As pitchers rise through the minor leagues, fans are bombarded with projections of their supposed ceilings: “future ace”; “mid-rotation guy”; “back-end at best.”
These are just that: projections. The projections tell us about the pitcher’s potential, but not usually where the pitcher is now. That’s particularly true if the pitcher is an established major-leaguer, and it has been several years since that pitcher was a prospect in any sense.
So, how do we assess the “stuff” — the pitch quality, really — of a current major-league pitcher? Earned Run Average (ERA) tells us how effective the pitcher probably was overall, but doesn’t tell us how the pitcher did it. Fielder Independent Pitching (FIP) tells us how good the pitcher was at avoiding home runs and walks, and at grabbing strikeouts, but doesn’t tell us which pitches were most responsible — or at fault — for those peripherals.
And let’s face it: if you’re watching a game, and benefiting from PITCHf/x, you’re watching the particular pitches that get thrown. You’re watching the pitcher’s sequencing, his placement, and his velocity. You want to know which pitches merely keep the batter honest, and which pitches are the true weapons. And, you’d strongly prefer to get your answers to those questions from objective, verifiable facts, rather than the speculation of your broadcast analyst.
Fortunately, this is doable. The same linear weights that gave us transformative concepts like Weighted on Base Average (wOBA) and True Average (TAv), which finally allowed accurate assessments of batter production, can also be applied to different pitch types. In fact, for a few years now, Fangraphs has published the “run values” of these pitch types, as categorized by Sportvision’s PITCHf/x system, and then calculated by Fangraphs, for each pitcher.
It’s not a huge secret that these values exist, but you will find few people who cite or rely on them. In general, readers seem to be intimidated by the run values or just dismiss them as unreliable. But if you’re willing to put in the work and filter them meaningfully, these pitch run values tell you quite a bit. Today, they’ll give us an idea of the rotation talent we have around the NL Central coming into 2014.
Method: Using Fangraphs data, I downloaded the values for all major league pitchers who threw at least 50 innings in a starting role last year. (There are 187 of them). I then looked at most of the significant pitch categories tracked by PITCHf/x , including:
- 4-seam fastball (FA)
- 2-seam fastball (FT)
- Sinker (SI)
- Split-Finger Fastball (FS)
- Slider (SL),
- Curveball (CU)
- Changeup (CH)
- Knuckle-Curve (KC) and
- Knuckleball (KB).
To qualify for this study, a starter needed to use an off-speed pitch (including a splitter) at least 10% of the time, and a fastball (except the splitter) pitch at least 15% of the time. I then “binned” each starter’s pitch type into one of three categories:
- Plus: a pitch in roughly the top third by effectiveness in the league
- Average: a pitch in the middle, either slightly above or below the mean;
- Minus: a pitch in roughly the bottom third by effectiveness in the league.
Finally, I provided my guess of the five likely starting five pitchers for each NL Central rotation, and compiled where each of their most common pitches ranked.
So, who brought the “plus” pitches last year, and who did not? We’ll go through each NL Central team; pitchers are not listed in any particular order. Please remember those pitch abbreviations from just above, because using them makes the charts much more readable:
|Brewers Pitcher||Plus Pitches||Average Pitches||Below Average Pitches|
|Marco Estrada||1 (CH)||2 (FA, CU)|
|Matt Garza||1 (SL)||2 (FA, FT)|
|Yovani Gallardo||3 (FA, SL, CU)||1 (FT)|
|Kyle Lohse||4 (FA, CU, SL, CH)|
|Wily Peralta||2 (FA, SL)||1 (FT)|
|Cardinals Pitcher||Plus Pitches||Average Pitches||Below Average Pitches|
|Adam Wainwright||3 (FA, FC, CU)||1 (SI)|
|Lance Lynn||1 (FA)||1 (FT)||2 (SL, CU)|
|Shelby Miller||2 (FA, CU)|
|Michael Wacha||1 (FA)||1 (CU)|
|Joe Kelly||1 (FT)||1 (CH)||1 (SL)|
|Cubs Pitcher||Plus Pitches||Average Pitches||Below Average Pitches|
|Jeff Samardzija||4 (FA, FT, FS, SL)|
|Travis Wood||1 (CT)||2 (FA, SL)|
|Edwin Jackson||2 (FA, SL)|
|Jake Arrieta||1 (SI)||2 (FA, SL)||1 (CU)|
|Chris Rusin||1 (SI)||2 (FA, SL)||1 (CH)|
|Pirates Pitcher||Plus Pitches||Average Pitches||Below Average Pitches|
|Francisco Liriano||1 (SL)||2 (FT, CH)|
|Jeff Locke||3 (FA, CH, KC*)|
|Gerrit Cole||2 (FT, SL)||1 (FA)|
|Charlie Morton||1 (FT)||1 (CU)|
|Edinson Volquez||2 (FA, SI)||2 (CU, CH)|
|Reds Pitcher||Plus Pitches||Average Pitches||Below Average Pitches|
|Matt Latos||1 (SL)||2 (FA, FC)|
|Johnny Cueto||3 (FA, FC, CH)||1 (FT)||1 (SL)|
|Homer Bailey||2 (FA, FT)||2 (SL, CH)|
|Mike Leake||1 (CU)||3 (FC, SI, CH)|
|Tony Cingrani||1 (FA)|
League-wide, you can sort these results to into three overall categories of pitchers, somewhat corresponding to the scouting categories I mentioned earlier:
- Top of the rotation: possessing at least 2 plus pitches;
- Mid-rotation: 1 plus pitch or at least multiple average pitches;
- Back-end starter: Mix of average and below-average pitches.
