Often, in evaluating players, scouts and analysts will throw out comparisons to other players to simplify their report. Wil Myers was drafted as a catcher but, being lanky and an advanced hitter, many compared him to Dale Murphy, an easily recognized name who went from catcher to the outfield. Comps are usually used for prospect evaluation and, while no two players are alike, they provide a quick snapshot of what kind of a player someone might turn out to be.
But these comparisons aren’t seen often beyond the minor league level. A player hits the big leagues and they become their own player and other players become compared to them. Beyond the Box Score has come up with an interesting project recently, though, that examines a combination of Pitch F/X data to compare current pitchers to each other.
There’s a lot of math that is beyond my comprehension, but basically, Stephen Loftus has done the heavy lifting by analyzing pitch type, velocity, movement, location and release point to come up with correlations. There are obvious outliers – R.A. Dickey isn’t similar to anybody, for instance – but it’s an interesting idea. I went through to find who the strongest comps (correlation of .800 or better) were for the Royals pitching staff. Here’s what turned up:
Drabek has a career WHIP of 1.671 and has already had two Tommy John surgeries.
Nicasio is a high risk, high upside guy for the Rockies. The Guthrie similarity is interesting but not surprising (more on that in a moment) and Bud Norris was a popular trade target for many teams.
Jeremy Guthrie, it turns out, compares to a great number of pitchers. Some are big time names. Cy Young candidates like Justin Verlander (hence the title of this column), Zack Greinke, Jordan Zimmermann, CC Sabathia and others. Given the way that Loftuss approached his study, it tends towards the prototypical pitcher. Dickey is far from anyone comparatively, because nobody else throws a knuckleball, nor would most throw it as hard as Dickey if they did. Similarly, Joe Smith of the Indians is a sidearm/submariner, so his unique release point would skew results.
Guthrie is a typical pitcher, though, and since he throws hard, has a typical assortment of pitches, and mixes locations, he gets very similar scores to the rest of the group. My favorite comp here? Jason Hammel, who was traded for Guthrie prior to the 2012 season, gets the top correlation.
These comparisons come from 2012 data, so finding to late-inning relievers as comps for Davis isn’t too surprising since Davis spent the whole season in Tampa’s bullpen.
That’s a nice list of names. That backs up the feeling of many that Hochevar has strong stuff, but it never comes together.
Wolf is a soft-tossing lefty. I included Dickey because Chen has the highest similarity score of anyone on the team, mostly because his velocity will be most similar to Dickey’s.
Last year’s relief pitchers are a bit more difficult to correlate. Part of that is probably sample size as well as the idea that they’ll use their top two or three pitches, but won’t have to dip into their fourth or fifth pitch often if at all, so they’ll have less to connect with other pitchers.
|Tim Collins||Tyler Clippard||0.730|
|Kelvin Herrera||Stephen Strasburg||0.776|
|Greg Holland||Alexi Ogando||0.766|
I included Halladay in Holland’s similarity scores because it’s the least similar (by these measurements) score of any Royals pitcher to another.
Obviously, since the data is just based on Pitch F/X, the actual performance of the pitcher in question isn’t accounted for, so that allows for Luke Hochevar to be compared to Zack Greinke, but it’s an illustration of how approach and an arsenal of pitches can be similar, but other factors come in to play. There’s no distinction between high leverage or low leverage situations. They don’t factor in pitch sequencing or the batters faced. There are also intangibles in play. Justin Verlander may reach back and go for the 99 mph fastball in a count with a batter while Jeremy Guthrie may go with an entirely different pitch (mostly because he doesn’t hit 99, but he can touch 96) or location in the exact same situation.
So it’s tempting to make these similarities more than they are. It’s an idea that might be incorporated down the line for other rookies or minor leaguers though. It’s one thing to see it on the field, but to also apply the data could make those “This guy is the next Verlander” comparisons even stronger. Does it mean the Royals have Verlander-light in Guthrie? Of course not. But it’s an interesting application of data and mathematics to player analysis.