Watch Now: MLBPA 'Extremely Disappointed' With Proposal (1:06)

Let me preface this whole exercise by saying I don't know what to make of xERA yet.

I imagine it's like other ERA estimators in that it doesn't tell the whole story but does a good enough job simplifying it. Whether your preference is for FIP, xFIP or SIERA, you probably know to treat it more as guide than as gospel, and I'm fairly confident there will never be a single stat that perfectly articulates all of pitcher performance. Still, one can be better at it than others that came before it, and Statcast is shooting its shot with xERA.

I like what Statcast has brought to the table so far. In fact, over the past couple years, I'd say it's become my most trusted (but not exclusive) source of data. The way it tracks the flight of the ball and uses probability to determine the likely outcome of every batted-ball event has revolutionized our understanding of the game in fairly short order. And as successfully as it has used its batted-ball data to predict batting average, slugging percentage and wOBA, it was only a matter of time before it attempted the same for ERA.

My preferred ERA estimator as of now is xFIP. I used to be a FIP guy, subscribing to the idea that strikeouts, walks and home runs are the outcomes most within a pitcher's control and, therefore, the ones that say the most about individual performance, but with the way the ball carries now, seemingly anything put in the air has a chance of leaving the park. Home runs have become a more random outcome, then, but what a pitcher can still control to a large degree is how often the hitter puts the ball in the air. And that's what xFIP measures instead.

So what makes xERA different? Well, it's accounting for the outcomes those other ERA estimators can't: balls in play. Instead of dismissing them as too luck-dependent to count, it relies on the same probabilities that it so successfully uses to estimate hitter performance. Quality of contact is a thing, after all. Not every ball put in play has an equal chance of landing for a hit, which of course we always knew but couldn't account for the way Statcast can.

Because xERA measures more of what a pitcher contributes, I suspect it'll be more accurate, and at a cursory glance, it generally comes closer to depicting each pitcher's actual ERA than xFIP does. But just because it more accurately reflects what already happened doesn't mean it'll be more predictive — not necessarily.

In case it is, though, let's see where it and xFIP have the most disagreement. Below are some of the pitchers who gain the most using xERA instead of xFIP, as well as some of the pitchers who lose the most. Neither is a straight leaderboard but rather a sampling of interesting cases.

Notable gainers according to (x)ERA

ERA(x)ERA(x)FIP(x)ERA-(x)FIP
3.604.085.48-1.40
2.343.364.61-1.25
2.493.094.28-1.19
4.043.264.04-0.78
3.333.284.05-0.77
3.593.914.60-0.69
1.782.292.94-0.65
2.582.683.18-0.50
3.513.944.33-0.39
3.684.034.32-0.29
2.753.363.64-0.28
2.933.463.74-0.28
3.413.413.66-0.25
3.673.613.85-0.24
3.823.804.03-0.23

Welp, I guess that explains why some people are so high on Aaron Civale. I wasn't seeing it. A low strikeout rate along with poor ground-ball tendencies earns a "no thanks" from me, particularly in today's homer-happy environment, but if he can keep inducing low-quality contact, he has a chance to be pretty good. John Means appears to be in the same boat, though with the added misfortune of pitching for the Orioles. I won't be adding either to my sleepers list, but I won't be so resistant to picking them up if they get off to a hot start.

Julio Urias' xFIP gave me some pause when deciding whether to get fully on board with the young left-hander in his move to the starting rotation. The talent was ultimately enough to win me over, but xERA makes for a most comforting confirmation. Chris Paddack, meanwhile, has genuinely concerned me, at least as ace-caliber pitchers go. His fly-ball tendencies would seem to make him vulnerable to the long ball, as reflected by his xFIP, but xERA tells a different story.

Look at all the Twins pitchers who benefit. I was no fan of either Jose Berrios or Jake Odorizzi in early drafts, but xERA, along with the increasing likelihood of them making an even higher percentage of their starts against rebuilding AL Central clubs, has me rethinking that stance. Kenta Maeda's xERA only underscores the more obvious ways he underachieved last year and is yet another reason to be excited about his prospects for 2020.

A couple of surprising Rangers pitchers, Mike Minor and Lance Lynn, may not have overachieved by as much last year as xFIP would have you believe. Meanwhile, Tyler Glasnow's dominance becomes all the more apparent with xERA, which is crazy considering his relatively modest 2.94 xFIP would have itself ranked third among all qualifiers, behind only Gerrit Cole and Max Scherzer.

I've focused exclusively on starting pitchers here for sample size reasons, but it's worth pointing out that a couple of concerning closers, Kenley Jansen and Raisel Iglesias, fared much better in xERA than xFIP last year. Jansen's was 3.06 vs. 3.77, and Iglesias' was 3.43 vs. 3.72.

Notable losers according to (x)ERA

ERA(x)ERA(x)FIP(x)ERA-(x)FIP
4.764.663.54+1.12
4.844.893.80+1.09
5.195.444.36+1.08
6.735.975.07+0.90
4.344.573.76+0.81
3.844.693.89+0.80
3.283.943.23+0.71
4.023.863.32+0.54
3.254.053.59+0.46
3.354.994.55+0.44
4.284.173.73+0.44
3.983.803.39+0.41
4.073.833.44+0.39
2.944.113.87+0.24
2.684.053.85+0.20

What a disaster for Rockies pitchers, with German Marquez, Kyle Freeland and Jon Gray all seeing a big change for the worse. It stands to reason xFIP wouldn't fully account for the Coors Effect since it assumes a fly ball hit in any venue has an equal chance of going out of the park.

You know what else stands out to me here? A bunch of ground-ball pitchers, led by Braves up-and-comers Max Fried and Mike Soroka but also including Patrick Corbin and Dakota Hudson. I do have a big investment in both Fried and Soroka, so it's not something I care to see, but I think we all suspected that what Hudson did last year was too good to be true. The xERA makes an even more convincing case than the xFIP does. Interestingly enough, Corbin's 3.15 ERA in 2018 is supported (retroactively, of course) by a 3.28 xERA, so the disparity last year is a curious one.

Shane Bieber has his share of detractors after a Cy Young-caliber breakthrough last year, and the xERA demonstrates why. Dude gives up some hard contact. I still say his command and bat-missing ability are both so stellar that it won't matter in the end, but the Statcast data says otherwise. You shouldn't get any ideas about Robbie Ray taking a big step forward, though. His hard-hit tendencies are well documented by now.

Dinelson Lamet's presence here isn't going to win many fans, seeing as he's a popular breakout pick, but honestly, a 3.44 ERA, like his xFIP projects, is too much to ask from someone so unproven. Take it from a guy who's gotten burned the past couple years by Kyle Gibson's xFIP. Maybe xERA will be what finally convinces me to stop obsessing about his swinging-strike rate.