xWHIP and eFIP

All statistics current through May 27, 2011.

Sabermetricians have shown that once the ball is out of the hands of the pitcher, the pitcher has very little control over the outcome of the ball put into play.

DIPS theory tells us that a pitcher controls a few things about the outcome of an at-bat, each to a variable degree. The pitcher has the most control over the elements of the game that only involve himself, and omit others. A pitcher is in almost absolute control over the general location of his pitch. There are some marginal or unpredictable variants such as temperature, weather, wind speed and wind direction that factor into pitch location, but a pitcher who does not hit his spots can generally only blame himself. He either gripped the ball incorrectly, released it too late, did not properly adjust his throw for weather condition, “has the jitters,” just can’t pitch, etc. A pitcher likewise almost completely (though less completely than location) controls intentional walks, which essentially eliminate the batter from the equation of the outcome. Though players like Miguel Cabrera and Jeff Francouer come to mind as the rare player who interferes with the pitcher’s attempt to intentionally walk them, it is generally true that if a pitcher wants to intentionally walk a batter, it will happen.

Once the batter comes into the equation, the pitcher loses his control over the outcome. Once the pitch is released from his hands, he has done all he can do. It is then in the batter’s hand as to whether contact is made, where on the ball the contact is made, how hard the ball is hit, and whether or not it is pulled, amongst various other variables. In this regard, a pitcher has some, but not total, control over unintentional walks and strikeouts. The pitcher tries to fool the batter, but the batter may or may not buy the bait. The fielders are irrelevant before the ball is put into play, so the outcome is largely dominated by an exercise in game theory between the batter and pitcher.

A pitcher also controls the tendency of the ball to be on the ground or in the air. As noted above, once the ball is released from the pitcher’s hand, what happens to it is ultimately a question of what the batter does. A pitcher can throw the ball with heavy sink to induce ground balls, or throw it high in the zone to induce a popup, but the hitter, not the pitcher, ultimately controls the angle of trajectory, the force of contact, and the direction of the ball off the bat. Once contact is made, the ball is either put in play, in which case the pitcher’s fielders have the ultimate control over what happens next, or the ball is foul, in which case the batter-pitcher game repeats, or the ball is a home run, over which the pitcher has some, but not ultimate, control. (This is the theory behind xFIP and using a normalized home run rate in lieu of the pitcher’s actual home run rate as traditionally used in FIP.)

BABIP research by ball in play type out indicates that league fielding per batted ball type tends to be relatively stable. It tends to fluctuate annually, but only slightly and negligibly. For example, the expected hits rate on ground balls in play between 2004 and 2008 was .239. From 2008 to 2010, it was .236. So far this year, it is .238. The same is true for infield fly balls, line drives and outfield fly balls (though the latter two tend to fluctuate more, which is probably the result of scorer bias*). It is also true that line drive rate seems to remain relatively stable and out of pitcher control as well in the long run. Only a handful of pitchers have cumulative line drive rates that are not between 18 and 20 percent over the past five years, and most of those pitchers tend to be extreme batted-ball players. Even in the outlier, however, no pitcher has a line drive rate below 16 percent or above 22 percent. Noting this, you can probably say that Mat Latos’ 9.2 percent line drive rate and Travis Woods’ 24.5 percent line drive rate on the season are either the result of bad luck or funky scoring and that we should expect such to persist in the future.

*Note: scorer bias might make one skeptical of batted ball-based evaluation/prediction tools, but it is important to note that I am not, nor are most, preaching a black-and-white bible of truth with sabermetrics and sabermetric tools, but rather commenting upon the tendency of outcome or a rough baseline from which to make better, more informed decisions. Tools like tRA and the xWHIP Calculator are hardly perfect, but they lead to more informed analysis and decision-making.

From this research, I stood on the shoulders of men much smarter than myself and created the Expected WHIP (xWHIP) Calculator. (You can download the beta version for xWHIP3 by clicking here.) In case you are not familiar with how the xWHIP Calculator works, let me give you the quick rundown of how to use it and what it does. Refer to the picture of the beta of version 3.0 below (note: the 2008-2010 environment is loaded in the hits/outs created field; I do not have the runs created data for 2008-2010 to provide at this time).

