are a baseball, fan you will see the following: a team comes up to bat in the late innings of a ball game. The lead-off batter gets on base with a walk or single. The next batter comes up and the manager instructs him to “sacrifice”: bunt the ball to the right side of the infield for a sure out in order to advance the runner to second base. If he succeeds, the colour commentator will rave about him “getting the job done” and “advancing the runner”. After the next two batters strike out and fly out, no one ever says, “Gee, I guess giving up that out on the sacrifice was pretty stupid, eh?”
Does the sacrifice bunt make sense? It must work. Almost every manager in the league does it, often two or three times a game. If everyone does it, it must be right.
The sacrifice bunt emerged as a strategy at a time in baseball history before there was such a thing as a designated hitter. Late in a close game, if a runner got on and the pitcher was the next batter, it made sense, because:
a) pitchers didn’t hit very well (a .200 average was considered good), and,
b) pitchers didn’t run very fast, making them very susceptible to the double play, and,
c) it was usually a good time to bring in a relief pitcher anyway. But nowadays, American League teams use it just as often.
It is possible, with a bit of computer programming and lots of free time, to create a “simulation” of thousands of baseball games. I set up such a simulation once to test the theory that the sacrifice bunt is a stupid strategy. I ran thousands of games in which, after the sixth inning, every time the lead-off batter got on, the manager used the sacrifice bunt. Then I ran the same series of games with no sacrifice bunt. Since there is no way of knowing which hitter exactly is coming up to bat in these situations, I created an average team with a set of batting and on-base percentages that reflected the abilities of a normal range of players.
It didn’t surprise me that the second simulation showed many, many more runs scored than the simulations using the sacrifice bunt. Consider this: the following batter, in most situations, will have an on base percentage (hits + walks) in the neighborhood of least .325. So roughly 1/3 of the time, he will advance the runner anyway, without giving up an out. The next batter has the same 1/3 chance of advancing the runner without giving up an out. And so does the next. And… here is the key point… so does the next batter. Without a sacrifice, you still have three outs to work with. With the sacrifice, you only have two. How significant is that? Consider some other factors. The next batter will, of course, often hit a double, a triple, or a home run instead of a single. With the runner on first, the first baseman has to hold the runner on, leaving a gap in the infield. The pitcher is often distracted by the runner. A fast runner has a chance of stealing the base anyway– I saw Tim Johnson use a sacrifice when he had Alex Gonzalez– a good base-stealer–on first. Finally, with the runner on second and one out, if the next hitter is “hot”, he will get walked anyway, setting up a potential double play. And don’t forget that without a runner at first, the first base man doesn’t have to cover the runner.
I’m not saying the sacrifice never works. But a lot of people make the foolish assumption that the odds of getting the hit you need to score a run are roughly the same after a sacrifice as they are before. In fact, they are substantially less, because one less batter is going to get a chance to drive in that run, and because the sacrifice ensures that the “batter” following a lead-off single or double invariably “hits” a single. In other words, over a season, or even a short series, the sacrifice will fail to achieve it’s desired objective– scoring a run– far more often than simply letting the next three batters do their job.
Well, if a sacrifice is so stupid, why do managers do it? The answer is simpler than you might imagine. Consider the World Series Champions of 1992 and 1993, the Toronto Blue Jays, who were managed by Cito Gaston. I don’t think anybody in this world would think that Cito was a smarter manager than Bobby Cox, one of the shrewdest skippers in the league. Why did he win? He simply put good talent on the field and let them play the way they were capable of. He put Roberto Alomar at second base and watched him make unbelievable fielding plays. He penciled in Devon White in centre field and watched him swallow up every fly ball hit there. He had the finest defensive third baseman in the league that year in Kelly Gruber (’92), and he had terrific pitchers, including Henke, Ward, and Wells, who didn’t give up a single run in relief until the final game. He won in spite of his questionable management. He won in 1993 in spite of the idiotic managerial decision to let Jimmy Key leave as a free agent and so they could retain Jack Morris, who contributed nothing to the 1993 victory. He won because Paul Molitor, John Olerud, Roberto Alomar, and Devon White, had terrific years.
Well, what exactly, then, is the role of the manager? The role of the manager is to call for the sacrifice bunt. What if the sacrifice bunt is a stupid strategy? Then what would the manager do? Send for the closer in the 9th inning of close games?
It’s like the famous question asked of Keith Richards, of the Rolling Stones: “Why don’t you sing more often?”
Keith replied, “then what would Mick do?”