The other day, I discussed predicting pitchers’ strikeout rates using xK%. I will conduct the same exercise today in regard to predicting walks. Using my best intuition, I want to see how well a pitcher’s walk rate (BB%) actually correlates with what his walk rate should be (expected BB%, henceforth “xBB%”). Similarly to xK%, I used my intuition to best identify reliable indicators of a pitcher’s true walk rate using readily available data.
An xBB% metric, like xK%, would not only if a pitcher perennially over-performs (or under-performs) his walk rate but also if he happened to do so on a given year. This article will conclude by looking at how the difference in actual and expected walk rates (BB – xBB%) varied between 2014 and career numbers, lending some insight into the (un)luckiness of each pitcher.
Courtesy of FanGraphs, I constructed another set of pitching data spanning 2010 through 2014. This time, I focused primarily on what I thought would correlate with walk rate: inability to pitch in the zone and inability to incur swings on pitches out of the zone. I also throw in first-pitch strike rate: I predict that counts that start with a ball are more likely to end in a walk than those that start with a strike. Because FanGraphs’ data measures ability rather than inability — “Zone%” measures how often a pitcher hits the zone; “O-Swing%” measures how often batters swing at pitches out of the zone; “F-Strike%” measures the rate of first-pitch strikes — each variable should have a negative coefficient attached to it.
I specify a handful of variations before deciding on a final version. Instead of using split-season data (that is, each pitcher’s individual seasons from 2010 to 2014) for qualified pitchers, I use aggregated statistics because the results better fit the data by a sizable margin. This surprised me because there were about half as many observations, but it’s also not surprising because each observation is, itself, a larger sample size than before.
At one point, I tried creating my own variable: looks (non-swings) at pitches out of the zone. I created a variable by finding the percentage of pitches out of the zone (1 – Zone%) and multiplied it by how often a batter refused to swing at them (1 – O-Swing%). This version of the model predicted a nice fit, but it was slightly worse than leaving the variables separated. Also, I ran separate-but-equal regressions for PITCHf/x data and FanGraphs’ own data. The PITCHf/x data appeared to be slightly more accurate, so I proceeded using them.
The graph plots actual walk rates versus expected walk rates. The regression yielded the following equation:
xBB% = .3766176 – .2103522*O-Swing%(pfx) – .1105723*Zone%(pfx) – .3062822*F-Strike%
R-squared = .6433
Again, R-squared indicates how well the model fits the data. An R-squared of .64 is not as exciting as the R-squared I got for xK%; it means the model predicts about 64 percent of the fit, and 36 percent is explained by things I haven’t included in the model. Certainly, more variables could help explain xBB%. I am already considering combining FanGraphs’ PITCHf/x data with some of Baseball Reference‘s data, which does a great job of keeping track of the number of 3-0 counts, four-pitch walks and so on.
And again, for the reader to use the equation above to his or her benefit, one would plug in the appropriate values for a player in a given season or time frame and determine his xBB%. Then one could compare the xBB% to single-season or career BB% to derive some kind of meaningful results. And (one more) again, I have already taken the liberty of doing this for you.
Instead of including every pitcher from the sample, I narrowed it down to only pitchers with at least three years’ worth of data in order to yield some kind of statistically significant results. (Note: a three-year sample is a small sample, but three individual samples of 160+ innings is large enough to produce some arguably robust results.) “Avg BB% – xBB%” (or “diff%”) takes the average of a pitcher’s difference between actual and expected walk rates from 2010 to 2014. It indicates how well (or poorly) he performs compared to his xBB%: the lower a number, the better. This time, I included “t-score”, which measures how reliable diff% is. The key value here is 1.96; anything greater than that means his diff% is reliable. (1.00 to 1.96 is somewhat reliable; anything less than 1.00 is very unreliable.) Again, this is slightly problematic because there are five observations (years) at most, but it’s the best and simplest usable indicator of simplicity.
