The Fans, The Metrics, and Dexter Fowler

Dexter Fowler played 959 innings in center field for the Astros in 2014, which is noteworthy mostly because he was traded to the Cubs on Monday, but also because of how poorly the defensive metrics rated him for those 959 innings. By Defensive Runs Saved (DRS) he was 20 runs below average and by Ultimate Zone Rating (UZR) he was slightly worse.

Both metrics dinged Fowler for a poor arm, something that’s followed him for his entire career, but for the third time in his career, UZR gave his range horrible marks relative to other center fielders. DRS had previously been friendlier, but didn’t hold back in 2014 with respect to his range.

Fowler’s defense rates poorly by the leading defensive metrics with his career DRS and UZR per 1,000 innings sitting somewhere between -7 runs and -11 runs. Now this is almost exclusively in center field so Fowler gets a few of those runs back on the positional adjustment, but we’re basically talking about a bad center fielder who might be an average guy in a corner. This is all according to the metrics which, as plenty of people will remind you, are imperfect.

This is particularly true in Fowler’s case as there’s a sizable contingent of people who think more of his defense than the metrics do. The Inside Edge numbers also look more favorably on Fowler, as indicated by Jeff Zimmerman’s Plays Made Ratio.

As a rule, DRS and UZR are better than your eyes and my eyes. If someone handed us an envelope with divinely created defensive metrics, I would wager that our optical intuition about players would be less accurate than the metrics. But that wouldn’t always be true. There are certainly some cases where the metrics miss something that we are capable of seeing and understanding.

The first thing that comes to mind when you look at Fowler’s ratings is positioning. We know that DRS and UZR assume normal positioning but Inside Edge is basing their ratings on where the player actually starts (to the best of their ability). Given that this problem followed Fowler to Houston, it’s probably not just about his Rockies coaches or something goofy about Coors Field. We can’t discount that entirely, but the season in Houston helps sort some things out.

If Fowler is really bad at deciding where to position himself and his coaching staffs have been relatively hands off on plays without extreme shifts, then it’s not surprising that the disagreements about him exist. It’s really hard for fans to see if a player starts out of position. They can judge his closing speed and his ability to cover ground, but if he started fifteen feet out of position, then he might not get to a baseball even if he was flying. Hopefully StatCast will solve some of these problems, but we don’t have StatCast today.

One of the keys here is that positioning is extremely important and if you’re out of position often enough, it doesn’t really matter how fast you are or how well you climb walls. After nearly 6,000 innings in the big leagues, we should be pretty confident that Fowler is a below average center fielder even if we can’t be sure exactly why that’s true.

Yet there’s something interesting going on with the Dexter Fowler perception. Let’s start with our new FAN Projections compared to Steamer for 2015.

Source	PA	BABIP	AVG	OBP	SLG	wRC+	BsR	Off	Def	WAR
Steamer	582	0.318	0.252	0.353	0.379	107	1.8	6.5	-10.8	1.4
Fans (29)	593	0.340	0.270	0.368	0.399	117	0.2	11.6	-6.6	2.5

Not surprisingly, the fans are more bullish on Fowler. They like him for four or so more runs with his glove than Steamer and five or so more runs coming from a higher BABIP. You can probably sell the BABIP more easily given his career marks, but there’s still the matter of Coors Field. Given everything we know about Fowler’s perception and his stats, it’s perfectly reasonable that the fans would like him more, especially with respect to his glove.

If we turn this around, however, the other fans are starting to notice Fowler’s defensive issues. Glance at his Fan Scouting Report defensive numbers over the last few seasons and you’ll notice an obvious trend. Fowler’s first step, speed, hands, release, arm strength, and arm accuracy ratings are all trending south and some took a big jump down in 2014. His overall rating has dropped ten points in just two years.

