Interesting article by Ken Rosenthal on lineup protection. One quote that caught my eye was this...
Bill James, godfather of the statistical movement, called such protection [the idea that the Phillies wanted to sign Alfonso Soriano to protect Ryan Howard] "overrated" and told the Philadelphia Inquirer that he would project Howard's numbers to be the same whether Albert Pujols or Corey Patterson were hitting behind him.He then goes on to say:
With all due respect to James, Pujols and Patterson are not equal, never will be equal and do not merit equal consideration. "Do not let this guy beat you," is a mantra at pitchers' meetings. Previous matchups, left-right concerns and ballpark effects all figure into a pitcher's approach. Still, according to statistical research, the difference in the Phillies' No. 5 spot between relatively comparable batters such as Burrell and Jimmy Rollins would be marginal.Of course James was not saying that Pujols and Patterson are equals, only that their effect on the performance of Howard would essentially be equal.
Rosenthal goes on to cite J.C. Bradbury's study on protection where he concludes that contrary to a protection effect, a good hitter actually has a negative effect on the previous hitter. Bradbury calls this the "effort externality", which occurs when a good on-deck batter causes the pitcher to use more effort to pitch to the current batter, thereby lowering the hit probability of the current batter. In addition, this means the pitcher will throw more pitches in the strike zone which decreases the walk rate. Bradbury sums up the study in this way.
Using MLB play-by-play data from 1984–1992, we are able to control for many factors, including the quality of the on-deck hitter, that influence the probability that a player reaches base. We find that the quality of the on-deck hitter negatively impacts the preceding hitter, which is consistent with pitchers using extra effort to minimize the damage of the batter who follows. This finding contradicts traditional baseball thinking that a good on-deck hitter protects the hitter who precedes him in the batting order. Though we find that a spillover does exist, its magnitude is very small. It takes a very large difference in the quality of the on-deck hitter to have only a minor impact on the current batter.