Using x-Stats To Find Sleepers and Busts for 2017 [Spreadsheet]

Modern baseball projection research has really benefited from the research done on x-stats (expected stats). Using new batted ball data, researchers are able to better identify a hitter’s true skill set and compare his outcomes over a season to the expected outcomes for someone with his hitting profile. We can take advantage of this data to make better fantasy baseball picks in 2017.

At Fangraphs, Alex Chamberlain posted the RotoGraphs x-Stats Omnibus last year that showcases some of the main x-stats and their formulas. I found these to be helpful but they are used to calculate expected stats that are not core traditional baseball statistics (ISO, K%, BABIP). I decided to take those x-stats and try to calculate their cumulative effect on a player’s totals from last year. Instead of just seeing that a player’s ISO should have been higher or lower last year, my calculations allowed me to see the difference between actual HRs and expected HRs.

By comparing a player’s actual 2016 stats to their expected 2016 stats based on their skill, we can really see who might be poised to progress or regress in 2017. These stats aren’t necessarily infallible but they do give us a good indication of some statistical anomalies from last season. Before we look at what I came up with and analyze it a bit, I need to give a big hat tip to the people who developed the following x-stat formulas that I used:

Compare Last Year’s Expected Stats To Actual Results

Using the K% formula, I was able to project how that would have affected a player’s batting average and other stats. I took that and then proejcted out the xISO and xBABIP for a player and calculated how that would further affect their stat totals from 2016. When all was said and done, I put all the data into a spreadsheet to be able to easily see who to target and avoid in 2017.

Check out the spreadsheet here!

Within that spreadsheet, I indicated which hitters had numbers that were way below or above expectations last year. I also put in their actual numbers and their expected numbers from my calculations.

I’ll be going into more detail in an upcoming post about some of the big differences in AVG and xAVG. I want to draw some attention to the players who seem to have discrepancies in a both AVG and power numbers.

5 Players Due For A Rebound in 2017

Let’s take a look at some players who had unlucky hitting and power numbers according to x-stats. It’s certainly not a guarantee that these players will bounce back but the x-stats seem to indicate that a correction is upcoming in 2017.

Buster Posey (C, SF) | 36 ADP | 2016 Total: .288 AVG, 14 HR | 2016 x-stats: .312 AVG, 17 HR

Posey had the lowest AVG and ISO of his career in 2016 but all indications point to an unlucky season. He had the best Hard Hit % of his career so it’s not like he wasn’t hitting well. He suffered from an unlucky HR/FB% and his lowest career BABIP but the x-stats both indicate that those should return to normal in 2017.

Matt Wieters (C, BAL) | 191 ADP | 2016 Total: .243 AVG, 17 HR | 2016 x-stats: .280 AVG, 20 HR

I do still have a bit of concern about Wieters’ waning power but the good news is that his BABIP and xBABIP indicates that his AVG was majorly unlucky last year. He’ll never be an elite catcher but he’s still a good higher-tier option at the position.

James McCann (C, DET) | 315 ADP | 2016 Total: .221 AVG, 12 HR | 2016 x-stats: .252 AVG, 18 HR

It seems like it was a bad year for catchers last year. McCann barely stayed above the Mendoza line with his poor AVG but it was seemingly more a symptom of bad luck. His power numbers were actually the best of his short career in 2016 but it seems like he could have even done better, according to his xISO. The fact that his career numbers haven’t matched up that xISO gives me some pause but his price tag is low and makes him worth a late-round gamble.

Alex Gordon (OF, KC) | 316 ADP | 2016 Total: .220 AVG, 17 HR | 2016 x-stats: .267 AVG, 21 HR

Gordon has never been elite in any fantasy category but he’s been valuable because he’s usually serviceable in all categories. With an unlucky BABIP last year, all of his numbers suffered and made him a fantasy liability. He’s a career .264 hitter and his xAVG indicates he should return to that in 2017.

Danny Espinosa (SS, LAA) | 365 ADP | 2016 Total: .209 AVG, 24 HR | 2016 x-stats: .231 AVG, 30 HR

Espinosa’s poor AVG didn’t outweigh his good power numbers last year and he found himself without a position in Washington after Trea Turner’s promotion. His trade to the Angels gives him a shot in 2017 to show that last year was a fluke. His AVG will never be respectable but last year was a bit lower than it should have been so expect a bit of a rebound.

5 Players Due For A Decline in 2017

Sometimes good players have great seasons and we over-inflate their value a bit too much. The following players are all good players but their x-stats indicate that they may be due for at least a bit of a decline.

Mookie Betts (OF, BOS) | 3 ADP | 2016 Total: .318 AVG, 31 HR | 2016 x-stats: .288 AVG, 25 HR

I pegged Mookie as a sleeper in 2015 and he was but then he turned into a legitimate star in 2016. He delivered elite value in every roto category. The bad news is that he is due for a dip in power and AVG based on xISO and xBABIP numbers. If that happens, he’ll still likely deliver good numbers across the board but “good” isn’t what you’re hoping for from the 3rd overall pick.

