Switch hitting can’t inherently be a bad thing. Regardless of who the pitcher is, the batter can pick whichever side of the batter’s box he would be better on for the plate appearance. Players generally take the strategy of just hitting from the side opposite the pitcher’s hand. Players want the platoon advantage, but I am not convinced that this is the optimal strategy for each switch hitter.
Tune in to enough baseball games, and you will come across an announcer commenting on how valuable a switch hitter is because you always have the platoon advantage. But isn’t there also value in hitting from your stronger side all the time? No matter how good you are from both sides, you must be at least slightly better at hitting from one side of the plate than you are from the other. Essentially, this is the “there is no such thing as a tie” argument applied to switch hitting.
If your lefty/righty split is close, then switching sides when the pitcher switches hands makes sense. For switch hitters with large lefty/righty splits, then it might be smarter to just hit with your dominant hand the entire time. I went through all switch hitters, both past and present to see which ones would be better off hitting from one side, and how many runs they would add by doing it.
The idea is to project how a switch hitter would have done if he had stuck to his better side when he would lose the platoon advantage and compare it to how he did when he just switched sides to preserve the platoon advantage.
For players that showed a preference for the right-hand side, I assumed that their performance as a right-handed batter versus right-handed pitching would be equivalent to their performance as a right-handed batter versus left-handed pitching minus the average righty platoon split. I used weighted on-base average (wOBA) as my performance variable. For switch hitters who are naturally right-handed, I calculated the following:
To calculate the average platoon splits, I took all right-handed and left-handed batters with at least 1000 at-bats against both right-handed and left-handed pitchers. I then averaged their platoon splits. Similarly, I only calculated the switch hitter advantage for players with at least 1000 at-bats from both sides of the plate. With this large a sample from each player, I didn’t have to worry about the stabilization point for wOBA.
Finally, I converted the wOBA advantage for each switch hitter into runs added, by simply converting wOBA into runs. For more information on how to do that, check out this FanGraphs page.
Some might argue that I should add in some factor to adjust for the fact that switch hitters have no experience facing same-sided pitching, which is what I would be asking them to do. I am skeptical that the inexperience against same-sided pitching would affect much. James Gentile of Beyond the Box Score looked at the numbers of switch hitters who eventually stopped switch hitting when Shane Victorino considered the change in 2012. Gentile notes that there isn’t much data to draw conclusions from, and that makes it hard to come up with an adjustment factor. For the few examples shown in Gentile’s article, I don’t see any reason to include an adjustment, although it could play a role when the player first makes the switch.
I went into this project thinking I would find a high number of players who are switch hitters but should really stick to one side of the plate. Instead, I found that most players who are switch hitters should keep hitting from both sides of the plate.
I ran the numbers for the 129 switch hitters who reached the minimum number of at-bats. Of the 129, I estimate that only 23 of them have greater than a 50 percent chance of doing better at the plate by staying on one side of the plate. Those 23 players are the ones where a coach might suggest a change, but the actual number of players who would benefit by hitting from just one side of the plate is higher.
The probabilities that the other 106 players would be better off by not switching hitting is greater than 0. Therefore, we would expect some to benefit by a change in hitting approach, we just wouldn’t necessarily bet on that being the case for any one of those individuals. By summing the probability that each switch hitter would be better by staying on one side of the plate, we get a more accurate estimate of the total number of switch hitters who shouldn’t be switch hitting. This exercise shows that there are probably 39 switch hitters who should not be switch hitting. 39 of 129 switch hitters (30 percent) might sound like a high number, but most of those players should actually continue doing what they’re doing.
While those players are not taking the optimal approach, it wouldn’t be worth making an adjustment unless you are losing a high number of runs by switch hitting. A manager must make sure that his players are happy in his clubhouse. That means he can’t be telling his players to make big changes just for an extra one or two runs per season. In some cases, I estimate that a player would produce less than one extra run per season if he stopped switch hitting.
Using the formulas discussed in the methodology section, I found just 5 switch hitters who would have produced more than 20 additional runs if they had not been switch hitters. In other words, there have been only five cases in the history of baseball where I believe that a manager should have talked to his player about how his switch hitting was hurting the team. Those five players are: Dave Hollins (-77 weighted runs added due to switch hitting), Lance Berkman (-60), Jose Valentin (-55), Spike Owen (-53), and Bill North (-48).
