How To Evaluate Mets Prospects in Triple-A Las Vegas

dilly1

It is hard to evaluate Mets position player prospects in the hitter’s heaven known as the Triple-A Pacific Coast League. Offense in the PCL is off the charts: the Mets affiliate, the Las Vegas 51’s, is hitting .288/.363/.448 as a team right now. Mets pitchers have an aggregate ERA of 5.25, but that’s been good enough to lead to a winning record on the year so far (31-27).

So how do we know which Mets hitters are good wilmer.JPGenough to be promoted to the major leagues and which ones are duds? We all already know we can’t look at the PCL statistics and expect anything close to that in the major leagues, but we also know that some players make the transition better than others. So, how can you tell the 2014 Wilmer Flores (.323/.367/.568) from 2014 Eric Campbell (.355/.442/.525)? Power.

*     *     *

This column isn’t going to provide a perfect answer, but it’s a fine place to start. At the very least, we can begin to narrow down the things that don’t matter, and hopefully work toward identifying what does matter, and how much.

Unsurprisingly, my quick-and-dirty review found that batting average, slugging percentage, OPS, and other measures of hitting skill correlated to success in the major leagues. What was most interesting, however, is the other finding: that players’ strikeout and walk rate in the PCL had no correlation to major league success.  Intuitively this makes a good deal of sense, and we’ll get into more detail below.

To do this project, I took a random sampling of players who had 300 or more plate appearances in Las Vegas in the years 2002, 2007, 2011 and 2014 (these are years in which a decent number of players went on to the major leagues). From that sample, I threw out any player that was older than 30. This left me with 22 player-seasons, from Joe Thurston in 2002 to Wilson Valdez in 2007 to Josh Satin in 2014. A small-ish sample, but a decent start.

I then collected the career major league games played, wRC+, walk rates and strikeout rates for each of the players who made it to the major leagues.

*     *     *

Overall, the group performed pretty terribly in the major leagues. As a group, they averaged a 25% decline in walks and a 20% increase in strikeouts upon reaching the major leagues, and that strikeout increase percentage increase upon making it to the majors is actually 35% for the most recent group I looked at (Andrew Brown, Josh Satin, Allan Dykstra, Daniel Muno, Matt den Dekker, Brandon Allen, Eric Campbell) in their short samples.

mlb-pcl-il-ops.png

These findings are in line with the work done by Dan Szymborski and Grantland who determined that only “about 80 percent of a Triple-A hitter’s production survives his promotion to the majors.” (Grantland).

Next, I took a look to see at which Triple-A statistics, if any, correlated to success in the major leagues among this group. Of course, the usual caveats apply — the sample is limited, it is going to skew depending on how you weight the inputs (e.g. is each player equal, whether they’ve had 100 games or 500). But we can begin to narrow down what might matter.

*     *     *

First up was strikeout and walk rates, which as we know are incredibly important in the major leagues. However, given the offensive environment that is the Pacific Coast League, I had a hunch that they weren’t quite as important to major league success. After all, you might be able to hunt for a walk in the PCL, but pitchers can still knock the bat out of your hands in the Show.

These are sorted from left-to-right from the best-to-worst, highest walk rates and lowest strikeout rates on the left.

wrc bb.JPG

wrc k.JPG

As it turns out, strikeout and walk rates weren’t just unimportant – they had no discernible meaning at all for this sample of players. We won’t be able to tell who the best major leaguers will be based on their control of the strike zone in the PCL. In fact, strikeout-to-walk ratio actually had a slightly negative correlation to major league success in this sample.

wrc bbk.JPG

*     *     *

I had long had a theory that batting average in Triple-A Las Vegas hardly mattered due to the thin, dry air which flattened breaking balls and allowed hitters to hit home runs with relative ease. It was hard not to think that, having seen so many players with high averages come to the majors and flop year after year, from Josh Satin to Eric Campbell to Matt den Dekker and so on. I thought there had to be something else, and I honed in on Isolated Power % (which is slugging percentage minus batting average) as a way to try and demonstrate the quality of contact.

I was half-right and half-wrong, because as it turns out, both batting average and isolated power had about equal correlations to success in the major leagues.

wrc bawrc iso

That one point on the right which is ruining my ISO% graph is Josh Satin’s career 103 wRC+ in 251 major league platoon at-bats, illustrating the peril of drawing too strong a conclusion from this study. But nonetheless, there appears to be some meaning here.

On-base percentage has a positive correlation, but it’s less important than batting average (which makes sense, since we determined above that walks have no predictive value). So with batting average and ISO% both appearing to have strong correlations, that leads us to the following conclusion: slugging percentage ought to be the most meaningful.