The first and only manual step is data entry. Begin by entering data into the gray cells by using the player’s page on Fangraphs. The xWHIP Calculator is calibrated to Baseball Info Solutions (BIS) batted ball data, and, to avoid unnecessary scorer bias through consistency, you need to enter BIS data, which is what is available on Fangraphs. You can also change the pink cells of “specialized data points,” but will likely require information that is not publicly available to properly modify such. Hence, you should probably leave them untouched (well, unless you want to use a player’s career home run per outfield flyball rate in lieu of the league average mark*).

*If you modify the HR/OFFB% cell, do not change the park factor cell, or you risk double counting.

Once you’ve entered the data, the rest is all automatic, courtesy of my tireless hours of work in creating the xWHIP Calculator. My xWHIP formula first adds up all the batted ball data, and then normalizes it based on a regressed line drive rate. I use 19 percent, or about the league average. Then, with my new “normalized” batted ball distribution, I multiply each ball in play form by its expected hits rate. This gives you expected hits. The calculator also calculates expected home runs based on the normalized data for outfield flies and line drives.

Once a normalized batted ball distribution is created, the xWHIP Calculator also calculates an expected innings total (xIP). I calculate expected innings because actual innings pitched, like hits, is a function not only of player skill, but fielder interference/assistance and other statistical noise. A great or poor play is the difference between an extra batter faced and the end of the inning*. Expected innings is calculated by multiplying events by expected outs created by event. For example, a ground ball put in to play tends to result in 0.808 outs per occurrence. Caught stealing and pick off rates are something that varies by catcher and pitcher, but calculating such to be effectively utilized is something that I am not properly equipped to do. Hence, I use a league-average rate of .02 outs created per base runner to account for expected pickoffs and players who might get caught stealing. This formula gives me expected outs, which I then divide by three.

*This is why K% is more stable, indicative and, as a predictive/evaluational tool, valuable than K/9.

Using expected hits, expected home runs, and expected innings, as well as the other calculated data, we get a few valuable output points from the xWHIP Calculator. The primary purpose of the calculator and its calculations is to give a pitcher’s expected WHIP. xWHIP may not be important from a purely sabermetrics standpoint, but fantasy baseball players find it quite useful. xWHIP is calculated in three ways. First, xWHIP1 calculates a pitcher’s expected WHIP using actual innings pitched. Second, xWHIP2 calculates a pitcher’s expected WHIP using expected innings (xIP). Finally, quick xWHIP, or qxWHIP, calculates a player’s expected WHIP based purely on a player’s actual innings pitched, strikeout total, and WHIP. qxWHIP was created by Alex Hambrick, and the theory behind it is explained here.

The xWHIP Calculator also has a quick-and-dirty defensive adjustment for pitchers that converts a team’s defensive results into an expected “hits saved” compared to the hypothetical “league-average defense” per inning. This defensive adjustment has severe limitations (defense is hardly uniform infield-to-outfield, or player-to-player), and is optional, but it gives some sense of how a team’s overall defense can be roughly expected to affect a player’s “true talent” line.

In addition to xWHIP, however, the new versions of my xWHIP Calculator also tabulate two mainstream ERA estimators using normalized data. The first ERA estimator is eFIP, which is based on xFIP. xFIP is traditionally calculated by subtracting two times a pitcher’s strikeout total by the sum of three times a pitcher’s walk total plus 13 times .105 times that player’s flyball total, all divided by innings pitched. The resulting figure is then added to some constant, usually 3.2, to scale xFIP to look like ERA. xFIP tends to be my ERA estimator of choice, but I have several problems with the popular version of the formula. First, it uses a pitcher’s flyball total to calculate expected home runs. Flyball total is a composite of outfield fly balls and infield fly balls. As popups can never be home runs, it is silly to include them. In addition, home-run-per-outfield-fly-ball rates tend to be more stable over the long term than home-run-per-fly-ball rates.

Mental Health and the CBA
A particular bit of language in the latest CBA could have negative consequences for some players.

Second, and perhaps this is offset by including popups in the traditional expected home run formula, xFIP does not account for line-drive home runs. Line-drive home runs are few and far between, but they do occur a few times per 100 hits that are scored as line drives.

