With the plethora of outfielders on the New York Mets roster it will be hard to determine who should play where and how much they should play. While Yoenis Cespedes will get the bulk of the time in left field, playing time in center field and right field will likely be shared among Jay Bruce, Michael Conforto, Curtis Granderson, and Juan Lagares. Brandon Nimmo is also an option but given his flexibility and youth he will probably spend most of the season in Las Vegas. (Tim Tebow is not considered despite his appearances in the outfield this spring as he will hopefully never play in Queens.)
With so many options, it could become very difficult to decide the share of playing time in the outfield. Luckily, there is more data than ever to help the Mets make smart decisions. Also, the public has more information thanks to the release of Catch Probability metrics from MLBAM’s Statcast on March 12. Historically, some of this data has been used in video replays showing what the probability percentage of an outfielder making a catch is; however, it has never been released to the public in this form. The exact definition of Catch Probability from MLB.com is as follows:
“Catch Probability represents the likelihood that a batted ball to the outfield will be caught. With Statcast tracking the exact start position on the field for each fielder and also measuring the hang time of each batted ball, the two most important pieces of data to define the difficulty of a catch opportunity are: 1. How far did the fielder have to go? 2. How much time did he have to get there?”
In summary, it is the likelihood of the average defender is making that same exact catch based on how far the fielder would have to travel and how long the air stayed in the air. The way Statcast Catch Probability data was released to the public is that they broke the data into five different categories by probability of the catch being made.
- 5 Star: 0 – 25%
- 4 Star: 26 – 50%
- 3 Star: 51 – 75%
- 2 Star: 76 – 90%
- 1 Star: 91 – 95%
The current defensive metrics are under a lot of scrutiny for being tough to understand and not representing true talent levels … however, the probability of a ball being caught is simple enough statistic that hopefully analytical and casual fans can get behind it.
The current defensive metrics and most fans can agree that the Mets in 2017 are expected to start an outfield of below-average defensive players; Yoenis Cespedes (with the exception of his arm), Curtis Granderson, and Jay Bruce are all poor defenders. Potential backups Michael Conforto and Juan Lagares, are considered to be better defenders, especially Lagares who is considered to be one of the league’s best. My question is this: do these new Catch Probability stats confirm what we think we know or raise new doubts.
The data available from Statcast is available for each individual player who spent time in the outfield during 2015 or 2016. To see how the Mets performed compared to the rest of the league, I combined data for both seasons for each player as well as for all of MLB to get a league average. The data is summarized below, red represents worse than league average and green represents better than league average in each of the five categories.
| Player | 5 Star (0-25%) | 4 Star (26-50%) | 3 Star (51-75%) | 2 Star (76-90%) | 1 Star (91-95%) |
| Jay Bruce | 7% | 45% | 53% | 89% | 88% |
| Yoenis Cespedes | 1% | 39% | 66% | 74% | 98% |
| Michael Conforto | 20% | 36% | 81% | 82% | 100% |
| Curtis Granderson | 6% | 35% | 57% | 82% | 93% |
| Juan Lagares | 20% | 83% | 95% | 95% | 100% |
| League Average | 8% | 40% | 67% | 83% | 93% |
1 Star (91–95%)
The one star category is those basic can of corn catches which have been made approximately 93 percent of the time over the past two seasons. All Mets are either in or above the expected range of 91 to 95 percent, except Jay Bruce who only makes the simplest catches 88 percent of the time. Given that there are other outfielders vying for the position it seems reasonable to not want to use a player that isn’t making the easiest catches. The impact of Jay Bruce’s poor defense based on his opportunities the past two seasons is two additional hits for the opposition over the course of the season.