And it is:

wrc slg no satin.JPG

If you don’t slug in Vegas, you should stay in Vegas.

(That missing point is Josh Satin, who ruins the visual. However even with Satin included, the correlation is still the strongest)

*     *     *

So who currently in Las Vegas deserves a call-up given this criteria? We are looking for a high batting average and a high slugging percentage, but someone who can still put up an isolated power that is on the higher end of the spectrum. We know from our study of statistics that batting average is one of the last things to “stabilize” and become meaningful, so if there is something that might be a fluke, it would be batting average. (Fangraphs Library).

First on the list is Ty Kelly, who hit .391/.478/.548 before being called up to the majors. However, Kelly is also a career .274 hitter in Triple-A, with most of those seasons coming in the PCL (not the comparatively low-offense International League). Kelly is simply not that good of a hitter, with a current BABIP of .457 on the year. To boot, his isolated power % is only .157, which would put him 18th out of the 23 players we looked at. He’s likely to have a rough transition to the majors.

Next up is TJ Rivera, who is slashing a respectable .362/.400/.529 in Las Vegas, good for a 144 wRC+. His BABIP is a high but not-impossible .389, and he is a career .323 hitter in the minor leagues. His isolated power leaves something to be desired, at .167, but the fact of the matter is that he’s potentially a true talent .330+ hitter in the Pacific Coast League. He doesn’t have the secondary skills necessary to become a star in the major leagues, but I would not be surprised if he met or exceeded Steamer’s optimistic projection for him of .268/.302/.358 (85 wRC+) if he were called up.

An interesting test case for the information above would be Travis Taijeron, who is slugging the ball in Triple-A to the tune of .285/.391/.545 (.260 ISO and 144 wRC+) in a season and a half. Taijeron has the slugging and the ISO (he leads the team in both) but his average is lacking despite a .400 BABIP. A 30% strikeout rate will do that to you, and he doesn’t appear to be good enough of a pure hitter to make up for that. In either case, it doesn’t seem like the Mets consider him a viable call-up option.

Three more players worth looking at: Eric Campbell, Dilson Herrera, and Brandon Nimmo.

Campbell, as we know well,  is a career .333/.447/.510 hitter in Las Vegas, with his lowest wRC+ there checking in at 148. Some of what we’ve learned here helps to explain him — for instance, his remarkable strikeout-to-walk ratio doesn’t matter. His batting average is nice, but his isolated power and slugging figures are both lackluster. His ISO% is squarely middle of the pack. He’s probably better than we’ve seen in the major leagues so far, but 17 home runs in over 750+ plate appearances in Las Vegas in your prime (ages 26-29) means there isn’t much there. Flores had 28 home runs there in 704 plate appearances at age 21 and 22.

dilly2.JPGFor Herrera and Nimmo, 22 and 23 respectively, there is hope. Herrera has already cranked out 21 home runs in Las Vegas in 597 plate appearances, slashing .314/.364/.513. He’s hit the ball even harder this year, with a .221 ISO%, but has been held back by an uncustomarily low .313 BABIP in the early-going. Herrera hit .292/.344/.558 in May and has hit .387/.424/.677(!) in June. Herrera is clearly on his way to being a regular everyday player in the major leagues. (Photo credit: Michael Baron)

Nimmo is somewhat of a tweener right now. He’s finally shown some life this year after several years of lackluster performance and high walk rates. He’s slashing .325/.403/.521 with a .379 BABIP after last night which, according to our normal rules, would make him a moderately good prospect thanks to that fringe-good slugging and ISO% (thank you @tpgmets for the update). However, Nimmo has truly come on as of late with a .483/.515/.897(!) line in June, showing actual power to go along with the high average and plate discipline. If Nimmo is over his nagging injuries and has turned a corner power-wise, this would be huge for the Mets (Nimmo also had a .229 ISO% in May).

*     *    *

So, in summary, the Mets called up the one guy who had no real chance of success in the major leagues in Ty Kelly, while the equivalent TJ Rivera continues to work away in anonymity. Kelly, Rivera, and Campbell are pretty much interchangeable parts.

Herrera, on the other hand, has little left to prove in Triple-A. (Fangraphs). If the Mets were willing to move Neil Walker to another position, I’m confident that we’d have already seen Herrera called up. With their offense flagging, I hope they find a way to use him sooner or later.

*     *    *

This is merely an initial attempt to find meaning for the statistics of Mets position players in Triple-A Las Vegas. It is by no means comprehensive, and I hope and expect that someone might be willing to continue down this path and research it more thoroughly.

It goes without saying that hitters need to hit, but it appears that is even more true in an offensive launching pad like Las Vegas. What doesn’t slug in Vegas, stays in Vegas.