The RotoGraphs x-Stats Omnibus, with Embedded Calculators

Updated Feb. 25, 2017

Aug. 16, 2016: Updated Alex’s xBABIP equation and added Andrew Dominijanni’s xISO equation.
May 23, 2016: Published.

Jump around in this post:
Hitter metrics: xBABIP | xISO | xHR/FB | xOBA | xK%
Pitcher metrics: xHR/FB | xLOB% | xK% | xBB%

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
Do you frequently use RotoGraphs’ “X” (expected) metrics? Do you wish they were easier to find? Have you ever commented to ask if they could be added to the leaderboards or at least wished they were all located in one spot? If so, you may want to…

BOOKMARK THIS PAGE!

I don’t know if there will ever be a time when FanGraphs has a leaderboard devoted to “X” metrics. The fantasy analysts at RotoGraphs have taken a largely vigilante approach to creating descriptive and predictive expected metrics over the years. Moreover, each metric typically undergoes an iterative process by which we improve it when new data is made publicly available to the authors.

So, this is it. This is my best attempt, on behalf of RotoGraphs’ staff and at the polite and enthusiastic behest of its readers, to centralize the freshest versions of the relevant metrics the RotoGraphs staff most frequently cites. I have also built primitive Microsoft Excel-based calculators for some (but not all) of the metrics that crunch the numbers as long as you provide the appropriate inputs. It should save us all an extra minute or two and preserve our sanity a little bit.

  • I will do my best to update this as soon as possible, whenever necessary.
  • I intend to include only FanGraphs/RotoGraphs-generated metrics, as opposed to accumulating all metrics from all websites.
  • If you notice something missing, please let me know in the comments and I’ll update ASAP!

Hitter Metrics

xBABIP: Expected Batting Average on Balls in Play

All right. So there are a few different versions of xBABIP floating around on RotoGraphs. I’ll list all of them, from newest to oldest, because I think all of them have their merits. I’ll detail the pros and cons of each as “notes” directly following the “source data” for each iteration. If you’re looking for BABIP over- and under-achievers, your best bet might be to calculate an average xBABIP by aggregating all three sources.

Andrew Perpetua’s xBABIP

Author: Andrew Perpetua
Original post URL: May 16, 2016
Source data: MLB/Baseball Savant’s Statcast data

This methodology uses Statcast data, easily the most granular data we have available. From a purely ball-in-play perspective, it is certainly the most advanced, and I think we will come to rely heavily on Andrew’s early work.

But the data are almost too granular, in that you can’t easily calculate xBABIP because there are literally thousands of angle-velocity combinations. Fortunately, Andrew performs the calculations on a daily basis and keeps them up to date in this workbook. It’s sort of a black box, but it’s not too difficult to comprehend: Andrew’s xBABIP determines the probability that a batted ball with a particular angle and velocity will become a hit, and it calculates BABIP accordingly.

Andrew acknowledges this strict batted ball approach as limiting; what it gains in batted ball granularity, it loses in runner speed and shift data. So, a ball in play by Jarrod Dyson with the exact same angle and velocity as a ball in play by David Ortiz would not have equal probabilities for landing as a hit. Yet, xBABIP treats them as equals. We know in our hearts this isn’t true. But, statistically speaking, for most other players, they probably should be treated as equals. These kinds of calculations will always be more accurate at the mean/median of the distribution than the tails, where Ortiz and Dyson exist.

The lack of speed and shift data are very minor faults for an otherwise excellent use of really nice data. But that’s why, for now, we should at least consider referencing older iterations of xBABIP for what they think about the importance of a player’s speed/baserunning — as a check on the Statcast data.

Alex Chamberlain’s xBABIP 2.0

Author: Alex Chamberlain
Original post URL: 2.0: July 13, 2016 (1.0: May 6, 2015)
Source data: FanGraphs’ batted ball data, courtesy of Baseball Info Solutions (BIS)

I wanted to create an xBABIP equation that not only used the (at the time) new BIS data but also use (1) public data that was (2) conveniently located in one spot. FanGraphs’ batted ball data, introduced a year ago, filled that role.

The batted ball data is aggregated, so it makes some (comparatively) large assumptions about balls in play. But it also makes an attempt at quantifying a hitter’s speed — something that Andrew’s currently does not do. A good example of why Andrew’s and this xBABIP might be useful: as of last week, there existed a 60-point disparity between Jackie Bradley Jr.’s xBABIPs, .307 to .367. The latter xBABIP incorporates the less-granular ball-in-play data but also the speed score (Spd); it probably overrates JBJ’s batted ball abilities, but it seems to give him some credit for being a fast dude.

Update (8/16/16): xBABIP 2.0 now uses year fixed effects to control for annual fluctuations in batted ball measurements. Use the “Year” dropdown to select the corresponding season for your analysis.

My xBABIP calculator:

If you delete the formula in the highlighted cell, simply refresh this page.

