Astros Saber Notes: Harrell, Dominguez, Carter

Sabermetrics encourages the improvising of new metrics and statistical applications---frequently aimed at answering specific narrow questions. From a pure curiosity standpoint, this can be more fascinating to me than the popular general advanced metrics.. Bill James used to call his ad hoc measures "toys" ---which he meant as term of affection.

In today's Talking Sabermetrics column, I'll look at some new applications which have relevance to Astros' pitcher Lucas Harrell, and Astros' hitters Matt Dominguez and Chris Carter. All three metrics were developed by Jeff Zimmerman, who writes on health and injury issues for Fangraphs' rotographs section. Fantasy baseball is a fertile ground for new statistical tools, and each of Zimmerman's tools is aimed at helping roto players refine their drafting and trading activities. This also points out the heavy influence of fantasy baseball on sabermetrics research.

Jeff Zimmerman and fellow Fangraphs' writer Bill Petti recently won the SABR Award for Contemporary Analytic Research.

Lucas Harrell

Zimmerman has been working on analytic profiles to detect players who are playing through injury or may be likely to be injured in the future. Unfortunately, Lucas Harrell shows up as one of the top candidates for an injury affected performance. The PAIN (Pitcher Abuse Index) metric uses pitch f/x characteristics to determine how well a pitcher lines up against the profile of injured pitchers. An index value over 100 means that the pitcher is a higher injury risk, based on the changes in pitch f/x performance of injured pitchers in the past. Loss of velocity, a change in the ability to put pitches in the strike zone, inconsistent release points, and a slow down in the player's normal pitching pace are among the factors considered.

Harrell has shown up with the 2d - 4th highest index value for PAIN over the past 40 days or so.Harrell's PAIN index value has been increasing steadily over the last month. Harrell's PAIN ranking reached 2d on July 5th, just before his start against the Rangers, which resulted in a demotion to the bullpen. Although his ranking decreased from 2d to 4th between July 5 and July 8, Harrell's index value increased from 166 to 185. Keep in mind that an index value greater than 100 indicates high injury risk.

Zimmerman wrote about Harrell on June 10. He points to a drop in velocity, a 9% reduction in his zone percent-- including a zone percent well below the 47% threshold associated with injured pitchers-- and a slower pitching pace. He then examines Harrell's release points and wraps it up:

While he tries to use different release points, they are not consistent in any way. Putting it all together, I would not be surprised to see Harrell end up on the D.L. at some point this season.

According to fangraphs' leaderboard, Harrell's current 38.7% zone percentage is third worst among starting ML pitchers. Only Jeff Locke and Jason Marquis have been worse at pitching into the strike zone. Compare that to Harrell's 49.2% zone percentage in 2012 and the 47% risk threshold used by Zimmerman.

I compared Harrell's pitch f/x results for his last start against the Rangers with his average pitch f/x results for all pitches thrown in 2012 and 2013 prior to that game. (Texasleaguers.com.) I show the velocity for types of pitches below:

(Through July 1 / July 5th vs. Rangers)

4 seam FB * 92.3 / 88.1

2 seam FB 92.2 / 91.7

Cutter 90.1 / 90.2

Slider 87.8/ 87.9

Change Up 82.8/ 83.9

Curve 82.4 / 80.5

*Note- Only one 4 seam FB observed on July 5

The reduced velocity (about 1/2 mph) primarily affected Harrell's 2 seam FB, which is his primary pitch. In addition, Harrell wasn't getting as much vertical movement on his 2 seam FB on July 5. Perhaps that contributed to the HRs given up by Harrell in the game.

Is Harrell attempting to conceal that he is pitching through soreness or pain? We don't know, but it is a possibility for almost any professional athlete, particularly a competitive personality like Harrell. However, that would explain some of the inconsistency and performance decline. We also don't know whether his errant mechanics could set him up for injury in the future.

