The Change: Hughes, Shoemaker, and The Limits of K-BB

Yes, the best in-season ERA estimator is strikeouts minus walks. You can use that tidbit to find a few sleepers in season, for sure.

But when you look at Matt Shoemaker and Phil Hughes right now, you realize there are perils when it comes to using the stat in the offseason. And really, you may start to see some of the limits of the stat even when it comes to in-season work.

Take a quick look at the K-BB leaderboard, with a few choice stats added in that will help provide context for our ‘sleepers.’

At the very top, we already have an interesting test case. Both Corey Kluber and James Shields are elite when it comes to getting strikeouts and limiting walks, but they have different outcomes when it comes to balls in play. Kluber is suffering from a million paper cuts with a bad batting average on balls in play, and Shields is allowing too many homers. Make me pick one, and I’ll take Kluber — his average fastball velocity is better, and he doesn’t have a history of homer problems like Shields. A few of those inside righty-on-righty changeups are always going to leave the yard.

You’ll see this in other players down the leaderboard, too. It’s easy enough to use strikeouts minus walks to appreciate Carlos Carrasco and Clayton Kershaw despite iffy ERAs, but what about Mike Fiers and Bartolo Colon? They’ve both had homer issues in the past, and their fastballs don’t crack 90 on average. For a while, Ted Lilly had great K-BB% and yet gave up all those homers and had ERAs in the high threes.

Of course, a pitcher’s true talent home run rate takes a long time to determine. Two or three years, even. So maybe a guy like Rubby De La Rosa doesn’t actually have a homer problem, maybe he’s just been a little unlucky so far in his small sample seasons. And this is also why K-BB is such a great small-sample in-season predictor: it ignores that data which takes the longest to become meaningful.

But not all pitchers are created equal when it comes to the long ball. Eventually, they do show their stripes. As exciting as Mike Fiers may be by K-BB%, he’s pitched 270 innings in the big leagues and given up more than a homer per nine. With that fastball, it makes sense. When he misses, the ball goes boom. Colon’s sample is even larger, though it might have as much to do with his limited repertoire (85% fastballs) as it does with fastball velocity.

Just off the list, but still in the top 35 by strikeouts minus walks, are Anibal Sanchez and CC Sabathia. Both suffering from homer problems, these two veterans don’t actually have a long history of homer problems. But Sanchez is still showing above-average fastball velocity, lives in a nicer park, and hasn’t had a long run-up into this problem. CC has been giving up gophers for four years now, and his fastball velocity is almost Colonian.

And so we come to the case of Phil Hughes and Matt Shoemaker, who finished last season 14th and 19th on the K-BB% leaderboard, respectively.

The first way to separate these two is recognize the role of season-to-season correlations. Shoemaker finished the season 19th, yes, but he also may have outpaced his true-talent strikeout and walk rates in 136 rookie innings. His minor league numbers weren’t that great, and the projections sniffed that out. Our depth chart projections had him 55th in the big leagues in K-BB going into the season.

But then Shoemaker came out and did it again, at least by K-BB. His rate would have been 21st on this list — just ahead of Madison Bumgarner’s — if he qualified. Still, Shoemaker’s velocity is down to 89 mph and he’s had a history of bad home run rates in the minors. Any excitement about his future needs to go more in the Mike Fiers bin then in the Carlos Carrasco one.

Phil Hughes has average velocity, and was projected to have a great K-BB this year (12th in projections). His homer rate is back to where it was with the Yankees, and so it’s hard to project him for much better, too. Added in is the fact that his strikeout rate has really suffered this year, and we know he’s had a history of searching for a good breaking ball, and a long track record of meh strikeout and swinging strike rates. With strikeout rate stabilizing, and homers always an issue for him, it’s hard to call him much of a sleeper.

So homers are the missing link between a true K-BB sleeper and a false idol, probably. The projections will ignore them, but we can’t. Corey Kluber, Carlos Carrasco, Anibal Sanchez, and maybe Rubby de la Rosa? Sure, their K-BB numbers make them exciting.

But Bartolo Colon, Matt Shoemaker, Phil Hughes, and Mike Fiers? Look at their arsenals. They’re going to give up homers, and it may not get better as the weather warms.