Season	Inn	Votes	Instincts	First Step	Speed	Hands	Release	Arm Strength	Arm Accuracy	Overall
2009	977.2	44	63	83	89	50	56	53	51	64
2010	948.1	26	59	79	87	68	54	44	49	63
2011	1072.2	19	56	72	85	62	51	48	53	61
2012	1026	16	56	68	83	58	50	48	62	61
2013	921.1	11	52	65	76	58	46	41	41	54
2014	959	12	55	65	66	52	47	32	38	51

We’re dealing with only a dozen votes so there’s definitely a sample size issue, but given that Tom Tango pleads with voters to only vote if they watched the player regularly, I feel okay about using it as a proxy for what the informed fan believes on average.

Either the fans are finally catching on to Fowler’s warts or Fowler happens to be declining steeply over the last two seasons. Neither would be a good indicator of his true talent ability at the moment, but we have one more tool to leverage; Fowler’s base running.

There’s more to running the bases than good speed, but speed is very helpful. Reaction time, decision making, and judging the ball also matter a great deal. In other words, a similar set of skills makes one a good outfielder (minus the arm) as makes one a good base runner. It’s not exactly the same, but it’s the same family of abilities.

Fowler remains a terrific base runner, posting above average to well above average BsR totals in each of his six full seasons in the majors. He frequently steals 10-20 bases a season, but he gets caught plenty as well. His real strength is going first to third, second to home, etc. When Fowler gets going, he makes good choices and executes quite well on basepaths.

This seems to run counter to the Fan Scouting Report which indicates his speed is declining significantly. Perhaps we can wash away the “instincts” and “reaction time” comparisons by saying it’s a different sort of each on the bases and in field, but the speed variable is hard to ignore. The fans think he’s getting slower but we’re only seeing negative outcomes in the field as he remains as good as ever on the bases.

The logical conclusion of all of this is pretty interesting. Fowler’s defensive metrics are starting to creep into fans’ evaluation of his tools. If Fowler was getting objectively slower, his defensive metrics should be plummeting and his base running should also be suffering. We only observe poor defense statistically, even though the fans are starting to say their eyes show a slower Fowler.

In the earlier FSR, the voters judged Fowler on his raw abilities, as they are asked to do. But over time, they have started to notice that Fowler is getting to fewer balls than they think he should likely because of the huge negatives on his player page. They started to notice flaws, and so they chalked it up to poor speed or instincts because there are only so many ways to communicate the problem on the ballots.

We don’t have comprehensive speed data so there’s only so much we can say for sure, but if we use base running as an instrument variable, we can be pretty confident that Fowler’s athleticism hasn’t cratered. Yet the Fan Scouting Report seems to think he’s really losing something on defense, indicating the fans are starting to catch up with Fowler’s flaws in center in some capacity.

The voters in our FAN projections so far, no doubt sampled from a different population than the FSR, still buy Fowler’s ability, probably because they’re not the kind of people who watched 120 Astros games in 2014 and are therefore a touch less informed about the player.

Fowler is probably a relatively poor defender with relatively good raw tools. It’s common for fans to misjudge players like that because it’s hard to track defensive performance in your brain and remember all of the poor routes, bad jumps, etc while not getting excited about a few great plays that stand out in your mind. That kind of mental accounting is difficult and it’s where DRS and UZR thrive.

But there are two bits of good news, lest you think this is a reason for Cubs fans to worry. This seems to indicate that Fowler is coachable and could improve his performance with better positioning, which is something his new team could offer. He has the raw ability to be an average center fielder perhaps, but he needs someone to help him reach that potential. You can’t teach speed, but maybe you can teach pre-pitch preparation.

Second, Fowler remains a league average player despite his flaws in the outfield. We’re talking about a 1.5 to 2.5 win player for most of his career. A below average glove in center with a slightly above average bat isn’t a superstar, but it’s a guy you’d like on your roster.

Fowler is an important reminder that being a good athlete doesn’t guarantee that you’ll save a lot of runs in the outfield. He has the ability to cover ground and make leaping grabs, but putting yourself in position to make plays is a vital piece of the overall puzzle and that’s much harder to observe. It’s where the defensive metrics’ blindness to positioning can beat us.

If you aren’t consistently turning batted balls into outs, you’re not playing quality defense. It might not be a skill problem, but it’s a problem. And with respect to Fowler, it’s one that is getting difficult for eye-test evaluators to explain away.