By Peter Jensen
Recently Mitchel Lichtman, known as MGL, wrote a two part article for The Hardball Times that looked at the strategy of intentionally walking a batter with men on second and third and one out in innings 1 through 6. He compared the runs scored in the inning after an intentional walk to the run expectancy for the second and third occupied one out base out situation adjusting for the quality of the batter and concluded that the intentional walks in that situation cost the defensive team runs and that “except perhaps in rare, ideal situations, you are probably reducing your team’s chances of winning.” I have used a different methodology to analyze the same situation and have come to a different conclusion.
I looked at a more recent set of data, 2005-2007 using just the American League just as MGL did for the years 1998-2007. Instead of looking at runs scored I looked at whether a team won or lost the game when it was ahead or tied at the time it either intentionally walked a batter in the 231 base out situation or decided not to intentionally walk the batter. For 2005 through 2007 in the American League teams faced a 231 base out situation prior to the 7th inning 1647 times. They intentionally walked the batter 141 times. They never intentionally walked the batter when they were ahead by three or more runs. They did walk the batter when they were behind by three or more runs. But at that point, facing a situation that was likely to put them further behind no matter what they decided to do, it is unlikely that walking the batter negatively affected their win probability appreciably.
So the relevant situations are when the team is between 2 runs ahead and two runs behind. I divided these situations into two groups for analysis. In Group 1 I looked at the situation where the defensive team was up by 1 or 2 runs or the game was tied. There were a total of 696 situations in this group and 42 intentional walks were issued. I then defined a bad decision as one in which the defensive team, who was tied or winning, ended up losing the game. In the 654 times that a team didn’t issue an intentional walk they ended up losing the game 373 times or 57% of the time. MGL’s analysis would predict that when the teams did issue an intentional walk in this situation that they would lose more often. In fact, they lost slightly less often, 23 of the 42 times the batter was intentionally walked or 54.7% of the time. This is true even before we consider that the batters that were intentionally walked were better hitters as a group than the batters that were not intentionally walked. The batters that were walked batted third, fourth, or fifth in the lineup 53.2% of the time while the batters that were not walked intentionally were number 3, 4, or 5 hitters only 33.2% of the time. Presumably, if these middle of the order batters had not been intentionally walked they would have added to the percentage of “bad decisions” in the not walked column. There is certainly no evidence here that intentionally walking a batter in this situation is creating extra lost games.
But perhaps intentional walks that a team issues when they are already behind a run or two are hampering their ability to overcome that deficit and come back and win the game. That is group 2, the situations where the defensive team is behind by one or two runs. There were 448 of these situations and the batter was intentionally walked 50 times. Teams did not come back to win the game from this situation very often. Of the 398 times when the batter was not intentionally walked the defensive team won only 106 times or 26.6% of the time. When the defensive team intentionally walked the batter they did indeed lose some of their ability to make a comeback. They walked the batter 50 times when behind 1 or 2 runs and only came back 10 times when they might have come back 13 times had they not walked him. But remember this does not take into account the intentionally walked batter’s better ability.
Overall the strategy of intentionally walking the batter may have resulted in a net loss of two games over 3 years, but probably not considering the difference in ability of the batters that were walked and the batters that were allowed to hit. Another factor that comes into play is that 30 of the 141 intentional walks did not start as intentional walks, but only became intentional after the batter had gone ahead in the count. Being ahead in the count would increase the advantage that the walked batters already had. This factor was not addressed in MGL’s analysis. Overall it is likely that the way that American League managers have employed the intentional walk during the last three years has not resulted in any reduction of win potential, and there is no good reason to advise them to change their current strategy. In fact, there is some evidence that they may be wise to adopt the unorthodox strategy of putting the tying run on base on at least some occasions when they are three runs ahead with men on second and third and one out. As mentioned earlier teams never walk the batter in this situation, but when the third, fourth, or fifth batter in the lineup is allowed to hit in this situation they lose 50% (8 of 16) of the time. The expected runs with the 3, 4, or 5 batter hitting in this situation is 1.875, but the expected runs with the 4, 5, or 6 batter hitting with the bases loaded and 1 out is only 1.657. But any statistical advantage would probably not be worth the avalanche of criticism that a manager would receive for going against the conventional wisdom of never putting the tying run on base.
Both the method that MGL employed and that method that I employed are only rough estimates that are too crude on which to base strategy decisions. The only accurate way to assess any particular strategy decision is to use a comprehensive simulation that takes into account every conceivable factor that might affect that decision. MGL has come to the conclusion that he has enough information to be able to offer advice to managers that will save a team wins. My conclusion is that I don’t know enough yet to advise a manager to change decisions that I am pretty sure are neutral at worst.