I’m Over Dexter Fowler: Part 2 of 3

In Part 1 of this three-part series (did you read Part 1? What are you doing? Go read it. Now. I’ll wait.), I tried to convince myself that Dexter Fowler may not have stayed on the field enough to be a valuable player had he come to the Blue Jays. Assuming he could stay on the field, let’s see how he might have done in right field with the help of some retrospective projections. Caveat: this article is biased so that I can overcome missing out on Fowler.

Part 2: Fielding

A counterpoint to Part 1’s argument is that Dexter Fowler would probably have significantly boosted the Jays’ defense due to the fact that his fielding is better than that of the man he would replace in right field. Let’s be blunt about Jose Bautista’s outfield defense: it wasn’t always bad, but it’s bad now and it’s getting worse. This can be made fairly obvious using Fangraphs’ defensive runs saved (DRS) metric:

DRS 2012 2013 2014 2015 2016 Avg 5 Avg 3
Bautista 6 6 (1) (3) (8) 0 (4)

Alternatively, let’s look at Fowler’s DRS:

DRS 2012 2013 2014 2015 2016 Avg 5 Avg 3
Fowler (12) (4) (20) (12) 1 (9.4) (10.3)

Granted, Fowler played entirely in center field during this time, whereas Jose played mostly right field, except for 54 innings in 2014 during which he played center. One would reasonably assume that in moving from center field to corner outfield, a player should be able to put up better defensive figures, so we could further assume that Fowler would have been a serious upgrade over Bautista in right field. The question is: how serious?

In order to get an idea of how much improvement occurs when switching positions, I took a sample of 40 players (active in the last 10 years) who played at least approximately 500 innings in both center field and in a corner outfield position (I say approximately because, for example, Kevin Pillar played 497.2 innings in left field and my 500 innings minimum is pretty arbitrary anyway). It’s just to make sure that players played enough innings at that position to produce a big enough historical sample in defensive stats. 500 innings is about 55 games, and while I could have reduced the innings limit to around 270 innings (30 games), this would have likely produced a bigger sample and created more work. I don’t like more work. But I digress.

The results of this 40-player sample, for change in DRS/inning, were as follows:

DRS/Inning CF to LF CF to RF
Median 0.00633 0.00336
Average 0.00312 0.00357
Max 0.01536 0.01544
Min (0.02898) (0.01263)
StDev 0.01053 0.00861

Conclusion: on average, most players in this sample group perform better defensively (in terms of DRS) in corner outfield positions compared to center field. The average figures for transitioning from center to either right or left field are also pretty close, with a similar amount of variation.

If we therefore apply the 0.00357 average increase in DRS/inning (CF to RF) to Dexter Fowler (adding them to his historical DRS/inning and multiplying by his innings played in each of the respective years to produce retrospective projections), we get the following results:

Fowler 2012 2013 2014 2015 2016 Avg 5 Avg 3
Inn (CF) 1,026.0 921.1 959.0 1,324.1 1,027.1
DRS (12) (4) (20) (12) 1 (9.4) (10.3)
Avg Proj DRS (RF) (8.3) (0.7) (16.6) (7.3) 4.7 (5.6) (6.4)
Med Proj DRS (RF) (8.6) (0.9) (16.8) (7.5) 4.5 (5.9) (6.6)
Max Proj DRS (RF) 3.8 10.1 (5.3) 8.3 16.8 6.8 6.6

And what did Bautista’s numbers look like again?

Bautista 2012 2013 2014 2015 2016 Avg 5 Avg 3
DRS 6 6 (1) (3) (8) 0 (4)

Not far off Fowler’s projections. In fact, in order to match Bautista’s 0 average DRS over the last five seasons, and -4 average DRS over the last three seasons, Fowler would only need to have boosted his DRS/inning by 0.00895 and 0.00574 respectively. Given the maximum difference in DRS/inning for center field to right field in the 40-player sample (min ~500 innings) was 0.01544 (Shane Victorino), an increase of 0.0089 DRS/inning is probably not unattainable for Fowler. But boosts in DRS/inning of 0.00894 and 0.00574 over the last five and three years would represent respective increases of approximately 1.5 times and 0.61 times the average increase of the 40-player sample.

Some players who produced differences between right field and center field in the range of +0.00894 DRS/inning in the 40-player sample are Colby Rasmus (0.00873) and Nick Swisher (0.00849). So is Fowler a better fielder than these guys? Here are their DRS over their most recent five years in center field:

Rasmus 2012 2013 2014 2015 2016 Avg 5 Avg 3
DRS (CF) 7 12 (6) 0 3 3.2 (1)
Swisher 2004 2006 2007 2008 2010 Avg 5 Avg 3
DRS (CF) 0 1 1 (11) 0 (1.8) (3.3)

And their most recent seasons in right field:

Rasmus 2009 2015 2016 Avg 3
DRS (RF) (1) 3 3 1.7
Swisher 2011 2012 2013 2014 2015 Avg 5 Avg 3
DRS (RF) (2) 2 1 0 (1) 0 1

Interestingly, Rasmus and Swisher were both superior center fielders to Fowler based on DRS, so these might not be bad comps. But for argument’s sake, here’s the list of players in the sample who produced differences greater than 0.00894 DRS/inning between right field and center field (in descending order of DRS/inning increase, after Victorino, ~500 innings):

These players can be further divided into two groups:

  • Victorino (0.01544), Betts (0.01532), Span (0.01445)
  • Saunders (0.01314), Choo (0.01095)

The difference? The first group produced positive DRS in center field and even higher DRS in right field (not to mention the three greatest differences), whereas the second group produced negative DRS in center field but somehow still managed to increase their DRS/inning by more than our target increase of 0.00894.

Other players who would likely be considered better outfielders than the above second group produced more modest differences in DRS/inning between center field and corner outfield (let’s call these guys the third group):

  • Ender Inciarte and Nyjer Morgan at 0.0075; both with positive DRS in center field to even greater DRS in right field
  • Michael Brantley at 0.0074; very negative DRS in center field to very positive DRS in left field
  • Marlon Byrd at 0.0073 and Ichiro Suzuki at 0.0069; both with slightly negative DRS in center field to fairly positive DRS in right field

Overall, it’s clear that results vary, especially within a 40-player sample, so any number of outcomes for Fowler moving to right field is possible. However, in a 40-player sample (and a lot of these guys are no slouches), only six players (15%) produced differences of greater than 0.00894 DRS/inning between right and center field, and only two of those six players did so with generally negative DRS/inning in center field; so while it may be possible that Fowler could achieve the same level of increase as perhaps Saunders or even Choo, I won’t be easily convinced that he’d be as proficient as someone like Inciarte (despite what his numbers say), let alone Betts.

Consequently, while there are several other defensive metrics that could be used to project Dexter Fowler’s production as a right fielder, and while DRS is just a small part of rating a player’s overall defensive ability, I’ve talked myself into believing that he would have been a “meh” right fielder had he come to Toronto.

Let’s discuss offense in Part 3.

P.S. Fun stuff: These recognizable players rank among the lowest in the sample in terms of difference in DRS/inning between center field to corner outfield, meaning they were actually more productive in center field:

DRS/Inning CF to LF CF to RF
Vernon Wells (0.0009) (0.0126)
Bryce Harper (0.0104) (0.0125)
Carlos Beltran (0.0084)
Kevin Kiermaier (0.0039)
Torii Hunter (0.0028)

 

Adam

Adam likes baseball and uses Instagram mainly for the Joe Biden memes.