On the flip side, 24 switch hitters added at least 50 runs over the course of their careers by switch hitting. Some of the players who benefitted from switch hitting the most are Jimmy Rollins (+136 weighted runs added due to switch hitting), Carlos Beltran (+125), and Ozzie Smith (+100). For those three players, and a few others, switch hitting helped add more than half a win each season. During Rollins’ peak, he was adding more than .7 wins each season by switch hitting (sometimes more than a full win). However, these players represent the extreme case. Most of the 129 switch hitters that I looked at hardly added much by hitting from both sides of the plate.
On average, switch hitters add 2 runs per season by being able to hit from both sides of the plate. In other words, it takes nearly five full seasons for the mere act of switch hitting to add one win. Or, a team needs five switch hitters to add one win solely from this rare skill. In fact, only 27 of the 129 switch hitters have added more than one-third of a win per season by hitting from both sides of the plate.
The data that I compiled does not necessarily say that switch hitting doesn’t help, it just says that it usually doesn’t help that much. I don’t know how valuable switch hitting would have to be to make it definitively worth it for the players who do it. Switch hitting certainly helped the careers of both Rollins and Beltran, but think of all the time a player must spend from a young age getting comfortable at both sides of the plate.
Instead of taking the time to refine two batting stances, you could just spend all that extra time becoming a better fielder or becoming a better hitter from one side. Of course, with diminishing returns to extra time in the batting cages, I doubt that the effect of getting even more practice from one side of the plate would translate to much at the major league level. The better question is how much better a player could be in other aspects of the game.
Players are rarely polished fielders by the time they reach the big leagues. I would be surprised if most of the players on that list couldn’t have added more than 2 runs each season by spending more time with the glove instead of learning how to hit in two different ways. My argument is hurt by the fact that Beltran and Ozzie Smith were two of the top three in terms of weighted runs added due to switch hitting. I still contend that those two are the extreme cases, and while it made sense for them, it rarely makes sense for the rest of the players who hit from both sides.
Smith is probably the most appropriate switch hitter of all time. What I mean by that is, The Wizard was a great fielder and baserunner. He had little room for improvement in either category, and thus it made sense to spend extra time working on hitting. Given that he added 100 runs over the course of his career by switch hitting, it’s hard to argue that he shouldn’t have done what he did. But I doubt that a player can figure out early in his career that he will be as good a fielder as Ozzie Smith.
Beltran is perhaps the most interesting case of all the switch hitters I analyzed. The average Hall of Fame center fielder had a career WAR of 71.2 and seven-year peak of 44.6 WAR. Beltran currently sits at 70.4 WAR with a seven-year peak of 44.3 WAR. These numbers back up the consensus that Beltran is a borderline Hall of Famer. What my research shows is that he would not have nearly as great a case for the Hall if he were not a switch hitter. If Beltran had only gone to bat as a right-hander, I estimate that his career would be worth just 58.2 wins above replacement. I didn’t go through each season to see how that would affect his peak, but he likely would have been knocked down 2-3 wins there, too. A WAR of 58.2 would place Beltran below Jim Edmonds, and he recently just dropped off the ballot.
The Nature of Lefty/Right Splits
There is one problem with my analysis that I have been ignoring to this point, and it has to do with the nature of lefty/righty splits in baseball. I have been acting like my estimates of what these players would do if they had hit from just one side of the plate are very accurate. The truth is that treating each player as though he would otherwise have the average lefty/righty split causes a problem.
Although the distribution of splits is normal, it is hardly concentrated towards the center. If you put a 95 percent confidence interval around each estimate, every player ends up with a range that includes a negative number, including Rollins, Beltran, and Smith. This problem is still apparent even with a smaller confidence interval. I calculated the range of outcomes we would get if we assumed that each player’s split could be within one standard deviation of the average split. The average range between the minimum estimate and maximum estimate for weighted runs added was 144 runs (more than 14 wins) over a player’s career.
The unpredictability of splits does not necessarily mean that we should throw the results out. However, it does mean that the individual estimates that I came up with are probably not accurate enough to be taken for granted. The analysis of how many players are helping or hurting their teams by switch hitting is still valid, since that is probability based. I may not be able to tell you which players are hurting their teams with confidence, but it’s still likely that 30 percent of switch hitters would be better off not doing it.
Similarly, the average of two runs added per season is likely accurate. Players are equally likely to have a larger split than the average as they are to have a smaller one. Since I averaged the weighted runs added per season across all players in the sample, the value I got has a smaller range of possible outcomes.
Ultimately, I still don’t think it’s smart to work on hitting from both sides of the plate at a young age. You could end up adding more than 10 wins from it, but you are more likely to add less than three. Work on a different part of your game with that time; you will surely find something that adds more than three wins over the course of a ten-plus year career.