Nolan Arenado (3B, COL) | 5 ADP | 2016 Total: .294 AVG, 41 HR | 2016 x-stats: .277 AVG, 35 HR

It is worthy of further investigation but some players show a knack for outperforming x-stats. Arenado might be an example of that as he’s had a similar ISO and BABIP in the past couple years despite his xISO and xBABIP painting a different story. Despite his x-stats saying he should decline, Arenado may defy expectations. Even with a decline, he’s still an extremely valuable player but there’s reason to have caution in taking him 5th overall.

Brian Dozier (2B, MIN) | 32 ADP | 2016 Total: .268 AVG, 42 HR | 2016 x-stats: .252 AVG, 33 HR

I can’t even begin to explain how Dozier hit 42 HR’s last year with a .278 ISO. Nothing in his past suggests that should have been possible. The good news is that he does show evidence of being able to hit for power but around 30 HR is more realistic than 40 HR. He’s a valuable player at a thin position but expect some regression across the board.

Gary Sanchez (C, NYY) | 47 ADP | 2016 Total: .299 AVG, 20 HR | 2016 x-stats: .254 AVG, 14 HR

As I was running the numbers for this analysis, Gary Sanchez popped up at the top of the list as far as hitters that most benefited from luck in 2016. The catcher position is thin and he’s a young player who came off a productive year so his value is really high in 2017 drafts. I would stay far away unless he proves that he can defy expected stats consistently. If you expect him to sustain a .358 ISO moving forward then, man, you’re going to be disappointed.

Aledmys Diaz (SS, STL) | 145 ADP | 2016 Total: .300 AVG, 17 HR | 2016 x-stats: .254 AVG, 13 HR

Diaz doesn’t have much of a track record yet so we can’t really establish a baseline for him. The x-stats are telling us that he should decline across the board though. Hitting for .300 would make him a great mid-round shortstop but hitting for .254 would make him a liability at that spot. I’d steer clear of him based on the x-stats profile for him.

You Might Also Like

  • DS
    01/26/2017 at 3:40 PM

    Thanks for this. A couple questions:

    1) Have you looked at Perpetua’s versions? (

    2) Is there any way to compare xStats projections to something like Steamer? I ask bc if both Steamer and xStats are projecting the same level of bounceback for a player, he likely won’t be very undervalued.

    3) Are the xStats you use above park-adjusted? If not, I would think all Rockies would be expected to outperform their xStats, on average.


    • Luke Gloeckner
      01/27/2017 at 11:25 AM

      Hey, DS. Nope, I hadn’t heard of that site so that’s pretty nifty too. Glad to see others doing similar work. As far as comparing the xStats to Steamer, it certainly could be possible though it would probably have to be done on a per PA basis since xStats would be for the 2016 season and Steamer would be for 2017 season. I think Fangraphs already has Steamer projections up this year.

      The xStats aren’t park-adjusted but I’d say they’re park-neutral. The xStats are calculated by using data like how often a player hit flyballs, line drives or pulled the ball. So they’re not generally looking at things that should be affected by park factors.

      • DS
        01/27/2017 at 11:31 AM

        Here’s some more info. on his stuff:

        I don’t know the author at all, but have followed his posts at FG. None of his numbers are park-adjusted either, so for extreme parks, they prob. need to be tweaked to be used as projections.

        • Andrew Perpetua
          02/09/2017 at 3:01 AM

          I’ve recently rolled out an update that does three things:

          first, corrects for ballpark measurement bias. Each ballpark has a certain error in exit velocity etc, so that is corrected for.

          second, it corrects for game time temperature

          third, it corrects for small deviations in park factors.

          With these three factors in consideration, it appears to cover even the most extreme ballparks pretty well.

  • Colin
    01/27/2017 at 1:00 AM

    You have Sanchez posed for a big regression this season. He was going to be one of my keepers at $13. I was going to grab another catcher just to back up the investment incase he was a bust.

    Which catcher do you think I could get for cheap that might be poised to bounce back this season? For some reason I think Gomes and Norris may have a decent season. I’m in an OBP/SLG league.

    • Luke Gloeckner
      01/27/2017 at 11:21 AM

      I do like Derek Norris and James McCann as cheap options that may help give you a nice backup to Sanchez. You pointed out Gomes too and I do think he’ll bounceback a little bit so he’s a decent cheap gamble too.

  • Randy
    02/11/2017 at 2:47 PM

    Is it possible to d something like this with pitchers? The information provided here is absolutely fantastic!!

    • Kris
      02/23/2017 at 3:10 PM

      I was kinda wondering the same thing Randy!

  • Kaley
    03/03/2017 at 8:53 PM

    Very great post. I just stumbled upon your weblog and wished to
    mention that I’ve really loved surfing around your weblog posts.
    After all I will be subscribing on your rss feed and I’m hoping you write once more soon!

  • Randy
    03/23/2017 at 9:38 AM

    Do you have xstats data on more batters? I noticed Jean Segura isn’t in list and I am interested in what he was expected to do.