Reviewing 2015 Pod Projections: Jacob deGrom

It’s time for another Pod Projection review and today it’s new top 10 starter Jacob deGrom. After a dominating performance during his 2014 rookie campaign that essentially came out of nowhere, we all wondered how much regression he would experience in 2015. Instead, he laughed at the notion of regression and took his performance to further heights, earning more than $24 and finishing as the ninth most valuable starter. Refresh your memory by reviewing my initial Pod Projection post.

IP: 175 projected vs 191 actual

Since he was initially slated to open the season as the Mets fifth starter, I had projected just 29 starts for him. I was close, he started 30 games. The big difference came from his innings pitched per start. My 175 innings over 29 starters equated to 6.0 innings a start, but he actually averaged about 6.4, no doubt thanks to better run prevention than we all predicted. He should hop over the 200 innings plateau in 2016.

K%: 22.7% projected vs 27.3% actual

What do you do with a pitcher who posted lackluster strikeout rates in the minors, but suddenly learned the art of the strikeout seemingly overnight in the Majors? deGrom’s track record presented a forecasting challenge. In my original writeup, I opined that “I think we can ignore those minor league rates and treat him as a new pitcher.” I was proven correct, but that still didn’t help me come anywhere close to the strikeout rate he ended up posting. Then again, I may have projected an even lower rate if I didn’t feel that way.

deGrom gained fastball velocity, which significantly boosted his two-seamer’s SwStk% and all his non-two-seamers generated SwStk% in the double digits. Talk about downside protection. Interestingly, his slider is arguably his worst pitch, as it has generated the lowest SwStk% of his three secondary offerings and the lowest ground ball rate. Yet he throws it most often. Normally I might suggest some upside if he reduces its usage in favor of his more effective changeup and curve ball, but I cannot imagine there would be any additional upside. Plus, those pitches might become less effective if used more.

BB%: 7.4% projected vs 5.1% actual

It wasn’t enough for deGrom to increase his strikeout rate, as he sharpened his control as well. So now he’s a whiff-inducing and strike-throwing machine. That’s tough to beat.

GB%/LD%/FB%: 47% / 21% / 32% projected vs 44% / 21% / 35%

He traded some 2014 line drives for 2015 fly balls, which was a good swap, and I had forecasted some additional grounders, which were instead fly balls. He posted a high ground ball rate at Triple-A in 2014 which provided some optimism, but it looks like he’s more of a league average distribution type guy now.

HR/FB%: 10.0% projected vs 9.5% actual

You had to figure his 6.1% mark in 2014 was going to regress toward the league average, and that’s exactly what happened. You also had to be concerned about the fences being moved in at Citi Field, but that clearly did not affect him as he posted a minuscule 2.9% HR/FB rate at home this season.

BABIP: .300 projected vs .271 actual

What concerned me was a projected slight ground ball tilt in his batted ball profile and a relatively weak ability to induce pop-ups. Add to that what figured to be a poor defensive Mets unit behind him and you come up with a BABIP that is projected to be worse than the league average. Turns out, the Mets weren’t bad in the field, posting a UZR/150 just barely in positive territory, but ranking in the middle of the pack of all teams, while their BABIP allowed ranked in the top 10.

But even given the better than expected defense, deGrom’s batted ball profile was league average and his Soft%, if that matters, only ranked 26th. There is nothing in his statistical profile that could explain such a suppressed BABIP, so you have to assume a reversion in 2016.

–

Below is a summary of how deGrom was projected by all systems, along with his actual totals.

Obviously, every system (even the Fans!) expected regression from his 2014 performance and Steamer believed in his performance the least. Even though I forecasted better underlying skills, ZiPS’ projected ERA was lower because it projected a low HR/9 rate. Basically, the system believed his 6.1% HR/FB rate in 2014 was far more real than I did. ZiPS does not regress the luck metrics as much as I, or Steamer, does, believing skill is more involved. It’s an odd thing to believe considering we had just one season of performance with which to make that determination for deGrom.