xStats Projects Hendricks, Uehara, Arrieta to Lead Cubs Pitchers in 2017
We recently listed the 2017 projections for Cubs hitters according to xStats, a projection model developed by Andrew Perpetua. xStats is unique because it is the only Statcast-centric pipeline that converts exit velocity, launch angle, and batted ball location to nearly every relevant stat (e.g., wOBA, FIP, batting average, homers, etc.). Those same conversions can be made for pitchers, which is why we’re now flipping the coin.
One of the best features of xStats is that it solves a big problem with isolated pitching metrics, which is that they ignore batted-ball data. For example, Kyle Hendricks is known to induce weak contact, but the extent to which we can quantify that ability has been clouded by lagging technology. Steamer and ZiPS — two of the most-used projection systems — peg Hendricks as a 3.50+ FIP and 3.30 ERA starter in 2017 even though he owns respective career marks of 3.29 FIP and 2.92 ERA.
Rather than ignoring batted balls, xStats created scFIP to estimate the probability a fly ball leaves the yard for a homer. The new model is better than FIP because ballpark influences confound a pitcher’s true ability to keep the ball in the field of play. It’s also more precise than xFIP because that stat assumes each fly ball has a 10.5 percent (league average) chance of being a homer. Perhaps most importantly, xStats allows us to complement FIP by using expected weighted on base average stat (xOBA).
xStats estimates 2017 performances using 2015 and 2016 data, but giving the latter 8 percent more weight. How do Cubs pitchers project in 2017 based on all that? Very good.
| Name | Sample | IP | scFIP | xOBA |
| Koji Uehara | 343 | 150 | 2.39 | .238 |
| Hector Rondon | 500 | 150 | 2.71 | .260 |
| Wade Davis | 463 | 150 | 2.52 | .262 |
| Pedro Strop | 475 | 150 | 2.56 | .267 |
| Jake Arrieta | 1741 | 150 | 2.98 | .269 |
| Kyle Hendricks | 1516 | 150 | 3.17 | .270 |
| Carl Edwards | 157 | 150 | 2.86 | .272 |
| Jon Lester | 1680 | 150 | 3.16 | .276 |
| Justin Grimm | 438 | 150 | 3.32 | .295 |
| Brett Anderson | 828 | 150 | 4.20 | .297 |
| John Lackey | 1676 | 150 | 3.62 | .300 |
| Mike Montgomery | 803 | 150 | 4.20 | .325 |
| Brian Duensing | 264 | 150 | 4.55 | .328 |
| Williams Perez | 746 | 150 | 4.62 | .333 |
| Eddie Butler | 676 | 150 | 5.20 | .367 |
| Aaron Brooks | 250 | 150 | 5.04 | .369 |
| Alec Mills | 19 | 150 | 6.84 | .407 |
xStats projects the Cubs rotation to be a strongpoint once again. Jake Arrieta is forecast to be the best of the crew because of his historical 2015 Cy Young campaign. But despite Arrieta’s monstrous 2015, he shares nearly the same xOBA prognosis as Hendricks (.269 and .270, respectively). Hendricks is also projected to have a 3.17 FIP, which is much better than Steamer (3.67) and ZiPS (3.51).
Rounding out the rotation is a robust 3.17 FIP from Jon Lester and 3.62 FIP from John Lackey, with Mike Montgomery and Brett Anderson both at 4.20 FIPs.
The bullpen projects very favorably as well. While we might think of Wade Davis as the clear-cut most valuable reliever, xStats actually pegs Koji Uehara to lead the pack with a 2.39 FIP. Almost every member of the relief corps is estimated to put up numbers below 3.0, although Justin Grimm is projected to have a 3.32 FIP. That’s still encouraging, though.
As always, I advise you to exercise caution when interpreting these forecasts because the model can’t capture potentially significant mechanical adjustments and small changes towards the conclusion of 2016. Nevertheless, I still find this system useful because it incorporates technology that so many fans and researchers used to dream of having.
For what it’s worth, xStats might be the first projection system in line with what we have seen from the convention-defying Hendricks. You can check out these pitching projections and so many more tools on xStats.org.