Third, xFIP is tabulated irrespective of expected flyball or outfield flyball rate. Pitchers, as noted above, do not seem to have much control over line-drive rate. If a pitcher, particularly in smaller samples (which give you less valuable data outcomes), has an atypically low or high line-drive rate, then a pitcher’s xFIP is skewed accordingly. The difference is, at most, a couple of home runs, but, like my infield flyball grudge with traditional xFIP, why use it if you don’t have to?

Fourth, xFIP does not account for park factors. Each of the 30 major league parks has different park dimensions that uniquely affect home run totals. Petco and Busch Stadium affect pitcher’s home runs allowed totals radically different than do the parks of Chicago. Players only play about 50 percent of their games at home, so you need to modify park factors accordingly, but the difference in expected ERA between Busch Stadium and Coors Field is substantial enough that it requires accounting, though that causes the xFIP formula to further sacrifice simplicity.

xFIP is a nice formula because it is simple and easy to calculate. Normally, accounting for my gripes would sacrifice much of xFIP’s simplicity appeal. However, given all the calculations the xWHIP Calculator makes, calculating a modified expected FIP to correct for my gripes is simple. I term this modified xFIP formula “eFIP.”

In addition to eFIP, the newest versions of the xWHIP Calculator will also calculated batted ball normalized versions of tRA or tERA, which I have termed “EXTRA.” EXTRA is calculated the exact same way as tRA, but it uses the pitcher’s normalized, not actual, batted ball data as the inputs.

Now that you know the parameters, let’s look at some of the major league’s leaders and losers in xWHIP, eFIP, and EXTRA using 2011’s runs environment and statistics through May 27, 2011. You can download the data file by clicking here.

Before reviewing the data, take note of the following. First, the following calculations use major league outs/runs/hits numbers, not league-specific numbers, so American League pitchers will tend to fare worse than these numbers, while National League hitters will tend to perform better. Second, changes in a player’s strikeout (xWHIP2) or walk rate (xWHIP1, xWHIP2) would have an appreciable effect on a pitcher’s expected WHIP. Third, while only starting pitchers (pitchers with at least one game started) are included in my data file, with the exception of Zack Greinke, only starting pitchers with 30 or more expected innings are included in my leaderboard (136 starting pitchers qualify). Fourth, I am calculating WHIP with unintentional walks (BB-IBB+HBP, or uBB); uBB better evaluates a pitcher’s control and expected baserunners. Finally, the league average xWHIP and eFIP are 1.33 and 4.00, respectively. The actual current major league average WHIP and FIP are 1.31 and 3.95, respectively.

First, the WHIP under-performers to date (calculated using “actual WHIP” (see above) minus the mean of a pitcher’s xWHIP1 and xWHIP2):

Name                   xIP        aWHIP     xWHIP     dWHIP
Davies, Kyle           44.54      1.90      1.57      0.34
Lackey, John           42.57      1.88      1.63      0.25
Reyes, Jo-Jo           55.06      1.66      1.42      0.24
Arroyo, Bronson        67.25      1.52      1.30      0.21
Greinke, Zack          29.99      1.14      0.93      0.21
Capuano, Chris         57.87      1.48      1.28      0.20
Dempster, Ryan         66.79      1.56      1.36      0.19
Garza, Matt            59.54      1.35      1.16      0.19
Holland, Derek         61.03      1.58      1.39      0.18
Jackson, Edwin         70.82      1.50      1.33      0.17
Tillman, Chris         50.41      1.67      1.50      0.17
Scherzer, Max          67.28      1.48      1.32      0.17
Carpenter, Chris       73.60      1.43      1.28      0.15
Lee, Cliff             78.18      1.25      1.10      0.15
Myers, Brett           69.78      1.51      1.39      0.13
McDonald, James        54.30      1.52      1.40      0.12
Norris, Bud            68.42      1.30      1.18      0.11
Dickey, R.A.           61.18      1.60      1.49      0.11
Francis, Jeff          69.59      1.40      1.30      0.10
Wood, Travis           62.79      1.41      1.31      0.10
Hudson, Dan            72.45      1.33      1.23      0.10
Rodriguez, Wandy       67.84      1.32      1.22      0.10
Lilly, Ted             63.55      1.34      1.25      0.09
Duensing, Brian        54.92      1.44      1.36      0.09
Morrow, Brandon        40.47      1.38      1.30      0.08
Baker, Scott           61.52      1.30      1.22      0.07
Stauffer, Tim          65.87      1.32      1.25      0.07
Niese, Jon             62.10      1.44      1.38      0.06
Volstad, Chris         52.02      1.44      1.38      0.06
Danks, John            66.59      1.40      1.34      0.06
Harang, Aaron          61.74      1.32      1.26      0.06
Narveson, Chris        57.44      1.35      1.29      0.06

