Pitcher Evaluation Tools by Jeff Zimmerman November 23, 2016 About a week ago, I went over the stats I use when examining hitters during the offseason. Today pitchers take center stage. With Pitchf/x and now Trackman publicly available, I find it quite a bit easier to evaluate pitchers and the changes they make. For pitchers, I have one sparkling new main source and one old standby. Pitchers are so much easier to evaluate compared to hitters. If a pitcher gains a couple ticks on his fastball, we know right away within a couple of pitches. If a hitter can no longer catch up with a 96 mph fastball, it may take a few months to know for sure and even then, we may not be sure why. Here are the tools I use to help find pitchers who have changed for the good or the bad. Pitch Type Metrics Ever since helping Eno dive into pitch-type metrics, I’ve determined that I will evaluate pitchers using this type of data. Even though it took a while to hammer everything out, the final results have been extremely promising. We can now determine what each small pitch change will mean for each pitcher and how pitchers can improve their results. The complete write up of the process is available, but here is a quick summary. The key change is to give each pitch an ERA value (pERA) based on the pitch’s swinging strike and groundball rate. All the values are based off the average values for starting pitcher. Closers will have higher grades because their stuff plays better coming out of the bullpen. The pitcher’s control is determined from their walk rate which is separate from the pitch grades. I’ve put each pitch on the 20-80 scale with 50 being average, 80 great, and 20 horrible. For starters, target pitchers with three average or better pitches. For relievers, they just need two pitches. I have all the data going back to 2010 available for everyone to download. The information seems like a lot to digest, but let me walk through an example using Charlie Morton. Often-injured Morton was just recently signed by the Astros. After I looked at the pitchers available this offseason, I was not surprised one bit to see Morton headed to sabermetric leaning team like the Astros. In just four games last season, Morton was putting up some insane numbers. First of all his fastball was up 2 mph from his previous season high. His two-seam fastball pERA went from grading 4.04 in 2014 and 3.98 in 2015 to 3.00 last season. The increase in velocity was also seen in his curveball (2.15 and 1.58 to -0.75 ) and splitter/change (3.05 and 2.83 to 0.78). The combination of the pitches put his combined pre-control pERA at 1.98 while being at 3.61 in 2014 and 3.44 in 2015. With the increase in velocity, Morton lost some command and saw his walk rate jump to 4.2 after being around 3.0 for the past three seasons. With control taken into account, his overall combined pERA is at 3.05. This value is closer to his FIP (3.09) and xFIP (3.01), but a ways off from his SIERA (3.56) and ERA (4.15). While it was just four games Morton threw in 2016 before getting injured, the changes in his pitches and the results made him an intriguing free agent option and the Astros decided to take the chance on him. Release Point Variability The other data sheet only has one piece of information, but I find it just as important. The value I measure is the amount of deviation in a pitcher’s release point. I use it to determine what pitchers may have good and bad mechanics. I give each pitcher a 20-80 grade where 50 is average, 80 is great, and 20 horrible. First, I am looking for pitchers with grades over 65 (e.g. Hisashi Iwakuma, Jacob deGrom, and Luke Weaver) since they will likely have great command and a higher K/BB rate. From some initial work I have done with looking at pitcher results and a tight release point, I find these pitchers have higher than expected K/BB rates. Some experts will say these pitchers will have just an improved walk rate, but the improvement can also be seen in a higher strike rate. The pitcher can put the ball where they want which is normally near the edge of the strike zone. Hitters will be forced to swing and they will either put the ball in play or get a strike. Besides the best pitchers, I am interested in pitchers with grades under 50 and then look to see why the bad value exists. These pitchers either have multiple release points (e.g. Raisel Iglesias), completely inconsistent delivery (e.g. Ubaldo Jimenez), or the pitcher moves his positioning on the mound. Here is an example of the last one. While working on player previews for 2017 season the other day, I noticed Mike Fiers had a mechanics grade of 28 after having a value of 48 the previous season. This change is a huge red flag. The first place I go to is the pitcher’s FanGraphs page and select the Pitchf/x tab and then Game Charts to look for any patterns. Here is Fiers’s chart: He looks to have two release point clusters. When a person sees this pattern, they know the person was moving on the rubber. It could be an in-season change or some pitchers change position depending on batter handedness (e.g. Trevor Bauer in the past). Off to BrooksBaseball.com we go to find the cause. Here is his chart of horizontal release point by game this past season. The change is obvious with the jump happening early June. Before the change, he had a 5.20 ERA. After he made the change his ERA was 4.13. Not a huge difference, but if Fiers showed marketable change at least there was a known event to be the cause. Hopefully, both of these sheets can help fantasy owners make more informed decisions when managing their roster. Finally, I know quite a bit of the information is new and possibly confusing. Let me know if you have any questions.