A recent look at leading indicators for the next few months
Baseball has just passed the half way mark of the season. And that makes us wonder what is in store for the second half. Players’ performance on the season can improve, decline, or stay about the same. And inquiring minds want some help in guessing at the answer.
Two types of stats can assist with answering the question about the next few months. The first group of stats are “x” (expected) stats calculated at Statcast, which can identify some hitters and pitchers whose actual stats reflect a good deal of good or bad luck. We utilize the Statcast “x” stats frequently. A second group of stats are “z” stats which the ZIPS model generates for its projections. Dan Symborski at Fangraphs says the z stats incorporate some factors similar to Statcast’s “expcted” stats, but also include variables which he finds are predictive.
Z-stats
Symborski, who developed the ZIPS projections, wrote articles for Fangraphs on hitter and pitcher over/under performance based on z-stats through mid-season, which he says are leading indicators of future performance. Basically the z-stats are based on indicators which are less susceptible to statistical noise and thus more predictive in smallish samples than the actual outcomes themselves.
For example, a walk or strike out is the culmination of a number of events during an at-bat—taking strikes and balls, fouling off pitches, and swinging and missing—and the underlying plate discipline stats like “first pitch strike” and various measures of swing% and contact%, can be a better predictor as to whether the existing K or BB rate will improve or regress. That is the concept of the z-stats. The z-stats in theory are more predictive of future performance than previous incurred events during the season.
The results cnn be fairly reliable at predicting the future direction of a player’s performance. Looking back at 2024, z stats predicted the direction of future OPS for 12 of 14 over achievers and 15 of 16 under achievers. The numbers were even better for z-stats predicting the direction of pitcher FIP.
With the explanation out of the way, let’s look at how this relates to Astros’ hitters and pitchers.
Christian Walker
Walker, the free agent signee at 1b, has experienced a very disappointing first half of the season. However, he is among the top 20 of z-stat performers expected to experience a rebound in OPS performance during the second half. Walker’s z-OPS is .729 vs. actual OPS of .635. (Note that these numbers are based on a 6/29 end date; Walker’s OPS has increased 10 points since then.) These findings are similar to the direction of Walker’s OPS indicated by Statcast x-stats. The z-stats are also projecting that both Isaac Paredes and Walker should see a regression toward a lower K rate.
Jeremy Pena
On the flip side, Pena is among the highest over performers in OPS. Pena has a .746 z-OPS versus an actual OPS of .867. So, his OPS may be likely to regress somewhat during the second half of the season. In addition, the z-stats show Pena over performing his K rate, meaning that he may increase his K rate in the future.
Hunter Brown
Hunter Brown is on the top 20 list for FIP over performers. Brown’s z-FIP is 3.44 vs. an actual FIP of 2.69. This would seem indicate some regression toward higher run scoring performance. The article notes that the z-stats still view Brown as an “excellent contributor,” but “not at the same extent” as his actual numbers. The z-stats indicate that Brown’s walk rate is likely to increase in the second half.
Framber Valdez
Framber Valdez is on the top 20 list for home run over performance. The z-stat indicates that Valdez should have 3 more HRs than the 6 HR actual number.
Notable x-stats from Statcast
Moving on from the z-stats, we can also examine some notable differences between x-stats and actual stats for Astros’ hitters.
Under Performing Hitters
[xwOBA minus wOBA] , x-BA, x-SLG
Walker [.26], .238, .418
Y. Diaz (.42), .272, .467
Caratini [.13], .252, .424
Walker, Diaz, and Caratini are in line for improvement in the second half, according to the “expected stats.”
On the other side, the x-stats indicate that Jose Altuve, Pena, and Paredes may be subject to significant regression in the second half. The difference between Pena’s wOBA and xwOBA is -27 points. The difference between wOBA and xwOBA for Altuve and Paredes is -74 and -29 points, respectively.
Hopefully regression by under performing hitters will exceed the impact of any regression by over performing hitters.