Thus, Mark Buehrle, Mike Leake, Hiroki Kuroda, Doug Fister, Tim Hudson, Zack Greinke, Dan Haren and Bartolo Colon can all reasonably be expected to consistently out-perform their xBB% in any given year. Likewise, Aaron Harang, Colby Lewis, Ervin Santana and Mat Latos can all reasonably be expected to under-perform their xBB%. For everyone else, their diff% values don’t mean a whole lot. For example, R.A. Dickey‘s diff% of +0.03% doesn’t mean he’s more likely than someone else to pitch exactly as good as his xBB% predicts him to; in fact, his standard deviation (StdDev) of 0.93% indicates he’s less likely than just about anyone to do so. (What it really means is there is only a two-thirds chance his diff% will be between -0.90% and +0.96%.)
As with xK%, I compiled a list of fantasy-relevant starters with only two years’ worth of data that see sizable fluctuations between 2013 and 2014. Their data, at this point, is impossible (nay, ill-advised) to interpret now, but it is worth monitoring.
Name: [2013 diff%, 2014 diff%]
- Chris Tillman: -0.76%, -1.51%
- Hisashi Iwakuma: +0.05%, -1.36%
- Jorge De La Rosa: +0.95%, -1.12%
- Julio Teheran: +0.43%, -1.13%
- Shelby Miller: +0.91%, +1.44%
- Wily Peralta: +0.29%, -1.46%
Miller is an interesting case: he was atrociously bad about gifting free passes in 2014, but his diff% was only marginally worse than it was in 2013. It’s possible that he was a smart buy-low for the braves — but it’s also possible that Miller not only perennially under-performs his xBB% but is also trending in the wrong direction.
Here are fantasy-relevant players with a) only 2014 data, and b) outlier diff% values:
- Alex Cobb: -0.58%
- Garrett Richards: -1.49%
- Jake Odorizzi: +1.04%
- Tanner Roark: -1.16%
- Tyson Ross: +0.78%
- Yordano Ventura: +1.34%
I’m not gonna lie, I have no idea why Cobb, Corey Kluber and others show up as only having one year of data when they have two in the xK% dataset. This is something I noticed now. Their exclusion doesn’t fundamentally change the model’s fit whatsoever because it did not rely on split-season data; I’m just curious why it didn’t show up in FanGraphs’ leaderboards. Oh well.
Implications: Richards and Roark perhaps over-performed. Meanwhile, it’s possible that Odorizzi, Ross and Ventura will improve (or regress) compared to last year. I’m excited about all of that. Richards will probably be pretty over-valued on draft day.
This just in, folks: Corey Kluber leads all MLB pitchers in wins above replacement (WAR). The great thing about running your own website is you have full discretion to toot your own horn when you please. As much as I find it tacky to do so, I made bold predictions for a reason: to see if my projections are actually worth a damn. I just wish I had time to make more; I should have started early in the offseason as I ran out of time the longer the academic year has worn on. (I’m a graduate student, so publishing to this website is not always the most optimal use of my time. According to societal expectations, at least — I think it’s a great use of my time!)
Anyway, let’s revisit my bold predictions to discuss a) their accuracy thus far, and b) why they have (or have not) been accurate. Here they are, in chronological order:
Ross is ranked 31st of all starters, according to ESPN’s Player Rater. Instead of rehashing details, you can read the linked article to see why I glowed about Ross this offseason and have chosen him as a streamer several times already this year (before he gained more recognition and, consequently, more ownership). That he qualifies as a reliever in ESPN leagues is a huge plus as well. I readily admit it’s not insane for a random names to rank highly in the player rater; just check out the names around Ross’, including Alfredo Simon, Josh Beckett, Aaron Harang and Collin McHugh. Unlike the names I mentioned, though, I think Ross has the natural ability to stay there, given his strikeout propensity that limit the damage done by walks (which, by the way, is a problem nowhere near as bad as Shelby Miller‘s — I guess six wins will mask his atrociously bad WHIP that will blow up in his face sooner rather than later.) Ross is still available in 21 percent of ESPN leagues, so if he’s out there, you should grab him. Just don’t expect him to keep winning as often in front of that terrible San Diego offense.