2 Star (76–90%)
In the two star category is where we start to see the New York Mets lag behind the rest of the league. The league average for Catch Probability is 83 percent and three players out of the five players are lower than that. Both Conforto and Granderson are lower but only by a small amount as they are both at 82 percent and still within the expected range of 76 to 90 percent. Cespedes with a catch percentage of 74 percent is noticeably underperforming the rest of the league in what should still be relatively simple catches. What is extremely odd about the data is that Jay Bruce is above average in this category and near the upper echelon at 89 percent. This is strange considering based on the data he is worse at making one star catches which should be easier than the two star catches. Perhaps this is due to a small sample issue?
3 Star (51–75%)
These catches are on average made one half to three-quarters of the time. Juan Lagares has only had 20 of these opportunities over the past two seasons but is a clear outlier in the data converting 95 percent of these opportunities into outs. Similarly Michael Conforto in his 21 opportunities has shown success (81 percent, above the expected range of 51 to 75 percent). Cespedes, Granderson, and Bruce are all still within the expected range, however they are all below the league average especially Bruce who is 53 percent, close to the bottom boundary of 51 percent.
4 Star (26–50%)
These catches are more difficult and only made between one-quarter and one-half of the time. Here Juan Lagares once again is extraordinarily above league average at 83 percent although this is only 12 opportunities over two years which is an extremely small sample. Most interesting about this is that Jay Bruce who struggles in the 3 Star category is now above average and towards the top of this range. As the data is currently only at an aggregate level it is not known if he is just genuinely better at making these more difficult catches than the rest of the league either due to skill or luck or if the data is biased somehow. For example, if his catches generally tend towards the 50 percent catch probability then it makes sense that he would be towards the top. On the other side, Conforto struggles in this range despite doing very well in the 50 to 75 percent range. He had only 14 opportunities though so it is still a small sample size but the flip-flop in this category is a little strange. Cespedes and Granderson still struggle compared to the rest of the league.
5 Star (0–25%)
These are the most difficult catches to make and are made less than one-quarter of the time. Juan Lagares and Michael Conforto are both are only at 20 percent, but that is still above average in this category. (Lagares had 30 opportunities and Conforto had 25 opportunities). Both Jay Bruce and Curtis Granderson are performing below average in this category making catches less than eight percent of the time. However the worst F for the Mets is Yoenis Cespedes. He has made only one catch in his 71 chances in these difficult-to-catch opportunities. Cespedes is actually second-worst in the league to only Matt Kemp who has zero catches in his 72 opportunities.
Conclusion
The defensive data released by Statcast confirms what most New York Mets fans thought: Juan Lagares is the best defensive outfielder on the team followed by Michael Conforto. However the main issue for the team is that right now Curtis Granderson and Jay Bruce are likely starting over Lagares and Conforto. Lagares has struggled to hit the past two years which makes it understandable that he is not likely to start; the bigger enigma is what to do with Michael Conforto. He is still very young and if given the opportunity he should be able to hit at the big league level. Although he put up only a .310 OBP in the major last year, he dominated at Las Vegas with a .483 OBP and a .727 slugging Percentage. This means he does not have any more room to grow in Triple-A and must be given the opportunity to play in the Major Leagues. Hopefully the New York Mets understand the defensive metrics, listen to Jarrett, and give Conforto the opportunity to start in 2017. Although Cespedes appears as clearly one of the worst defenders on the team, I will not try to convince anybody that the Mets should not play him as his defensive inefficiency is not worth nearly as much as his offensive value. If I was the Mets I would start Cespdes, Conforto, and a combination of Granderson and Bruce. If the Mets had been able to trade Jay Bruce, than they would not have had to worry about his defensive skills. Unfortunately with the surplus of power hitting outfielders this off-season, they could not find any takers.
Author’s Note
Unfortunately, Statcast defensive data so far is that it is only available at this limited aggregate level as of right now. Once more granular data is available these data points can be further weighted and analyzed properly. Darren Willman, Director of Baseball Research & Development for MLB, tweeted on March 19 about hoping to make the charts at least more dynamic. This granularity and customization of the data will further allow us to make better comparisons of the players and their defensive metrics.
Photo Credit: Jasen Vinlove-USA TODAY Sports