Baseball has clearly had technological and analytical improvements over the last few decades. Pitching and hitting advancements have led the way, and defense has lagged behind. In the much more recent past, more accurate defensive metrics have been created, and the older ones have been updated. Defensive Runs Saved (DRS), Ultimate Zone Rating (UZR), and Outs Above Average (OAA) are known as some of the best defensive metrics. They are all made up of different components and can show different ideas. In this case however, they all agree on one thing: Mike Trout is not a five-tool player.
One of the factors that almost all defensive metrics use is catch probability. Even though there are different catch probability models, they are all very similar and use a lot of the same variables. Some variables are distance the fielder has to run, direction the fielder will have to go, and opportunity time, which can be thought of as basically the hang time that a ball is in the air.
Statcast’s outs above average is completely based off of catch probability. For example, if there is a fly ball hit to center field that has a catch probability of 40%, the outfielder would get .60 added to their OAA total if they caught it and would lose .40 if they didn’t. Similarly, if the fly ball had a .99 catch probability, the fielder would get .01 if they caught it, and would lose .99 if they didn’t (MLB.com explains it here).
Two things to note about outs above average is that is all balls that land within eight feet of the wall are not counted against players to prevent players from being hurt by their ballpark. Outs above average also does not account for throws, or other outfield errors (like letting a ground ball get by). Since, 2016, the first year they have data for, Mike Trout has only had one above average season (2018). In all the others he has been below.
Ultimate Zone Rating is similar to Outs Above Average in terms of how each play is adjusted for difficulty. UZR utilizes information from Baseball Info Solutions, so their models and numbers are slightly different. Some differences between UZR and OAA is that UZR takes a few more things into account. The first one is UZR adjusts each player’s rating for their position. They also adjust for different ballparks. OAA doesn’t do this, which is a reason that centerfielders often end up with the highest number of outs above average (they get to the most balls) and have an average above zero. UZR also takes error runs, how often players make errors compared to others at their position, into account. One last variable that UZR accounts for is outfield arm runs, which is meant to see how many runs an outfielder can save by preventing runners from advancing. Mike Trout has also struggled with UZR over his career. Comparing him with Kiermaier again from 2013 (the first season they both played) you can tell that not only has he been worse, but he has been significantly worse in most seasons.
Defensive Runs Saved is a very similar metric to UZR, but more balls in play are grouped together. This just means that some plays are valued higher or lower by DRS than UZR because it is considering it a different type of batted ball. The other difference is that DRS is used for calculating wins above replacement (WAR) for baseball reference, and UZR is used to calculate WAR for fangraphs. This year though, DRS has made some pretty big changes. They are now standardizing for different positioning, shifts, and direction of the ball. They have also changed how they are looking at two players who can make the same play. For example, if there is a high fly ball hit to left-center field, and both the center fielder and left fielder can catch it, both could be eligible for credit, or blame. Similar to the previous two statistics, Mike Trout has not been comparable to Kiermaier, and has simply been bad. In five seasons since 2013, he has had negative defensive runs saved.
All three have their strengths and weaknesses, none of them are a great finished product yet, and a lot of the time they have similar views on players. OAA has been the more predictive of future season OAA (see example below), than UZR is for future UZR, and the same for DRS. While that doesn’t mean everything, it shows that there is less variability. Once there are at least two years’ worth of new DRS data it will be time to reassess though and see if the new advancements they have made have been successful.
Those are the three most used defensive metrics, but there is other data that can help explain Trout’s low defensive ratings. First, DRS and UZR both have ratings for Trout’s arm that are not very favorable. They both agree that over the course of his career Trout’s arm has been slightly below average. Statcast can show the total estimated catch probability and the actual catch probability, which you can use to find how far above or below average a player was. They also show the outfielder’s jump compared to the average. The components of the jump are also tracked such as reaction (how much ground they cover in the first 1.5 seconds), burst (how much ground they covered in the next 1.5 seconds), and route which compares how much ground was covered in the first 3 seconds compared to how much ground they would have covered if they took the perfect route.
Despite being in the 95th percentile in sprint speed, Trout’s jumps have been some of the worst in baseball. His reaction has traditionally been very bad and has been at least negative two feet below average every year, except this one (he’s at -1.8). His burst has normally been better, but still not great, and has only been above average once in his career. This year however, his burst is at -3.1 feet, which would have been good enough to finish dead last out of qualified fielders last year. His routes have been some of the best in baseball, and that has continued this year, but his overall jump is -5.8 feet (last year’s lowest was -4.2), and he is only covering 27.5 feet (last year’s low was 28.5).
Considering that this data is not adjusted for specific outfield positions it is even stranger that Trout is doing so poorly this year. This horrible stretch could just be part of a fluke bad year, or he could be nursing an injury, but there are five years of data that suggest at best he’s an average centerfielder. The reason I’ve been comparing him to Kevin Kiermaier is because he is a centerfielder who has been in the league for roughly the same amount of time, and is legitimately elite on defense. Comparing Trout, or anyone for that matter, to Kevin Kiermaier wouldn’t do them any favors, but it helps show just how far away he is from being elite, when a lot of the public thinks he might belong there. He finished just behind Kiermaier in gold glove voting last year and has been a finalist for the award three other times. That shows the value of advanced defensive metrics. It is impossible to watch every play be able to judge it perfectly and compare to other players. There will undoubtably be people who use little evidence to claim that Mike Trout’s defense was on par with other elite outfielders like Kiermaier’s. I’m definitely not a Mike Trout hater, he is the best player of this generation, but I am saying that right now he is a three-tool player.
|Crafting a Gameplan… on The Jordan Hicks Dilemma: A Qu…|
|SB on Zach Davies: What Changed?|
|Can Detroit Tigers B… on Improving Pythagorean Winning…|
|Avoid the Dead Zone:… on An Analysis of Jakob Junis, Ar…|
|Predictive Fitness a… on Press Release: BaseballCloud A…|