clock menu more-arrow no yes

Filed under:

Yankees’ players are far exceeding their mean PECOTA projections

New, comment

There are ten Bombers that are at or above their 90th percentile projections.

Baltimore Orioles v New York Yankees Photo by Jim McIsaac/Getty Images

Baseball can always be relied upon to deliver surprises. You can’t have a sport this random and not have that be true. It seems this season there has been an exceptional amount of surprises, and the Yankees have been right up there with the best of them.

The Yankees were expected to be a solid if unremarkable team. PECOTA had them as an 82-win team, and FanGraphs had them a bit below .500. Now PECOTA projects an 89-win season for the Yankees, and FanGraphs is up to 91 wins. A team that was seen as needing some luck to get a Wild Card might actually win the division. Obviously Aaron Judge is a big factor in this, but one player can’t carry a team, so it is a good thing that so many other Yankees’ players have joined him playing out of their minds.

One of the things that I really like about PECOTA projections (from Baseball Prospectus) is that they do not just give you the 50 percent regression point. Projections are actually a bell curve, and outside of PECOTA we are only given the peak of the curve. PECOTA goes the extra mile and gives us nine points on the bell curve. Here is Judge’s PECOTA card as an example.

Aaron Judge PECOTA

90 432 63 106 19 2 23 70 56 123 3 1 .286 .384 .533 .314 33.8 RF 7 4.5
80 412 58 97 17 2 21 65 51 121 3 1 .273 .368 .509 .302 27 RF 7 3.7
70 398 55 91 16 2 20 61 48 119 3 1 .263 .357 .491 .293 22.5 RF 7 3.2
60 386 52 85 15 1 19 58 45 117 2 1 .256 .348 .476 .285 18.8 RF 7 2.8
50 375 49 81 14 1 18 55 43 115 2 1 .248 .339 .462 .278 15.5 RF 6 2.4
40 364 46 77 14 1 17 52 40 113 2 1 .241 .331 .449 .270 12.3 RF 6 2
30 352 43 72 13 1 16 49 38 111 2 1 .233 .321 .434 .263 9.2 RF 6 1.7
20 338 40 66 12 1 14 46 35 108 2 1 .224 .310 .417 .254 5.7 RF 6 1.3
10 318 36 60 11 1 13 41 31 104 2 1 .212 .295 .394 .241 1.3 RF 5 0.7
Weighted Mean 378 50 82 15 1 18 56 43 115 2 1 .250 .342 .466 .280 16.4 RF 7 2.5
Baseball Prospectus

What these percentiles mean is simple. In the table above, the 20th percentile line lists Judge’s triple slash line at .224/.310/.417. That means that he has an 80 percent chance of performing better than that, and a 20 percent chance of performing worse. For another example, his 90th percentile line lists a slash line of .286/.384/.533. That means he has a 10 percent chance of performing better than that, and a 90 percent chance of performing worse. You can probably see where I am going with this.

Judge is currently laughing at the other puny mortals who play baseball by hitting .332/.439/.678 with a league-leading 19 HR. It can be difficult to compare players by slash lines -- that is what wOBA is for — but that is clearly far beyond what his 90th percentile projections had for him. That is probably above even his 95th percentile projections or more.

For a rookie like Judge, projections are especially difficult because there is no major league track record, and obviously it is too soon to make historical comparisons. Luis Severino and Gary Sánchez also fall into this category. Other players on this veteran team, however, do not.

I went through some of the most surprising performances from both position players and pitchers, and then compared them to their 90th percentile wOBA. Since Baseball Prospectus does not use wOBA, I had to calculate the projected wOBA myself using this wOBA calculator. I set HBP, IBB, and ROE at zero.

Yankees position player PECOTA — 90th percentiles

2017 wOBA Projected wOBA 2017 BABIP 2017 HR/FB%
2017 wOBA Projected wOBA 2017 BABIP 2017 HR/FB%
Aaron Judge .460 .388 .422 38.8
Aaron Hicks .423 .369 .333 20.0
Brett Gardner .364 .361 .294 22.0
Starlin Castro .374 .356 .361 16.4
Didi Gregorius .357 .354 .347 11.1
Matt Holliday .389 .382 .347 24.5
Gary Sánchez .377 .387 .318 28.1
Projected wOBA comes from 90th percentile PECOTA Baseball Prospectus, FanGraphs

That is unreal, and we are not even at the pitchers yet. Sánchez is the only one of the SEVEN hitters above who has not met or exceeded his 90th percentile PECOTA projections so far this season, and he is still pretty close. As one would expect, there are some ridiculous BABIP and HR/FB numbers in that table. Regression is very likely to come.

As for the pitchers, I will just stick to the starters. The Yankees bullpen is doing excellent even beyond Aroldis Chapman and Dellin Betances, but the sample sizes are way too small and reliever performance tends to be especially volatile.

Yankees starting pitcher PECOTA — 90th percentiles

2017 ERA 2017 DRA Projected ERA Projected DRA BABIP HR/FB%
2017 ERA 2017 DRA Projected ERA Projected DRA BABIP HR/FB%
Luis Severino 2.75 2.49 3.11 3.25 .267 16.4
Jordan Montgomery 3.55 3.50 3.45 3.51 .282 7.8
C.C Sabathia 3.66 4.71 3.73 3.91 .280 13.6
Projections are from 90th percentile PECOTA Baseball Prospectus, FanGraphs

Interestingly enough, none of the BABIPs are crazy, and the HR/FB of Severino and C.C. Sabathia are actually high. Sabathia’s DRA is much higher than his runs allowed, but the DRA Run Values list is not clear as to why that is.

Sabathia’s DRA indicates that he is likely to regress, but as for Jordan Montgomery and Severino, it is harder to tell. Baseball Prospectus scouted Montgomery in September, and projected him as a fifth starter. Right now I would argue that he has the results of a number three starter, but it is hard to imagine he has changed so much in less than a year.

As for Severino, I am curious as to whether or not his PECOTA projections assumed that he would be relieving for part of or all of the season. If that is true, his 90th percentile ERA purely as a starter should be higher than what is listed. Despite being only a two-pitch pitcher, I doubt that scouts are too surprised by his season to date. Scouts’ major concern with Severino was never his stuff, but his mechanics. He does not use his lower half in his delivery, which makes him a significant risk to land in the bullpen, though he would be expected to be a relief ace if that were to happen. It is great to see young players succeed, but if he continues to be a successful two-pitch starter who does not use his lower half, he might be the first.

It is entirely possible and even likely that PECOTA is off in measuring the true talent of some of the newer players. In other words, some of the newer players might “only” be exceeding their 60th or 70th percentile true talent. Let’s say for argument’s sake that each of the ten players listed above is meeting or exceeding his 90th percentile projections. Remember that a player has a one in ten chance to accomplish that. The chances that this continues for the entire season for all ten players are about 1 in 10 billion!

So I am going to go out on a limb and say that this Yankees season is not 1 in 10 billion and say that regression is going to come. Even if the Yankees fall off a bit and lose the division, they have LOTS for their fans to be excited about. Aaron Judge alone is enough for fans to go crazy with excitement, and the team is rebuilding itself without ever tanking, or even without being uncompetitive. And don’t forget, one of the best prospects in baseball, Gleyber Torres, might make his debut later this season. It sure is great to see a tough-luck team like the Yankees have everything go right for them (bold, italicized, and underlined sarcasm font).

All stats are current through Saturday, June 10.

Luis Torres is a Featured Writer at Beyond the Box Score. He is a medicinal chemist by day, baseball analyst by night. You can follow him on Twitter at @ChemTorres21.