Baseball’s (publicly available) defensive metrics are bad. I mean, there’s really no way around this conclusion – they do not correlate well with themselves year-to-year, many simply throw out plays with defensive shifts, and, as that link notes, there’s no way for current systems to account for positioning at the beginning of a play.
The last two points are important, as illustrated by this excellent BaseballProspectus article from last year, in which Colin Wyers examined Brett Lawrie’s insanely high DRS ratings. The gist of it was that Lawrie, a third-baseman, played out in short right-field against heavy-pull hitters but was being credited for plays out there as if he was ranging all the way over from his normal third base positioning.
While such a situation is extreme and rare, there’s a lesson to be taken regarding the rating of ever player in the game: don’t put too much stock in them if they’re not accounting for where a player starts the play.
It should go without saying, but pre-pitch positioning is an enormous important part of fielding a position. If you played little league (I’ll get to Dexter Fowler, give me a minute), you remember being instructed to move in for the tiny players at the plate, or move to the extreme pull-side of a good hitter facing a soft-tossing pitcher. (Er, well, I hope you remember, otherwise your coach or the random dude running your team wasn’t doing their job).
This same thing happens at the major league level to a very detailed extent. Infielders know the location and type of the upcoming pitch, and lean or move in a corresponding direction as the pitcher goes into his delivery. Outfielders and infielders alike know the tendencies of the batter and position themselves where hitters are most likely to hit the ball.
Coaches and managers have binders full of this data and direct shifts/positioning from the dugout. The half-step gained by proper positioning can be the difference between a highlight-reel play and a nondescript base hit.
As noted by FanGraphs’ Mike Petriello, new Cub Dexter Fowler’s positioning is extreme – almost no one plays as shallow a centerfield as Dexter Fowler. This is of note, as almost no one can match the horrific -44 Defensive Runs Saved he’s posted in his six full seasons as a centerfielder.
Is there a connection here? Is there reason to think that he’s been a better centerfielder than the numbers would suggest? You amateur detectives who’ve read the title to this piece may know the answer already, but let’s take a much deeper look at Fowler’s defensive chops, starting with his time as a prospect.
At one point in time, Dexter Fowler was a premiere prospect. Both Baseball America and Baseball Prospectus rated him as a Top 15 Overall prospect before 2009. He was coming off a year in which he torched AA to the tune of a .946 OPS, but he wasn’t expected to replicate that sort of production as a big-leaguer. Rather, his carrying facets were his hit tool, his excellent eye (which is technically a skill, not a tool), his speed, and his glove.
Now, call me naive, but I think that scouting a player’s glove is a relatively easy task. Trying to tell whether or not an 18-year-old is going to be able to hit major-league breaking stuff is next to impossible, but defense is much simpler. Speed, whether or not a guy gets a good jump, route running are all pretty apparent in the outfield. So when the scouts and the defensive metrics don’t line up, I side with the scouts.
As Fowler was progressing through the minors, his glove work and arm drew high praise. To quote two scouting reports from the late 2010’s:
There’s really nothing he can’t do… is a plus-plus runner, and he’s an outstanding center fielder with a good arm. – Baseball Prospectus’ Kevin Goldstein (current member of Astros’ Front Office)
It’s tough to imagine that Fowler immediately went from one of the best defensive center fielders in the minors to a guy who should be in a corner almost overnight, but that’s what popular defensive metrics would have you believe. And it’s not like his reputation as a solid defender has gone away among scouts – Bernie Pleskoff, a former MLB scout, tweeted some support for Fowler’s defense:
Dexter Fowler adds an outstanding defensive outfielder to the new #Cubs mix. Center field will be well played. Another piece that adds depth
— Bernie Pleskoff (@BerniePleskoff) January 19, 2015
I couldn’t disagree more. They don’t reveal what my eyes see when I watch. And no, over his career that isn’t a lot of Es. Not at all.
— Bernie Pleskoff (@BerniePleskoff) January 19, 2015
Still lots of people disagreeing with me on Fowler’s defense. I respect metrics-but I also think personal observation is the best method.
— Bernie Pleskoff (@BerniePleskoff) January 19, 2015
When you watch Fowler play defense, he doesn’t look like a butcher out there. You know what a league-worst defender at a position looks like if you watched Chris Coghlan play left field last year, and Fowler simply isn’t that kind of player. Players who are true atrocities in the outfield just do not make a play like this. Fowler gets a good jump, covers 150 feet of ground with a perfect route, and makes a full-extension diving catch. At the very least, the raw tools to be an adequate centerfielder are there.