There is much more to these categories than “stuff,” of course. People at the top of the ladder tend to add terrific control and strikeout ability. People at the bottom may have wonderful potential stuff, but can’t use it effectively. And people in the middle often play up or down based on other factors. Kyle Lohse, for example, has four average pitches, but has such good control that he projects the air of a top-of-the-rotation pitcher when you watch him. Wily Peralta, on the other hand, has dynamite stuff when he is on, but struggles to stack success or intimidate hitters.
Nonetheless, at least based on the charts above, this analysis suggests that:
- The Brewers’ and Cubs’ rotations consist entirely of mid-rotation talent;
- The Cardinals’ rotation is also primarily mid-rotation talent, except for Adam Wainwright, who has top-of-the rotation stuff;
- The Pirates are a mixture, with Gerrit Cole having top-of-the-rotation stuff, Edinson Volquez being a back-end starter at best, and the remainder having mid-rotation stuff; and
- The Reds, as I suggested a few weeks ago, look to be the best rotation in the division, with two starters who featured ace material last year (Cueto and Bailey), and three starters who brought mid-rotation stuff.
Also, for those of you who were bewildered by the disappearance of Shelby Miller last postseason for the Cardinals, the Cardinals’ chart provides a sabermetric explanation for the benching: of all the available starters, he was the only one who did not have a “plus” pitch in his arsenal last year. In the postseason, that tends to matter. Of course, that doesn’t explain why Miller needlessly took up a postseason roster spot.
For the Brewers, there’s nothing wrong with having a rotation featuring average starters. There’s much to be said for being consistent, particularly given the need to avoid the crippling losing streaks that helped knock the Brewers out of the division race early last year. But from a “stuff” perspective, barring some improvement, the Brewers will likely be looking up a bit at the Reds and Cardinals as the 2014 season gets underway.
Follow Jonathan on Twitter @bachlaw.
PS. For those of you who’d like to understand what I did, or to look up other starters in the league last year, I’m putting my spreadsheet online so you can review it yourself. I’ve even color-coded the pitch qualities (green, yellow, red) for easy reading. Please just credit me if you cite it. The link is here: https://skydrive.live.com/redir?page=view&resid=9EE1CA9222EEDD14!3558&authkey=!ABt-NOeNEC9UsTA
PPS. PITCHf/x categories are sometimes controversial, and the fine folks at Brooksbaseball.net re-categorize the PITCHf/x categories to make them more accurate. In another words, PITCHf/x sometimes calls certain pitch types something other than what they actually are. If you really want the most accurate information about a pitcher’s arsenal, you should make Brooks Baseball one of your regular stops. But the purposes of this exercise is to look at pitch patterns for each starter that, by whatever name you assign, tend to generate a certain result. The raw data from PITCHf/x seems adequate for that purpose, even if people don’t always agree with the labels it puts on a particular type of pitch.
PPPS. The asterisk next to Jeff Locke’s knuckle-curve is because I adjusted it downward to be a merely average pitch from where it was originally assigned. If you look at the column on my chart for knuckle curves, you’ll see why.
Finally, thanks to the New York Times for the graphic.