On the other hand, there may be no injury---just severe mechanical issues. Any performance issue which mimics an injured pitcher is a serious problem. Hopefully, Harrell's move to the bullpen will allow the Astros to sort out the cause or causes of Harrell's performance decline.

Matt Dominguez

The overall offensive stats tell us that Matt Dominguez hasn't been hitting well this year. (wRC+ of 71) However, the Roto stat called SLOW may give us one of the details for "Why?"

Zimmerman created SLOW in order to attempt to quantify terms like "slow bat," and "can't catch up to the fastball," which are often thrown around by broadcasters and scouts. SLOW uses pitch f/x data to quantify a hitter's ability to hit 91+ mph fastballs thrown down the middle of the strike zone. For these sub set of pitches, the statistic takes into account contact rate, foul contact percent (whether contact is in play or not), and the pull angle of line drives and grounders (to determine if the swing is late).

Matt Dominguez makes the Top 20 leaderboard of hitters with the biggest change (worsening) in SLOW from 2012 to 2013. In other words, he isn't catching up to the fastball like he did last year. At first, I thought this was surprising, since Dominguez showed the capabilty to turn around a 100+ mph pitch from Aroldis Chapman last year. But, then again, maybe that's why his offensive statistics have fallen off this year---he isn't repeating the good fastball hitting of 2012.

An examination of Dominguez's pitch value result on his fangraphs page supports the SLOW finding. Dominguez was +2.6 hitting fastballs in 2012 and he is -1.6 this year. He has similar declines in pitch values for 2 seam fastballs and cut fastballs. However, he has positive values hitting curveballs and change ups, which is more similar to last year.

We don't know if this is due to a change in his batting mechanics or if he is having problems "guessing" what the pitcher will throw. But it seems like Matty D and his batting coach need to figure it out.

Chris Carter

Based on some recent discussion and comments on the Lowrie trade, Chris Carter appears to be a polarizing player. However, even his critics admit that Carter has a lot of raw power. And we have another indicator of that fact.

Jeff Zimmerman has created an interesting tool, which you can access here, that calculates the average distance on batted balls for any major league hitter you request. This tool uses pitch f/x data and some physics. But it's a neat application. Not surprisingly, a hitter's average distance on fly balls is highly correlated with his HR / flyball rate. In addition, this data can be used to test whether a player's current HR total is sustainable. It may also be possible to predict whether a majorl change in a hitter's seasonal flyball/HR rate is likely to continue or likely to be an aberration.

At 303 feet, Chris Carter is No. 13 on the ML leaderboard for "flyball plus HR" distance. That puts him in the company of power hitters like Miguel Cabrera, Adam Dunn, Chris Davis, Bryce Harper, and Carlos Gonzalez, whom are among the other hitters with 300+ foot average flyballs. .J.D. Martinez and Carlos Pena are the closest Astros to Carter in distance, at No. 47 (293 feet) and No. 60 (292 feet), respectively.

Chris Carter's high BABIP on line drives (.811) led me to wonder whether the average distance on batted balls could be incorporated into the x-BABIP calculation in order to improve the accuracy of BABIP predictions. (Chris Davis and Chris Carter have the highest wRC+ on line drives, by the way.) Distance on batted balls is a potential measure of how hard a batter hits the ball, which in turn would affect BABIP for line drives. Seven of the 10 top distance batters also have line drive BABIP substantially above average (.695). However, a more thorough analysis would be required to demonstrate the relationship.

Want additional evidence that Carter has premium power? Among ML first basemen, Carter has the sixth highest isolated power (ISO). Carter's high strike out rate probably is the most controversial issue for his critics. Yet, a look at the high ISO first baseman shows that high strike out rates are part of the power hitter package. The top nine ISO hitters at first base have an average strike out rate of 25%. The average strike out rate for the remaining 21 qualified third basemen is 20%.

Do you have any thoughts on Harrell's pitching struggles, Dominguez's fastball hitting, or Carter's power?