2017 pERA Update With Exit Velocity Grades

Last offseason, I created an individual pitch metric, pERA, which gives each pitch an ERA and prospect grade based on its ground ball nature and swing-and-miss capability. With the 2017 season over, I’ve compiled the final 2017 values. This year, I’ve added in exit velocity (EV) grades for each pitch.

The process I used for creating pERA is in the article linked above but here is a quick rundown.

• 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 on the average values for starting pitchers. 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.

For example, here is how Justin Verlander’s pitches have graded out over the last few years. First, here’s his pitches with an ERA grade. Additionally, the “just pitches ERA” is for the pitches and the other one includes a control element. The ERA’s are weighted by times thrown.

His fastball’s ERA is higher than the other pitches but it just doesn’t get the swings-and-misses on the level of his breakers.

Besides the ERA grades, here are the pitches graded out on the 20-80 scale.

While the none of the pitches really stand out, two factors help Verlander. His above-average fastball and control help to put a good floor on his value. The curve and slider production just add to his overall production. Now onto the new information.

Besides the updated 2017 information, I’ve included per pitch exit velocity values and grades (2015 to 2017). For these calculations, I again only used starting pitcher values. Expect relievers to have higher values across the board since they can go all out for an inning or two.

First off, here are the median exit velocity values for different pitch types.

Fastballs occupying the top few spots is not surprising since they will have higher exit velocities because the ball is coming in faster. The standard deviations are centered average about 3.2 with some types being higher or lower. I compared the actual distribution versus the theoretical standard deviation distribution. The two values almost perfectly lined up with no unusual distribution tails in either direction.

The pitches can then each be given an exit velocity grade with 50 being league average and each 10-grade change being a standard deviation from the median. For example, a 60-grade value is two standard deviations better than the median.

Now to see if any year-to-year correlation exists between. I took the two matched seasons, 2015 to 2016 and 2016 to 2017 and found the correlation between starter pitches which had at least 50 pitches in each season. I didn’t expect to see correlation but ended up with some.

The split-finger correlates but it was in just a few samples. The rest display a bit of projectability except the cutter and slider which show none.

Simply, per pitch batted ball predictability needs to be better quantified. I’m not doing the gory math today but hopefully next week. For now, just understand some small amount of season-to-season correlation may exist.

The yearly pERA update is complete with some batted ball data added in. I’m sure to reference this article in future studies and player previews to help understand a pitch’s quality.