When to Walk David Ortiz by Matt Klaassen October 30, 2013 It is a bit of an exaggeration to say that David Ortiz has been the Red Sox’ whole offense during the World Series, but only a bit, as he is hitting .733/.750/1.267 in the Series so far. Of course, he hit .091/.200/.227 in the ALCS versus Detroit. The reader probably understands both sets of numbers to be both small sample fluctuations — Ortiz is an excellent hitter, but making decisions based on any short span of plate appearances is a bad idea. The temptation to give Ortiz the “Barry Bonds treatment,” i.e., walk him every time he comes to the plate, no matter what the situation, is understandable, but should be resisted. Even if Ortiz is no Bonds, he is an excellent hitter, and clearly the Red Sox’ best. Although the intentional walk in general seems to be overused, there are situations in which it makes sense, especially with the Cardinals sending the right-handed Michael Wacha to the mound tonight and the Red Sox (probably) hitting several right-handed hitters behind the left-handed hitting Ortiz. When to walk or not walk is not a clear cut situation, but using some general principles, we can at least outline some basic game situations when it is might be the right idea. I first should say that this post basically follows the analysis of intentional walks found in Chapter 10 of The Book. I will repeat as few of the specific details from the book as possible, since one can always pick up the book for those. Also, for the sake of space and sanity I will not be able to cover every combination of base, out, game situations, reliever combination, batting order possibilities, and so on. This post will just deal some of the basic ones. The matter of how to deal with “intentional unintentional walks” will also not be treated here. Almost every intentional walk will increases run expectancy. There is one general exception to this rule — runners on second and third, two outs, and an elite hitter — say, Albert Pujols in his prime — at the plate. This general analysis of run expectancy does not take into account the game situation (inning and relative score) and the quality of the hitters behind the elite hitter. The Book’s chapter on walks eventually gets to a chart of base/out/inning/score situations and when a walk is advisable. Each state has a specific ratio associated with it — this is the ratio of the potential walkee’s expected wOBA to that of the hitter(s) behind him. Depending on how many outs into the inning the game is at, the ratio is the elite hitter’s expected wOBA divided by a weighted average of the expected wOBA of the two, three, or four hitters following him (i.e., how many batters might expected to hit on average after the intentional walk). If the ratio is equal to or higher to the ratio on the chart, the intentional walk is a good idea. So we need a batting order. As of this writing, tonight’s batting order has not been posted. Based on recent batting orders and who is reported to be playing tonight, after Ortiz, I am guessing the next four batters will be Mike Napoli, Jonny Gomes, Xander Bogaerts, and David Ross. It may turn out to be different, especially the Bogaerts/Ross/Stephen Drew sequence, but this is my best guess after consulting with others, and the first two batters after Ortiz matter the most. Despite all the numbers involved, I also do not want to give the impression that this is a super-precise sort of calculation and decision. There is a bit of wiggle room, but taking all of that into account involves many details, most of which are not accessible to most of us (e.g., how the the pitcher feeling about a particular pitch he likes to use versus lefties). Since we are not going by hot and cold streaks or observed performance generally, we need an estimate of current true talent for the relevant players. For this, I used Steamer’s 2014 projections — yes, it is for next season, but it is probably close enough for the moment. The Book recommends removing the walks (and hit by pitch) from both the numerator and denominator of the wOBA calculations for the player in question (Ortiz) then multiplying by 1.12, since in each situation, an intentional walk, unintentional walk, and hit by pitch have the same game impact. Finally, on the basis of Steamer projections, I also estimated each player’s platoon skill. (Hopefully the reader can see why I will only look at a few situations. To be more precise one would also want to figure in each pitcher’s platoon skill and one can get even more complex from there, but I want to keep things relatively simple). Ortiz is no slouch versus left-handed pitchers, either, but he has a long history of being better versus righties than lefties. With the Red Sox probably following Ortiz with at least two right-handed hitters in Napoli and Gomes (the latter has a particularly large platoon split), there may be some situations in which a right-handed pitcher slated to face that bottom part of the order might want to walk Ortiz. In the bottom of the first, the Cardinals need to have fallen behind by a pretty wide margin in order to justify walking Ortiz. For example, they need to be down by five runs with two outs and a runner on second for the walk to Ortiz (who would be coming up bat for the second time in the inning!) to make sense. In the bottom of the second, much the same is true — the Cardinals would need to be down three or four runs (it is close) with one out and runners on second and third or two outs and a runner on second to walk Ortiz. As the game progresses, the number of situations in which intentionally walking Ortiz with a right-handed pitcher slated to face him and the following batters increases, but just slightly, and it is close. They mostly occur with one or two outs and runners on second and third, or two outs with runners on second or third. The Cardinals would need to be behind (and again, I’m simplifying to avoid just listing every case): by at least three in the third, two in the fourth, or one in the fifth. By the bottom of the sixth, walking Ortiz with a righty on the mound is recommended even in a tie game with runners on second and third and one out or a runner on second and two outs. In the seventh, with second and third occupied and one out, the Cardinals might even want to walk Ortiz if they are up by one. Let’s fast-forward to the ninth inning, since that is the “big inning” when the Cardinals would have their closer in, and would be ahead since it would be the bottom of the ninth in Fenway. Again remember that for the sake of simplicity, we are not calculating Trevor Rosenthal’s (assuming he is in the game) own platoon skill here, although it would be a good idea to do so. In the ninth, there are a number of situations in which walking Ortiz to let the right-handed pitcher face those behind him is a good idea: With the game tied and pretty much any number of outs and a runner on second or third (this is close), first and third, or second and third. If the Cardinals are up by one, a walk makes sense with runners on second and third and any number of outs, a runner on second and two outs, a runner on third with two outs, and runners on second and third with two outs (this last one works if the Cardinals are up by two, as well). As one can see, Ortiz is likely so much better against a right-handed pitcher than the batters following him that an intentional walk is defensible in a variety of situations. But it is hardly the Bonds treatment. Moreover, if the manager is ready to go to his bullpen, he could bring in a southpaw to face Ortiz straight up then a right-handed pitcher to face the batters following him, forgoing the intentional walk altogether (due to the length of this piece, I have not gone through the ratios for this possibility). Of course, no matter what the manager does, we know that the unexpected is to be expected, especially in this particular series. Ortiz might get a free pass in the wrong situation followed by a double play ball, or he might smoke a key homer off of a lefty brought in to face him. Even if the calculations above are basically correct (and once again, this is a simplified version), they are not guarantees. Thus, we watch the games.