Much of this leaderboard is populated with under-inspiring pitchers who, while unlikely, have pitched pretty poorly this year and are hardly worth a spot on your bench. Case in point: the injured John Lackey and “immutable” Kyle Davies. A few names do stand out, however. I think the ship sailed on Ryan Dempster (whose numbers are infinitely better if you omit his 0.1 inning pitched disaster at Arizona) after his 11-strikeout performance on May 13, but maybe some owner has not been paying close enough attention this past month (e.g., people in college). We all know Cliff Lee and Matt Garza have had their share of bad luck this year, but what about Chris Carpenter and Zack Greinke? Greinke’s performance to date puts him in company with the top three guys in the league, but his 5.79 ERA has been ugly. If any owner is having second thoughts about the Royals ex-Ace, or is willing to deal him at market value, I’d strongly considering biting. And what about Bud Norris? I wrote about him last week, but his ownership rate is still below 50 percent (it actually went down a notch). I think a lot of people are overlooking just how good Norris has been this year. Jeff Francis and Travis Wood are a pair of pitchers who could help you in other categories without hurting your future WHIP.

Next, the WHIP over-performers to date (calculated using “actual WHIP” (see above) minus the mean of a pitcher’s xWHIP1 and xWHIP2):

Name                   xIP        aWHIP     xWHIP     dWHIP
Tomlin, Josh           60.26      0.93      1.24     -0.32
Lohse, Kyle            67.89      0.91      1.22     -0.30
Humber, Philip         56.67      0.98      1.28     -0.30
Britton, Zachary       59.14      1.12      1.38     -0.26
Ogando, Alexi          54.57      0.94      1.18     -0.25
Johnson, Josh          56.03      0.96      1.19     -0.23
Hudson, Tim            62.97      1.14      1.35     -0.22
Morton, Charlie        58.11      1.31      1.52     -0.21
Harrison, Matt         55.43      1.25      1.46     -0.20
Chacin, Jhoulys        64.78      1.10      1.30     -0.20
Maholm, Paul           66.89      1.16      1.36     -0.20
Beckett, Josh          58.97      0.98      1.18     -0.19
Penny, Brad            59.28      1.32      1.51     -0.19
Hochevar, Luke         70.45      1.23      1.41     -0.18
Liriano, Francisco     46.26      1.48      1.66     -0.18
Verlander, Justin      75.59      0.96      1.14     -0.17
Jurrjens, Jair         54.41      1.02      1.19     -0.17
McClellan, Kyle        61.16      1.21      1.38     -0.17
Haren, Dan             77.03      0.89      1.06     -0.16
Kennedy, Ian           71.84      1.10      1.24     -0.14
Coke, Phil             49.98      1.27      1.41     -0.14
Moseley, Dustin        57.56      1.23      1.37     -0.14
Burnett, A.J.          65.88      1.29      1.42     -0.14
Hanson, Tommy          62.63      1.09      1.22     -0.13
Carmona, Fausto        70.32      1.22      1.34     -0.13
Correia, Kevin         68.79      1.19      1.31     -0.12
Billingsley, Chad      67.93      1.26      1.37     -0.12
Cahill, Trevor         69.18      1.23      1.35     -0.11
Pineda, Michael        61.86      0.99      1.11     -0.11
Lincecum, Tim          74.38      1.05      1.16     -0.11
Hellickson, Jeremy     55.45      1.24      1.35     -0.11
Weaver, Jered          82.98      0.96      1.06     -0.10

