Most fantasy players are aware of the pitching metrics FIP, xFIP, and SIERA, and use them when analyzing pitchers to draft for their team. These metrics are great at indicating when a pitcher may be outperforming his true skill, or have fallen victim to poor fortune. However, when analyzing pitchers it seems that a lot of people simply stake their flag in one of, or all of, the metrics to decide a player’s talent and value. This approach can be misguided because while FIP is calculated around the three outcomes a pitcher can control, walks, strikeouts, and home runs, it does not directly account for the types of contact a pitcher induces, the defense behind him, or in the case of xFIP, his ability to limit home runs. SIERA attempts to incorporate these factors by including net ground balls (GB-FB) and weighting them alongside a pitcher’s strikeout and walk rates. These stats however don’t paint a full picture, and can misguide you when drafting your team. So today, let me show you examples of how that can happen.
What if you’re looking at a pitcher’s Fangraphs page and his ERA is 3.13, his FIP is 4.40, his xFIP is 4.93, and his SIERA is 4.35? Do you:
- Laugh at the uninitiated fans that say he’s good and don’t know of our special sabermetric stats.
- Tell your league Taco that he is a great sleeper.
- Look at his batted ball numbers, pitch f/x data, and heatmaps to see why there is such a large gap between surface stats and peripheral stats.
There are no “wrong” answers here, except for a, b, and c. Those are wrong, you should pick d. The pitcher in question was Mr. Anti-FIP himself, Marco Estrada and his 2015 stats. If you ignored Estrada in the following draft you missed out on another solid season where he turned in a 3.48 ERA over 176 innings with 8.5 K/9. He wasn’t a league winner, but considering that everybody was pegging Marco Estrada as a bust last season means you could’ve gotten a great value if you picked him. Estrada maintained his success despite the underlying metrics because he induced a lot of soft contact and garnered plenty of popups by throwing his changeup low in the zone or out of the strike zone. With this pitching style FIP, xFIP, and SIERA will perpetually misjudge his true skill, since Estrada doesn’t rack up big strikeout numbers, allows a lot of flyballs, and his walk rate will be elevated since he has to go outside the zone often.
For every pitcher that outperforms ERA estimators, there will be one that under performs according to ERA estimators. Consider Michael Pineda, who for the last two years has had an ERA much higher than his FIP and SIERA, pair that with his K-BB rate and he’ll look like an ace in the making. The problem is that Pineda only throws two pitches, and while his slider is very good, his fastball is not so good. Pineda’s stellar walk rate is somewhat artificial, as he suffers from an affliction I like to call “Phil Hughes Disease”, where the sufferer repeatedly throws their primary pitch (for Pineda it’s the fastball) in the heart of the zone. The pitcher will maintain a good walk rate, but will allow more runs than he otherwise would because he throws the ball over the plate too much. Pineda has also had a high HR/FB ratio over the last two seasons, and when xFIP corrects for that it looks like Pineda got unlucky. Pineda pitches in a tough division in a ballpark that was the least forgiving for home runs in the MLB per ESPN park factors, higher than even Coors Field. Combine that with a straight fastball that is routinely over the plate and you’ll get a pitcher with a higher than average home run rate. If Pineda doesn’t improve or at least move to a better park, his FIP, xFIP, and SIERA will perpetually overestimate his true skill. He’ll always rack up strikeouts with his slider, he’ll limit walks by tossing his fastball over the plate, and it’ll seem like his home run rate is abnormally high.
The case of Marco Estrada has been documented before, and the case of Michael Pineda has frustrated fantasy owners before, but these players offer valuable insight into possible deficiencies of ERA estimators. If a pitcher has an ERA that is much lower or higher than their FIP or SIERA, look at their batted ball profiles, their pitch f/x data, and their heatmaps. If there is a noticeable outlier in some category, that may be the reason for the difference. The following three pitchers had big gaps between their ERA and ERA estimators in 2016, so I took a closer look to see why this may have happened and whether it affects their value for the upcoming 2017 season.
Robbie Ray finished with an ERA of 4.90, a FIP of 3.76, an xFIP of 3.45, and a SIERA of 3.59, along with an eye popping 11.25 K/9. He seems like the perfect breakout candidate, however his 36.6% hard contact surrendered should be a big warning sign. He, like Pineda, is a two pitch pitcher that relies heavily on his slider to strike people out. If he can’t locate his slider, which he struggled with often last season, he has to go to the fastball, and since hitters know it’s coming they crush it. Ray definitely got unlucky in regards to his BABIP, which was .352, the highest in the league by 13 points (Pineda was second highest at .339), but considering that he pitches in Chase Field, which is essentially Coors Light, and surrenders such a high amount of hard contact I wouldn’t expect his BABIP to regress to average. Ray isn’t a 5 ERA pitcher, but I wouldn’t expect his ERA to come down below 4 next season. With his current pitching style he’ll always have a higher than average BABIP, leading to a lower than average strand rate, and an artificially low FIP and SIERA. Keep an eye on him during spring training, because if he starts developing a third pitch he might rise up my draft board, but for now I wouldn’t reach much higher than his current ADP of 201.
FIP and SIERA say that Eickhoff should’ve been around a 4.10-4.20 ERA last season, but Eickhoff has been able to induce infield flyballs at an Estrada-esque 12.6% over his career as well as having 20.1% soft contact against. The way that he mixes pitches and velocities will help him maintain this success. His current ADP is 224th and he’s a nice value there, I would take him ahead of Adam Wainwright at 205th personally. I think prime Eickhoff could be what Jake Odorizzi is now, not in pitching style, but statistically. Reliable and perpetually underrated, with ERA estimators always about a half run below his ERA.
Brandon Finnegan had the biggest negative differential between his ERA and his FIP in 2016, so I’d be remiss not to mention him. As far as the three true outcomes go, Finnegan did everything wrong. He walked batters, allowed home runs, and had a below average strikeout rate. He had a 22.5% line drive rate and a 36% hard contact rate, yet he somehow escaped the season with a .256 BABIP against, 96 points lower than Robbie Ray despite similar batted ball profiles. I wouldn’t call anyone that has to play for the Reds lucky, but boy Brandon Finnegan got away with a lot last season. He’s still young, so there is room for improvement, but his current ADP is 398th and likewise should only be drafted in deep or NL only leagues.
ERA estimators are great tools to use when evaluating players, and this article isn’t about denouncing their viability for fantasy players. Most of the time ERA estimators will align with a pitcher’s true skill, but it’s important to recognize potential outliers. Too many times I’ve seen people quote FIP or SIERA as the end all be all of pitcher skills, but if you are using these metrics in a vacuum then it’s no better than quoting a pitcher’s ERA. If you take the extra step when analyzing pitchers for your team it could reap great rewards. You could have been in on Marco Estrada or Tanner Roark when no one else was.
Please direct all questions, comments, and hate mail to @Elliott_TFR
(Note from the Editor: As we continue to ramp up for the 2017 fantasy baseball season, we are still seeking writers. If you think you may be interested please click here for more details)