There are no other pitchers like Colin Poche

Something to remember as we talk about Colin Poche: no one really knows how this is going to work out. There just aren’t very many pitchers like Poche so comps are hard to find.

To begin with, Poche has thrown his four-seam fastball for 88% of his pitches this year. I went looking for other pitchers who had matched that rate over a full season of work. Dating back to 2014, I found four. That becomes six if one lowers the qualification bar to a comical 100 pitches thrown.

Jake McGee and Sean Doolittle have made excellent careers out of throwing greater than 85% four-seamers. Tony Cingrani and and Chad Green have done it for a single season each. John Holdzkom and Jack Leathersich did it in the majors for a hot second, but haven’t hung around.

All of those pitchers, other than Leathersich, can rely on mid-90s or better velocity. Poche’s fastball, however, hass averaged 93 mph. Rather than high-octane velo, Poche's fastball works in less obvious ways, with good extension, and elite rise.

So the first question about Poche, when he was promoted to the major leagues was "Will that fastball actually play?"

He's answered that one emphatically. It plays.

Among pitchers with at least 200 pitches thrown, Poche has generated a whiff on a higher percentage of his fastballs than every major league pitcher other than Josh Hader and Emilio Pagan. He's in a virtual tie with Pagan, but he throws more fastballs, so I'm comfortable ranking him ahead of Pagan in pitch effectiveness. Let that sink in: Colin Poche has the second best fastball in baseball.

The next question, though, is "Will his elite fastball, thrown almost all the time, make Poche a good back-end high-leverage reliever?"

His strikeout rate has been massive, at 35%. His walk rate has been a little bit too high at 10%. His groundball rate has been really really low 20% — lower than all of those comparable fastball-throwing pitchers — and that's where even more uncertainty comes in.

Fly balls are good for pitchers because they're almost always caught. Fly balls are bad for pitchers because they sometimes become home runs. Groundballs are more likely to end up as singles, but a ball on the ground is never a home run.

Sean Doolittle has carried a 30% groundball rate throughout his career, but in one season he got it down to 23%, so he might be the most analogous pitcher on the list. Josh Hader, the one reliever who definitely has a better fastball than Poche, is working with a 23% groundball rate this year.

Let's talk about Doolittle, Hader, ERA estimators, and edge cases in the pitching models.

On the Edge, Where Models Argue

Sean Doolittle has been an excellent major league reliever for the better part of eight years. Like Poche, he throws almost entirely four-seam fastballs. Like Poche, he gets big whiff numbers on those fastballs. Unlike Poche, he almost never walks batters, so he’s keeping men off the bases — if he’s going to give up a home run in one of those frequent fly balls, it’s probably going to be a solo home run.

But the interesting thing, and one of the reasons he’s been so good, is that Doolittle mostly doesn’t give up home runs. His 7.5% HR/FB is well below the average for that period; he’s been solidly better than the league average in every individual season. For more average pitchers, limiting home runs per fly balls is not a differentiating skill, but for Doolittle it almost surely is.

Poche hasn’t shown that skill yet in his first season, with a 14% HR/FB that’s just slightly better than the much-discussed and unusually-high 15% HR/FB rate that we’ve seen this year. But some models think that he has it.

By his 4.53 xFIP (loves the strikeouts, doesn’t like all those walks, assumes that 15% of fly balls will become homers, and cares about nothing else) he’s been better than his 5.03 ERA. Meanwhile, FIP knows that only 14% of the fly balls have actually been homers, so it has Poche slightly better at 4.36.

SIERA thinks he Poche should right now have a 3.44 ERA.

That’s the fourth biggest difference in baseball (minimum 30 innings pitch) between a player’s xFIP and their SIERA.

Something SIERA does differently than xFIP is that rather than assigning every pitcher an average HR/FB, it looks at their other stats and tries to make a few informed guesses. Players at the extreme of batted ball profiles (either flyball heavy or groundball heavy) tend to do better on their HR/FB. Same on players with extreme strikeout rates. And Poche checks both those boxes.

These aren’t just arbitrary adjustments by SIERA; they’re trends it pulled from real observation of the set of existing major league pitchers.

As we’ve said, though, for someone as far on the extremes as Colin Poche, that set of comparable major league pitchers is small.

Back to Josh Hader. He doesn’t throw his fastball quite as often as Poche, but at 81% he’s not far off, either. His fastball is similar to Poche’s, but even better. Over his three years in the league he’s averaged 28% groundballs, but this year, in 2019, he’s down at 23%. So Josh Hader is a pretty good comparable, and another pitcher that really is an outlier.

This season he’s carrying a 21.5% HR/FB, and his 15.6% career rate is above average for the past three years.

That should make Rays fans nervous.

A Wildcard in a Leveraged Position

Since his first big league appearance on June 8, Poche has had the third highest leverage when entering the game of any Rays pitcher, just barely behind Nick Anderson and Emilio Pagan.

Kevin Cash and the Rays FO have decided to trust him.

If they and SIERA are right, then the bullpen looks set up to carry the team to and in the playoffs.

Anderson, Pagan, and Poche are the one-inning guys who can face either righties or lefties. Oliver Drake can and does dominate lefties while facing some righties as well. Chaz Roe can and does dominate righties, often one batter at a time. Diego Castillo can either slide into that one-inning all-purpose role, be used specifically against righties, or he can go two innings at a time, perhaps as an opener, with some extra days of rest.

If the Rays and SIERA are right about Poche, then Cash has five innings of “A Bullpen” available without stretching, and eight if he were to stretch Poche, Drake, and Castillo.

If they are wrong — if it turns out that Poche is more Hader-lite and less Doolittle — then he’s going to lose the Rays some big games from his highly-leveraged spot, and Cash will need to start counting innings a little bit harder before being able to activate the A Bullpen, putting pressure on the bevy of Rays starters trying to work their way back from injury.

We won’t really know the answer for years, but we’ll start to finding it out right now, and it might just determine the Rays playoff fate.