Here we find a lot of players with BABIP-deflated ERAs who are on the Atkins diet when it comes to strikeouts: Kyle Lohse and Zach Britton’s combined strikeouts per nine rate (9.81) is equal to that of Bud Norris. Most of these “trailers” tend to be groundball pitchers because groundballs, while having a lower expected runs outcome per event, have a higher hits-resulting rate. A year of xWHIP has taught me that ERA and WHIP tend to be inversely related to groundball and flyball rates. Alexei Ogando throws hard, but can you really trust a flyball pitcher (64.5 percent AO%) in Texas (inflates home runs by 10 percent)? Ogando’s SwStr% (8.9 percent) indicates he is capable of slightly better than league-average strikeout totals. As you might notice not all players on this board are “bad” or have “bad” expected WHIPs (e.g., Josh Beckett). This is only a tool to help figuring out who has been under/over-performing, and an under-performer may very well be worth keeping.

Then we have the pitchers who are secretly better than their listed FIP.

Name               xIP      aFIP    EXFIP   dFIP
Karstens, Jeff     45.60    4.92    3.57    1.35
Volquez, Edinson   51.36    5.77    4.50    1.27
Arroyo, Bronson    67.25    5.48    4.33    1.15
Gorzelanny, Tom    52.31    5.28    4.14    1.14
Hochevar, Luke     70.45    5.46    4.35    1.11
Dempster, Ryan     66.79    4.80    3.78    1.03
Myers, Brett       69.78    5.44    4.49    0.95
Greinke, Zack      29.99    3.06    2.18    0.88
McDonald, James    54.30    5.11    4.29    0.82
Lester, Jon        68.85    4.25    3.48    0.77
Capuano, Chris     57.87    4.61    3.86    0.75
Bedard, Erik       51.52    4.32    3.68    0.65
O'Sullivan, Sean   50.48    6.26    5.61    0.64
Lewis, Colby       64.90    5.26    4.62    0.64
Latos, Mat         53.83    4.35    3.80    0.55
Buchholz, Clay     58.87    4.77    4.25    0.52
Romero, Ricky      63.02    3.98    3.48    0.50
Baker, Scott       61.52    4.19    3.70    0.49
Pelfrey, Mike      64.67    5.07    4.59    0.48
Carmona, Fausto    70.32    4.29    3.82    0.47
Colon, Bartolo     57.80    3.79    3.32    0.47
Blackburn, Nick    62.39    4.65    4.19    0.47
Norris, Bud        68.42    3.75    3.30    0.45
Chen, Bruce        41.62    5.12    4.68    0.45
Arrieta, Jake      60.54    4.66    4.23    0.43
Chacin, Jhoulys    64.78    3.96    3.56    0.40
Scherzer, Max      67.28    4.36    3.98    0.39
Lilly, Ted         63.55    4.67    4.28    0.39
Dickey, R.A.       61.18    4.75    4.37    0.38
Gallardo, Yovani   68.64    4.27    3.91    0.36
Litsch, Jesse      46.74    4.69    4.35    0.34
Kuroda, Hiroki     70.70    4.15    3.81    0.34
Rodriguez, Wandy   67.84    4.05    3.71    0.33
Liriano, Francis   46.26    5.44    5.11    0.33
Volstad, Chris     52.02    4.24    3.91    0.33

If there is any pitcher to avoid on this list, it’s Edinson Volquez. I took a lot of flack being a vocal Volquez hater this offseason, and while it’s only been 51 innings, I really want to say “I told you so” about how bad his control was going to burn him this year. Volquez has a 4.16 xFIP, so a lot of people might be tempted to buy, but even if you tinker with his batted ball distribution a bit, his expected FIP is putrid. A 4.50 FIP would be “average” by standards two or three years ago, but in the new era of the pitcher, it’s trade-or-cut material. Ryan Dempster’s a name on this list I really like, but, as noted above, the ship has probably sailed on him by now. Same goes with Erik Bedard, who has been lights out over his past five or so turns. And what about Bartolo Colon? Is he the real deal after injecting cheeseburgers from his belly into his elbow? No matter which you choose, all the metrics seem to check out (3.77 ERA, 1.20 WHIP, 3.61 FIP, 2.90 xFIP, 3.86 eFIP, 1.20 WHIP, 1.28 xWHIP), but something does not smell right. A 5.9 percent SwStr% ties for his second-lowest mark since 2002 and is well below his post-2002 average of 7.6 percent, but his strikeout rate (23.6 percent) is a career second-best at age 38? I’d use the “it checks out” line to hedge your risk.