Mike Podhorzer’s xBABIP 2.0

Author: Mike Podhorzer
Original post URL: 2.0: Feb. 6, 2017 (1.0: Feb. 8, 2015)
Source data: FanGraphs’ plate discipline, batted ball data, and splits leaderboard

Mike has brand new, shiny xBABIP equation that incorporates shift metrics! About dang time one of us did something about it.

Mike’s xBABIP calculator:

If you delete the formula in the highlighted cell, simply refresh this page.

Back to top

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

xISO: Expected Isolated Power

Andrew Dominijanni’s Statcast xISO

Author: Andrew Dominijanni
Original post URL: Updated July 13, 2016 (Originally May 26, 2016)
Source data: Statcast’s exit velocity data, and FanGraphs’ batted ball data, courtesy of Baseball Info Solutions (BIS)

Down in the FanGraphs farm system, Dominijanni produced great work with new Statcast data, improving upon the original xISO equation by incorporating exit velocity data. LDFBEV stands for “LD/FB exit velocity,” in mph. Dominijanni also maintains an updated spreadsheet of hitter xISOs here.

Andrew’s xISO calculator:

If you delete the formula in the highlighted cell, simply refresh this page.

Alex Chamberlain’s BIS xISO

Author: Alex Chamberlain
Original post URL: May 6, 2015
Source data: FanGraphs’ batted ball data, courtesy of Baseball Info Solutions (BIS)

Alex’s xISO calculator:

If you delete the formula in the highlighted cell, simply refresh this page.

Back to top

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

xHR/FB: Expected Home Runs per Fly Ball

Author: Mike Podhorzer
Original post URL: 2.0: Jan. 30, 2017 (1.0: Jan. 26, 2015)
Source data: Statcast and FanGraphs’ splits leaderboard

Updated to include new Statcast data!

Hitter xHR/FB calculator:

If you delete the formula in the highlighted cell, simply refresh this page.

Back to top

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

xOBA: Expected (Weighted) On-Base Average

Author: Andrew Perpetua
Original post URL: May 5, 2016
Source data: MLB/Baseball Savant’s Statcast data

Andrew’s xOBA uses a batted ball’s angle and velocity to determine the probability that each particular ball in play will go for a single, double, triple, home run or out. He does this for every ball in play for every hitter, and then he calculates his xOBA according to wOBA’s specifications.

Comprehensive xOBA data are included in Andrew’s xBABIP document, which updates daily.

Back to top

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

xK%: Expected Strikeout Rate

Author: Mike Podhorzer
Original post URL: May 13, 2013
Source data: FanGraphs’ plate discipline data, courtesy of Baseball Info Solutions (BIS)

Hitter xK% calculator:

If you delete the formula in the highlighted cell, simply refresh this page.

Back to top

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Pitcher Metrics

xHR/FB: Expected Home Runs per Fly Ball

Author: Mike Podhorzer
Original post URL: Jan. 8, 2015
Source data: Baseball Heat Maps

Absolute value of average angle (Abs Abv Angle) and standard deviation of distance (Std Dev Dist) are currently not publicly available.

Pitcher xHR/FB calculator:

If you delete the formula in the highlighted cell, simply refresh this page.

Back to top

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

xLOB%: Expected Left-on-Base Percentage

(LOB% is a primary driver in the difference between a pitcher’s FIP or xFIP and his ERA. Pitchers typically have little control over how their LOB% shakes out, although relievers tend to have higher LOB%s than starters.)

Author: Mike Podhorzer
Original post URL: Feb. 16, 2016
Source data: Traditional batted ball data, which can located at FanGraphs or Baseball Reference

Except for FB% and IFFB%, all cells should be inputted as raw counts. All data can be found at FanGraphs except for pickoff data, which Baseball Reference provides.

xLOB% calculator:

If you delete the formula in the highlighted cell, simply refresh this page.

Given the complexity of the equation, there’s a high probability that I botched the calculator. Let me know if something seems fishy.

Back to top

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

xK%: Expected Strikeout Rate

Author: Mike Podhorzer
Original post URL: 2.0: Jan. 16, 2017 (1.0: April 21, 2014
Source data: Baseball Reference’s pitching pitches data

Pitcher xK% calculator:

If you delete the formula in the highlighted cell, simply refresh this page.

Back to top

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

xBB%: Expected Walk Rate

Author: Alex Chamberlain
Original post URL: 2.0: Jan. 25, 2017 (1.0: Feb. 25, 2015)
Source data: Baseball Reference’s pitching pitches data

Here, K% represents strikeout rate, per usual — not to be confused with strike percentage (Str%).

xBB% calculator:

If you delete the formula in the highlighted cell, simply refresh this page.

Back to top



Print This Post

Currently investigating the relationship between pitcher effectiveness and beard density. Biased toward a nicely rolled baseball pant. Three-time FSWA finalist, one-time winner. Featured in this year's Lindy's Sports' Fantasy Baseball magazine. Doing everything I can to better understand (fantasy) baseball using only publicly available data.

newest oldest most voted
Dolemite
Member
Dolemite

xTRODINARY work!!! I am xTREMELY xCITED for this!