Why Terry Francona’s Aggressive 3-0 Approach is a Mistake

For all the talk we’ve heard about Terry Francona over the last few months, it’s easy to forget that he has yet to actually manage a non-exhibition game in Cleveland. He’s been a very visible figure this offseason and he’s clearly made his mark on the Indians’ clubhouse already, but for all the talk of his strategic genius we have yet to actually see him in action with the Tribe.

So when Francona offered some concrete information on how he plans to manage the Indians last week, it got some attention. Specifically, he says he will encourage his players to swing in 3-0 counts:

"Cleveland’s hitters will be given the go-ahead to attack 3-0 pitches in certain situations this year.“I never quite understood why so many people just automatically take 3-0,” Francona said. “If you’re a smart enough hitter, and you’re able to just say to yourself, ‘OK, I’m going to get a certain pitch in a certain spot,’ it’s like a free swing.“A lot of times, even if a guy swings 3-0 and fouls a pitch back, because they took a healthy swing, they feel better about the next pitch. I think it’s great, and I think it breeds confidence.”"

It’s an interesting philosophy and I love Francona’s willingness to think outside the box, but that doesn’t mean it’s the right idea. In fact, swinging away in 3-0 counts is exactly the opposite of what a smart hitter should do.

For those unfamiliar with the concept, game theory is the study of decision-making given the constraints of different circumstances and other actors’ desires. The “prisoner’s dilemma,” a hypothetical situation in which two conspiring criminals must decide independently whether or not to cooperate with an investigation, is perhaps the most famous game theory analysis. Though both criminals would be better off if both refused to participate, each knows that he is better off individually by turning on his partner. Given the impossibility of coordination and each criminal’s desire to improve his own situation, the Nash equilibrium (the combination of both men’s rational decisions) is for both criminals to turn each other in.

Kevin Jairaj-USA TODAY Sports

At the Society for American Baseball Research Analytics Conference earlier this month, Matt Swartz gave a fascinating presentation about game theory as it applies to pitchers’ choices about what pitches to throw and where to throw them and batters’ decisions of whether or not to swing. Though the 3-0 count model was not the focus of his presentation (it was just a simple illustration before he got into more the more specific questions that he has researched), he walked through it in an interesting way—and came to a very clear conclusion.

In determining the rational course of action for the hitter, one must first consider the different things the pitcher can do. If the pitcher throws a ball, it’s obviously in the batter’s best interest to take ball four and walk to first base. However, if the pitcher throws a strike, the batter is still better off not swinging—this might come as a surprise, but batters who make contact on a pitch in the zone on a 3-0 count don’t do as well as the average hitter in a 3-1 count. So it doesn’t matter whether the pitcher throws a ball or a strike because the batter is better off not swinging either way.

Admittedly this is an oversimplification of the situation. Not all strikes are created equal, and if a batter gets a pitch he knows he can drive you don’t want to get in his way. But when you’re that far ahead in the count, the point remains that swinging at a suboptimal pitch is worse than letting a good pitch go by. In this situation caution is not a vice—especially when Francona is talking about a change in approach that could make his hitters feel uncomfortable at the plate.

I applaud Francona for both his willingness to buck managerial orthodoxy and the fire he is trying to instill in this Indians ballclub, and I hope we continue to see that this summer and in the years ahead. But in this case, his enthusiasm is misguided—with a 3-0 count, an aggressive mistake is very much still a mistake.