On Tuesday, I looked at catch probability to evaluate the St. Louis Cardinals outfield. Using CPR+, I determined that Statcast was significantly lower on the Cardinals outfield than DRS and UZR.
At that time, however, I made no adjustments for positioning. For example, while Piscotty is 11% below average (89 CPR+) compared to all other outfielders, how is he when compared to only right fielders? Today, I set out to answer that question, as well as to quantify an expectation for the 2017 outfield.
Currently, as far as I can tell, the Statcast catch probability leaderboard does not include a positional filter. Therefore, to make a positional adjustment, I assigned an outfield position to players based on where they played the most outfield innings. I lowered the threshholds to 25 opportunities and 100 innings at a position to increase the player sample. I then averaged CPR by position, which you can see below.
With that information, I scaled CPR+ so that each positional average equals 100. Every point above 100 is 1% better than average, and every point below is 1% worse than average (again, think wRC+). I then applied the metric to the St. Louis Cardinals projected outfield for the upcoming season.
As expected, the move to left field should benefit Randal Grichuk defensively. In center field, Grichuk was a little below average, with a CPR+ of 95. In left field, he would have been a little above average.
As most everyone knows, the St. Louis Cardinals had a defensive weak spot in left field, courtesy of Matt Holliday. He played nearly half of the team’s LF innings and had a CPR+ of 75. Overall, the team had a left field CPR+ of 85, about 15% below average. Moving Grichuk to left results in a 21% improvement.
Unfortunately, the improvement in left field is mostly offset by Dexter Fowler, who is a much worse center fielder than Grichuk, at least per CPR+. Fowler’s CPR+ of 81 is a 14% decrease from Grichuk’s 2016 production in center. Additionally, Fowler has a worse arm than Grichuk as measured by both DRS (rARM) and UZR (ARM), which CPR+ does not account for.
In right, Stephen Piscotty was about average for the position in 2016 as measured by CPR+, which is basically in line with his ranking by DRS/1000 and UZR/150.
As shown in my article from Tuesday, CPR+ is fairly stable from year-to-year, with an r^2 value of approximately .65. Thus, it is fair to expect Grichuk, Piscotty, and Fowler to produce a CPR+ in 2017 near their 2016 levels.
So what does that mean for the overall Cardinals outfield? The easiest way to compare is to use 2016 batted balls to see how much better the new outfield would have been last season over the team’s actual 2016 outfield using a few simplifying assumptions.
Last year, weighting the Cardinals approximate positional CPR+ by the percentage of outfield batted balls to each position, we get an overall CPR+ of 93 for 2016. That number includes the likes of Matt Holliday and Brandon Moss.
Next, I moved Grichuk over to left and inserted Fowler into center. I then applied the same method, and calculated an overall CPR+ of 94. As I mentioned previously, it looked like the benefit from moving Grichuk to left was almost entirely offset by inserting Fowler into centerfield, and it appears this is exactly the case.
This is bad news for the Cardinals, who are banking on the realignment to improve the overall defense in 2017. Further, since catch probability does not buy into Fowler’s defensive renaissance, that five year contract looks more dubious. It’s likely that, as Fowler ages, his defense will get worse. That might necessitate a move to left field, but Fowler already would have been a below average left fielder last season; his CPR+ in LF would have been only 95.
The St. Louis Cardinals claimed to have emphasized athleticism and defense this offseason. Their success in doing so has been questioned, and catch probability doesn’t show that they were successful.