As terribly as this prediction has turned out — Miller is batting .151/.230/.247 with 3 HR and 3 SB — I do not regret making it. Miller has struck out in 28 percent of his plate appearances, which is way, way worse than he ever was in Triple-A or even last year, when he struck out 17 percent of the time. It pains me deeply that The Triple Machine hasn’t hit a triple. Have I given up on him this year? Honestly, yes. His batting average on balls in play is grossly unlucky right now, but even regression to the mean won’t fix what his strikeout tendency has broken. But I still like him as a sleeper for next year, or even as a late bloomer this year. If he can demonstrate an improvement in his plate discipline as the year wears on, I will give him another chance. It upsets me, though, that he had such a hot spring. It fuels the fire of analysts who criticize spring training stats as unreliable. I agree, to an extent, but Miller’s spring stats were an extension of his 2013 season — albeit an extension inflated by some good luck. It’s worth emphasizing here that strikeouts really aren’t luck-based, so to say the his spring training was lucky is an ignorant dismissal.
Corey Kluber is this year’s Hisashi Iwakuma (aka big breakout candidate)
There’s one thing I, at least, can privately appreciate about my bold predictions: I abided by all of them in every single I’m in, unless someone happened to grab a pitcher before me. Ultimately, in four leagues, I grabbed Ross and Miller in four of them, and Kluber in three — and in the fourth one, I promptly traded Jayson Werth and Tyson Ross (who I drafted in the last round) for Norichika Aoki and Corey Kluber (this is a points league, so Aoki carries some value for his lack of K’s and contact approach). Did I win the trade? Who knows — I traded one guy I liked for another I liked more. Point is, I actually rolled with my bold predictions. Might as well eat my words, right? (Is that how that saying goes?)
I got Kluber in the equivalent of the last round in every draft and for $1 in my primary keeper auction league. Yes, I’m bragging. But, more importantly, this isn’t a revelation to me. I knew Kluber would be good based on last year’s peripherals, as did a host of other people on FanGraphs (namely, Carson Cistulli and the Corey Kluber Society). But a lot of people didn’t see it coming, which is crazy to me, and it makes me question what it really takes to become a paid professional “fantasy expert.” Tristan H. Cockcroft ranked Kluber 58th of starting pitchers this preseason, which is better than I expected, but look at some of the names above him: Matt Garza? Justin Masterson? Zack Wheeler? For a guy who invests so much in seeing an improvement in skills, Wheeler has been, for his entire career, buying up billboards to plaster them with slogans such as I HAVE CONTROL ISSUES. Kluber is essentially the antithesis of Wheeler. And, yet, who has the smaller track record? Ridiculous… (In Eric Karabell’s defense, he said pitching is so deep this year that owners may not be able to draft Kluber, which was a roundabout way of indicating he liked him, at least somewhat, heading into draft day.)
Anyway, I’m clearly on a rant, and I need to get this train back on the rails. Kluber is somehow not 100-percent owned at this point — he’s 99.9-percent owned, but hey, at least I’m not lying — yet he’s striking out everyone and their mothers. I don’t know if he continues to strike out 10 per nine innings (10.28 K/9), but the percentage of swinging strikes he has produced has jumped 1.4 percent, placing in the top 1o in the category, behind Max Scherzer and ahead of Madison Bumgarner. This is all a long-winded way of saying he could, and perhaps should, be a 200-K guy this year. In that sense, maybe he’s not a buy-low guy, but his lack of name recognition and his .350 BABIP makes him a prime candidate to be exactly that. A handful of rankings have him in the 35-to-40 range; even then, I can give you a case to trade perhaps a dozen names ahead of him for Kluber, including Gio Gonzalez, Matt Cain and, yes, maybe even Justin Verlander (who, at this point, is still owned in most leagues simply because of name recognition and past performance; and while I understand the importance of past performance, do not let yourself be blinded by nostalgia).
This one is random, but hey, it’s legit: Haren has only a 6.89 K/9 right now. You can read the linked post to find out way. I may rip him a little too hard — his control still makes him a fairly solid starter — but he’s more of a Kyle Lohse these days than, well, a Corey Kluber. Lohse is serviceable, but he’s not elite, and Haren should be able to net you an extra win or two along the way in front of a lethal Dodgers offense.
OK, that’s it. I’m 3-for-4 in my bold predictions so far this year, which is a pretty good day at the plate, so I’ll take it.
Also, the academic year is winding down, and once it winds down completely, Need a Streamer will ramp up with more content. Stay tuned, and thanks for reading.