I understand that a series of highlights doesn’t tell the full story, but I would argue you can actually learn quite a bit about a player’s raw skills in the field. You won’t just see the potential upside of a guy like in the linked video, as there’s often something funny about highlight-reel catches: they’re the result of a poor route or jump.
Players athletic enough to patrol a MLB team’s center field can often recover from poor jumps or bad routes and make a flashy catch. I’m not gonna sit here and tell you that Fowler’s highlights make him a better outfielder than metrics say, but I strongly encourage you go through some to get a feel for his defense. I’ll link some videos in a comment on this post.
So if Fowler looks like a capable centerfielder, and scouts say he’s a capable centerfielder, why do metrics like DRS hate him so much? My guess is that it’s all in the positioning for Fowler, and I (as luck would have it) have numbers to support that.
Last spring, FanGraphs added fielding numbers from Inside Edge Fielding to their website. Inside Edge uses a team of evaluators to judge how difficult a given play is to make. They rate them as “impossible” (0% chance to make the play) to “routine” (90-100% chance to make the play) with a few different ranges in between.
I love this type of data because I think it eliminates the problem posed by the initial positioning issue. The evaluators who watch each play don’t have lines superimposed on the field letting them know where a player’s zone is, or know the trajectory angle of a fly ball.
Instead, they can see that a fly into the left-center gap will be particularly tough for a CF toward right field, or that the low screaming liner was hit right at a shallow defender. While this system is subjective, it helps solve an enormous problem that previous metrics cause.
Interestingly, and perhaps surprisingly, Dexter Fowler rates as a pretty decent center fielder according to the Inside Edge metrics. Below is Fowler’s performance from 2012-2014 according to Inside Edge. (How to read: The “Remote” column shows Fowler’s performance on plays that he had only a 1-10% chance of making. In 2012, Fowler made 11.1% of the 9 plays that were judged to be a “Remote” chance to make)
Of course, this isn’t much use without knowing how often all centerfielders make these plays. Using a sample of all outfielders to play at least 500 innings in CF in any of 2012-2014, I figured out what the average made-play rates were for starting center fielders over that time. This sample included 47 separate outfielders who have combined to play nearly 100,000 innings in center field since 2012. The results are below:
Fowler’s performance with respect to these averages are seen below in the “%Made+” column. The numbers you see are Fowler’s performance divided by the starting CF average for each play, normalized so that 100 is the same as average. Higher than 100 is better than average, below 100 is worse than average.
As you can see, Fowler’s performance with respect to his peers has been pretty impressive. He’s made a “remote” play in each of the last three years, something just 10 center fielders in the sample have done. His performance on “unlikely” plays has been great as well, as only 10 other center fielders made 6 or more “unlikely” plays.
Where Fowler does “struggle” relative to his peers is on plays is on plays rated “even”, “likely”, or “routine”, and even then he’s right there with the rest of the pack.
Of course, this doesn’t account for the discrepancy in chances from each player. To adjust for this, I used a plus/minus type system as detailed by this amazing post on Beyond the Box Score. Basically, if the average starting centerfielder in this sample was expected to make 8.1% of “remote” plays, a fielder would be rewarded 1.081 “points” for for making such a play, while being penalized .081 “points” for not making the play.
This is done for all plays in the Inside Edge database. Luman, the post author, calls this IERP. Fowler’s IERP of 32 over the last 3 seasons is good for 14th in this sample. From there, if you adjust IERP based on total chances, (not including “impossible” plays), you get an “IERP+”. Fowler’s IERP is a 93, which suggests that he’s a slightly below-average defensive center fielder.
An interesting note that I don’t know what to do with: Fowler’s IERP+ is below average due to a very low 2013 value. Why is this? Well, it appears that in his final season in Colorado, Fowler had far fewer chances at Unlikely, Even, and Likely plays than you would expect given his innings totals. His DRS and UZR for 2013 were also the highest they’ve been in the same time span. Did Walt Weiss figure out how to position Fowler correctly?
Given all this, I have to say I’m pretty bullish on Fowler’s chances to be a useful defensive outfielder in 2015. He still has the raw tools he had when everyone thought he’d be an above-average defender, at least one scout still thinks he’s a great defender, and Inside Edge fielding metrics (which integrate defensive positioning to an extent, unlike DRS, UZR, and others) rate him as at least a slightly-below-average defender.
If I had to guess, Fowler’s been working with poor coaches who have been unable to correct his positioning and place him in a position to succeed.
If any group of coaches is going to turn Fowler into a good defensive center fielder, it’s probably this one.
Thank you to FanGraphs, which supplied all metrics used in this post.