As it usually is, the strike zone was at the center of quite a few discussions during the recently completed trouncing billed as the World Series. Mark Bellhorn‘s tendency to strike out and walk prodigiously, as well as the entire Red Sox team’s reluctance to swing and miss in the World Series, were discussed near and far. This got me thinking about who the game’s best and worst hitters are when it comes to choosing when to swing and when to spectate, and how that translates to their performance, particularly power numbers.
Even if you had never seen baseball before, you could infer from the multitude of replays and even the superfluous dirt-cam introduced in the World Series this year that an at-bat is a complex series of events that requires lengthy analysis. Or you can divide it into two separate events: decision and result. From the batter’s perspective, the decision is simple: swing or do not swing (there is no try). Once that choice is made, the batter can additionally influence the outcome if he chooses to swing, or he transfers the decision to the umpire if he does not.
In order to determine if the decision was “good” or not, we must evaluate both the choice and the effectiveness of that choice. For example, if the batter chooses not to swing, that choice can be deemed “correct” if the pitch is called a ball or “incorrect” if it’s a strike. (Though many of you may debate that in light of some of the recent strike zone interpretations by our friendly umpiring crews.) Once the batter has decided to swing, the results become more varied and therefore more difficult to evaluate.
Before we evaluate the effectiveness of each choice, let’s take a look at who tends to make which choice. For all charts, pitchouts and intentional balls will be excluded because, unless the batter is Roy Hobbs, there’s no choice involved in those situations. Here are the leaders in both the percentage of pitches swung at and taken in 2004:
Player Take Swing ------------------------------------ Todd Zeile 66.8 33.2 Chris Snyder 65.8 34.2 D'Angelo Jimenez 64.9 35.1 Frank Menechino 64.5 35.5 Luis Castillo 64.3 35.7 Jason Kendall 64.3 35.7 John Olerud 64.2 35.8 Bobby Abreu 63.9 36.1 Nick Punto 63.9 36.1 Barry Bonds 63.8 36.2 Player Take Swing ------------------------------------ Vladimir Guerrero 39.7 60.3 A.J. Pierzynski 40.2 59.8 Jesse Garcia 41.9 58.1 Johnny Estrada 42.1 57.9 Todd Greene 42.7 57.3 Dmitri Young 43.1 56.9 Jose Molina 44.0 56.0 Rey Sanchez 44.2 55.8 Bengie Molina 44.4 55.6 Jose Vizcaino 44.6 55.2
Looking at these two groups of players, it appears at first glance that the decision whether or not to swing has virtually nothing to do with overall performance, especially considering that the likely NL MVP is among the top 10 in pitches taken and the likely AL MVP is at the extreme opposite end. There are both good and bad players on both lists, though perhaps a few more reputable ones in the “take a pitch” group. For now, we can assume that both approaches at the plate are equally viable.
Instead, let’s try to evaluate the results of the choice rather than the choice itself. First, how do our hacktastic friends fare when they swing so liberally? We can break up the results of a swing into three to four categories: miss, foul, in play, and hit:
Player Miss Foul In Play Hit in Play ------------------------------------------------------------ Vladimir Guerrero 17.0 39.9 43.1 37.7 A.J. Pierzynski 13.5 37.4 49.2 28.5 Jesse Garcia 17.7 33.3 49.0 29.0 Johnny Estrada 14.5 45.2 40.3 36.2 Todd Greene 31.9 29.7 38.4 34.8 Dmitri Young 21.2 40.4 38.5 32.9 Jose Molina 33.3 28.6 38.1 33.4 Rey Sanchez 10.1 36.6 53.3 26.5 Bengie Molina 14.7 39.0 46.3 30.2 Jose Vizcaino 12.4 38.2 49.4 30.1
To clarify, the first three categories are the result of the swing while the fourth is the percentage of balls put into play that become hits (note that this will not match up with pitcher BABIP or Defensive Efficiency since home runs are being included). Looking at this group, it’s hard to constrain all the players with any of the conventional ideas about free swingers.
There seem to be several different kinds of hitters in this list. Vladimir Guerrero, Johnny Estrada, and Bengie Molina all post similar Miss/Foul/In Play numbers, but Molina’s H/BIP ratio falls well short of the other two. So while their approaches are all free-swinging and their results when swinging are all very similar, there appears to be a marked difference in skill at producing hits from that performance.
Likewise, A.J. Pierzynski, Jesse Garcia, and Jose Vizcaino all do an excellent job of both making contact and putting the ball in play when they swing, but despite very similar H/BIP ratios, Pierzynski (.272/.319/.410) and Vizcaino (.274/.311/.374) come out significantly ahead of Garcia (.252/.265/.330). Again, with this group, even similar decisions and performance can lead to disparate results, making analysis based simply on these ratios dubious.
Looking at the other extreme approach, perhaps we can find some significant difference in results. Back to our list of patient hitters, here’s how that group fares when they do finally find a pitch they deem worthy of a swing:
Player Miss Foul In Play Hit in Play ------------------------------------------------------------ Todd Zeile 15.2 39.6 45.2 30.2 Chris Snyder 20.7 35.4 43.9 31.9 D'Angelo Jimenez 9.6 40.8 49.6 32.4 Frank Menechino 12.7 38.8 48.6 33.9 Luis Castillo 8.4 37.7 53.9 32.5 Jason Kendall 7.7 37.2 55.1 34.0 John Olerud 10.2 37.4 52.4 29.7 Bobby Abreu 15.8 41.7 42.5 37.2 Nick Punto 8.8 42.6 48.6 31.9 Barry Bonds 12.6 38.2 49.2 40.3
While the first group averaged 18.6% miss, 36.8% foul, 44.6% in play, and 32.0% H/BIP, this group averages 12.2, 38.9, 48.9, and 33.4. The more patient hitters certainly do a better job of hitting balls they swing at and a slightly better job of turning those balls in play into hits, but the different is minor at best. The variance among this group is nearly as wide as the free swingers, a fact that does not bode well for a player like Todd Zeile, who carefully selects which pitches at which to swing, and then only puts 45.2% of them in play, converting 30.2% for hits, numbers worse than Molina’s.
