FANTASY BASEBALL – The Invention and Use of xHard
I know. I know. If there was one thing baseball probably didn’t need – it was another needlessly complex stat, let alone one I’ve taken it upon myself to calculate. However, as the season has gone along, I’ve found my attention, especially in daily formats, often focusing on pitcher contact rates and the generally one-sided, but tempting nature of the stat. I mean, as FIP has shown us, pitchers allowing the game to fall outside of their control is generally unwanted. With the restriction of opponent contact then being the most obvious solution to this issue, it would make sense that pitchers with lower contact rates would not only be the most successful over the course of a season, but more likely to consistently produce that success on a day-to-day basis. Not to mention the suppression of contact also having a direct correlation to strikeout potential – arguably the most important facet of picking a starting pitcher in DFS.
Yet what is a stat without context? What is FIP without ERA? Herein lies the difference between real-life baseball and fantasy. As much as pitchers want strikeouts, an emphasis on the conservation of pitches has been paramount in the pitch-count era we currently dwell. Pitchers like Roy Halladay and Greg Maddux thrived throughout their careers by not missing bats, but instead missing barrels. The generation of poor contact is just as important as the suppression of it and, thanks to FanGraphs, in May the general public finally received a way to properly quantify this ideology. The website introduced Hard%, Medium%, and Soft% dating back to the 2002 season. Fantastic. Finally, the perfect pairing stat with overall contact rate. A way to not just measure the amount of contact a particular pitcher allows, but how consistently dangerous that contact is. Still, I needed a way to weight the two figures with the importance of hard contact obviously trumping just basic contact. I landed upon this formula, the structure of xHard:
(Contact Rate x 0.3) + (Hard Contact x 0.7) = xHard
Now, the numbers themselves are rough. While I decided to multiply contact rate with 0.3 (the league’s average BABIP rounded to the nearest single decimal place), BABIP itself describes batting average for balls put into play, while contact rate accounts for all contact, be it fair or foul. Hard contact’s multiplier is also more interpretation than anything as 0.7 represents ESPN writer Mark Simon’s estimate to the batting average on balls put into play at a hard rate. xHard is also more reflective than predictive – though, really, all stats are without reference. For example, Clayton Kershaw‘s mark of 42.5 is pretty unimpressive (as the scale generally sits on a 35 to 55 spectrum with the lower numbers signifying the better pitchers). However, when you consider that Kershaw’s 30.5% hard contact rate is not only a career-worst, but well higher than his 24.7% career figure, it all starts to make sense. Like any stat, it’s not the job of the formula to adjust for outliers – its yours. xHard is not the be all and end all, it is merely another tool for DFS enthusiasts to use. So, without any further verboseness, here are the Top 10 and Bottom 10 in xHard entering play on Monday, June 1st of the 111 starting pitchers who qualify:
- Garrett Richards, LAA – 35.7 (3.12 ERA)
- Francisco Liriano, PIT – 36.2 (3.47 ERA)
- Dallas Keuchel, HOU – 37.0 (1.76 ERA)
- Sonny Gray, OAK – 37.8 (1.82 ERA)
- Tyson Ross, SD – 38.8 (3.76 ERA)
- Chris Sale, CWS – 39.3 (3.66 ERA)
- Felix Hernandez, SEA – 39.4 (1.91 ERA)
- Jake Arrieta, CHC – 39.4 (3.18 ERA)
- Scott Kazmir, OAK – 39.5 (2.93 ERA)
- Gio Gonzalez, WSH – 39.5 (4.73 ERA)
So, some of this is pretty standard fare. Dallas Keuchel, Sonny Gray, and Felix Hernandez are good? Thanks, tips. However, aside from those three, none of the other members of the Top 10 in xHard rank in the Top 10 in ERA with the biggest outliers seemingly Francisco Liriano, Tyson Ross, and Gio Gonzalez whether it be based in statistics or reputation. Let’s begin with Liriano. The man has his flaws. The former Twin has the highest BB/9 of any pitcher to qualify since 2011 and has only worked deeper than six innings in five of his starts this season. Yet, for all his faults, no one can hit him. Like, literally. Liriano has by far the lowest contact rate in baseball at 66.6% which he combines with the tenth lowest hard contact rate at 23.2% – making him the only pitcher to sit inside the Top 10 in both categories. His 11.31 K/9 is fourth in the league. So what if he gives up a few runs? The same goes for Ross and Gonzalez, specifically the latter, who have struggled at times this season, but have never let it affect their strikeout rates per nine or fantastic ground ball rates – the reason their respective hard hit rates are so low. After looking at his xHard does it shock you to realize Gonzalez’s FIP is a stunning 3.03?
- Mike Fiers, MIL – 52.9 (4.53 ERA)
- Mark Buehrle, TOR – 52.4 (4.97 ERA)
- David Phelps, MIA – 51.2 (3.50 ERA)
- Tom Koehler, MIA – 50.4 (4.01 ERA)
- Jon Niese, NYM – 50.2 (4.42 ERA)
- Kyle Kendrick, COL – 50.1 (6.38 ERA)
- Chase Anderson, ARI – 49.6 (3.26 ERA)
- Nathan Eovaldi, NYY – 49.4 (4.40 ERA)
- Taijuan Walker, SEA – 49.1 (6.18 ERA)
- Josh Collmenter, ARI – 49.0 (5.05 ERA)
Here’s where xHard really comes into play: pitchers to stack against. Now, again, some of these gentlemen are just so obviously bad you don’t need my fancy, Skinemax-soundin’ stat to make it official. Eight of these men possess ERAs above 4.00. Kyle Kendrick has the league’s highest FIP at 6.09. Josh Collmenter and Mark Buehrle have two of the MLB’s slowest fastballs. However, this list does expose a few men Icarus-ing pretty hard. David Phelps for instance has a Top 50 ERA through two months, but his inclusion in both the Top 5 in general contact rate and hard contact rate (a status he shares with just Buehrle) suggests his freefall is forthcoming. Chase Anderson, a much more complex case with his 3.06 FIP actually better than his 3.26 ERA, still has the fifth highest hard contact rate at 34.9% and is even 24th in contact at 84%. Take Jon Niese‘s 2015 into account. On May 9th, Niese, who has consistently been near the top of both of these stats the entire season, had pitched to an impressive if incredibly misleading 1.95 ERA. Since that point, Niese has surrendered 20 runs in 20 innings, his ERA rising to a far more mundane 4.42. Even generating a career-best 53.2% ground ball rate, Niese could only survive so long generating so much dangerous contact.
Once again, xHard is not to be used on its lonesome. Ground ball rate, K/9, and, maybe most importantly, opponent should all be taken into account when choosing a DFS lineup for a given night. Weird outliers like Mike Pelfrey (39.6), who counteracts his 88% contact rate with a league-low 18.8% hard contact, are always going to be present. But, when used in proper conjunction with these other numbers and factors, xHard is an extremely useful stat – at least I’ve found it to be so far.
*Special thanks to Kyle Soppe for his help sound-boarding xHard (no credit for thinking of a better name)*