Jim Albert tries to find the optimum first swing percentage.

As we will see below, there is a lot of variability in the proportion of first pitch swings among regular players. It would see that a moderate value of this swinging proportion would be optimal. If a batter always will take the first pitch, then the pitch can just throw a fastball down the middle of the plate. On the other hand, if the pitcher knows that the batter will swing at a high proportion of first pitches, then the pitcher likely would throw a pitch outside of the strike zone that the batter would chase. To keep the pitcher guessing, it would seem advantageous (from the batter’s perspective) to swing at a “middle” proportion of first pitches.

I have seen very successful hitters who seldom swing at the first pitch, but they tend to be extremely good at strike zone judgement and putting the bat on the ball (Wade Boggs might be the prototype). His sample size is small, but the data does appear to support his theory.