Instead, a more complete look at each hitter would also include the results of not swinging. Here’s how our patient group does when taking a pitch, something they all know how to do quite well:
Player Strike Ball ------------------------------------ Todd Zeile 42.1 57.9 Chris Snyder 37.3 62.7 D'Angelo Jimenez 35.6 64.4 Frank Menechino 37.1 62.9 Luis Castillo 37.4 62.6 Jason Kendall 41.1 58.9 John Olerud 36.6 63.4 Bobby Abreu 34.9 65.1 Nick Punto 32.4 67.6 Barry Bonds 25.6 74.4
And the hackers:
Player Strike Ball ------------------------------------ Vladimir Guerrero 14.2 85.8 A.J. Pierzynski 25.6 74.4 Jesse Garcia 41.5 58.5 Johnny Estrada 27.8 72.2 Todd Greene 22.9 77.1 Dmitri Young 21.3 78.7 Jose Molina 31.8 68.2 Rey Sanchez 29.3 70.7 Bengie Molina 23.5 76.5 Jose Vizcaino 31.4 68.6
The free swingers average a 27:73 strike/ball ratio while the more patient hitters manage 36:64. Does this mean that our aggressive group possesses a nine percent better batting eye than the patient group? If you assume that all strikes are hittable pitches and that batters only swing at hittable pitches, then yes, but that’s far from the truth.
The difference isn’t as great as you might imagine. Remember that we found that our free swinging group put the ball in play 4.3% less often than our patient group, but still managed to nearly match the patient group in H/BIP. Based on that, the free swingers appear equally adept at converting their swings into hits; it would be difficult to argue that their aggressive approach is costing them anything other than a bit more effort from the opposing pitcher. (You could argue that they’re losing walks as well, but if they do reach base as often as the patient hitters when putting the ball in play, it’s likely that the fewer walks are mitigated by the increased hits.)
In fact, expanding the analysis to all major league hitters in 2004, there is absolutely no correlation between a hitter’s approach at the plate (as measured by Take v. Swing) and H/BIP. So whether a batter swings aggressively or not, he has just as likely a chance of getting a hit when he puts the ball in play.
The other main point about being patient and waiting for a good pitch to hit is that it yields higher slugging numbers. This past season, our two groups above averaged an ISO of .148 for the free swingers and .165 for the patient hitters. While that may look impressive initially, if we expand the lists to the top 25 at each extreme (to counteract the Barry Bonds effect), the free swingers actually pass the patient hitters with an ISO of .154 to .144. In 2004, sitting back and waiting for your pitch actually yielded less power than swinging at the first thing that comes your way, at least when looking at the extreme groups above.
While it seems clear that both approaches can work well, it may be possible to analyze how well certain hitters are making their decision based on the outcomes of their most frequent choice. Expanding the sample once again, out of 439 players this season who had a minimum 100 PA, if we take the top 125 at both ends of the Take/Swing spectrum, we can analyze within those groups if other characteristics lend themselves strongly to performance.
Running a quick regression for both sets, however, yields little to no correlation between any outcome percentages (such as taking strikes versus balls or putting the ball in play when swinging) and outcome performance metrics like SLG or ISO. Using multiple variables does increase the model towards significant correlation, but no combination of variables shows an R-squared of higher than .4734. Effectively, even in the most significant situation, over half of the variance in performance metrics cannot be accounted for using outputs from batter abilities once they choose to swing or not swing the bat.
Looking rather at the aggregate statistical lines for each of the two samples, the patient hitters hit .260/.348/.417 with a .157 ISO while the aggressive hitters produced .262/.313/.421 with a .159 ISO. So while it’s certainly true that players who take a few more pitches will walk more often, they actually show slightly less power and batting average than the more aggressive group. This confirms what we saw when trying to map various at-bat measures to batting performance in the previous paragraph: sitting back and waiting for a pitch to hit doesn’t necessarily translate into more power.
This point brings back into question who is really in control of the at-bat. It is likely that patient hitters with good power simply don’t see as many pitches to hit and that skews the results we’re using here. Even having removed intentional balls and pitchouts, it’s difficult to compensate for at-bats where pitchers simply don’t get the batter anything to hit, preferring a likely walk to the chance of a home run. However, if pitchers were using this approach based on a batter’s power numbers, we would still expect to see a correlation between power numbers and a high Take Percentage.
So what has all this meandering about shown us? First, that there truly are many varied approaches to hitting, any one of which has the potential to be successful for an individual player. Some players swing very aggressively while others wait for a pitch they believe they can handle; each of these approaches is equally likely to generate both a certain percentage of hits on balls in play and a certain ISO or SLG. While patient players (or at least player who recognize when they’re not being given a pitch to hit) show distinctly better walk totals, they show virtually identical power numbers to their free-swinging counterparts. While it’s not entirely convincing that patience doesn’t walk hand in hand with power at the major league level, their apparently complete independence as demonstrated above shows that power comes from a variety of batting approaches.
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.
Subscribe now