What’s up with Jeff Karstens? He’s been good on the surface (3.57 ERA, 1.28), but regular FIP says look out (4.70). Karsten’s improved strikeout rate (18.9 percent this season, 12.2 percent career) makes sense if you look at batters’ swing-and-miss rate against him (9.0 percent this year, 7.1 percent career, 8.4 percent major league average), but what is causing it? It’s not his velocity (88.4 MPH fastball this year, 88.5 career) or pitch usage (none of his four usage rates varies by more than a few percent points this season). His change-up has been wicked awesome, but both his fastball and slider (thrown almost a combined three-fourths of the time) have fared poorly both this year and for his career. Tread at your own caution.

Chris Capuano, on the other hand, has been secretly good for the Mets, even if the results do not say so. His ERA (4.94) and WHIP (1.45) have been atrocious, but his peripherals (3.86 eFIP, 1.28 xWHIP, 7.74 K/9, 19.4 percent K%) say this waiver wire fodder (2 percent Yahoo ownership) might be worth a careful look.

And the guys whose FIPs are not telling the whole story. Keep in mind that in the second “year of the pitcher,” ERAs are not what they used to seem.

Name               xIP      aFIP    EXFIP  dFIP
Bergesen, Brad     51.38    3.90    5.15   -1.26
Morrow, Brandon    40.47    2.53    3.63   -1.09
Tillman, Chris     50.41    3.85    4.89   -1.04
McCarthy, Brando   62.85    2.67    3.60   -0.93
Jurrjens, Jair     54.41    2.97    3.85   -0.88
Coke, Phil         49.98    3.83    4.65   -0.81
Hudson, Dan        72.45    2.88    3.66   -0.78
Buehrle, Mark      72.50    3.82    4.59   -0.76
Zimmermann, Jord   60.00    2.98    3.73   -0.75
Sabathia, CC       77.83    2.99    3.70   -0.71
Fister, Doug       63.85    3.57    4.27   -0.70
Garza, Matt        59.54    2.01    2.71   -0.69
Lohse, Kyle        67.89    3.23    3.91   -0.68
Bumgarner, Madis   57.50    3.16    3.84   -0.67
Zambrano, Carlos   66.71    3.94    4.57   -0.62
Humber, Philip     56.67    3.77    4.35   -0.59
Hernandez, Livan   68.75    3.94    4.47   -0.53
Halladay, Roy      86.09    1.93    2.46   -0.53
Billingsley, Cha   67.93    3.29    3.82   -0.53
Johnson, Josh      56.03    2.72    3.24   -0.52
Masterson, Justi   64.33    3.23    3.72   -0.49
Marquis, Jason     63.53    3.80    4.27   -0.47
Weaver, Jered      82.98    2.74    3.21   -0.46
Oswalt, Roy        44.97    3.29    3.73   -0.44
Kennedy, Ian       71.84    3.41    3.84   -0.42
Beckett, Josh      58.97    3.03    3.45   -0.42
Pineda, Michael    61.86    2.84    3.25   -0.42
Morton, Charlie    58.11    3.91    4.29   -0.38
Reyes, Jo-Jo       55.06    4.25    4.61   -0.36
Maholm, Paul       66.89    3.60    3.96   -0.36
Chatwood, Tyler    53.24    5.09    5.45   -0.36
Nova, Ivan         54.62    4.61    4.96   -0.36

My mother always told me to never trust Brandon Morrow. As I noted last week, you’re better off selling him at cost to another saber-friendly owner and investing the funds elsewhere. Jordan Zimmerman is much better than he’s been or his presence here indicates, and I would sit tight with him. Is Justin Masterson finally putting it all together as a post-hype sleeper? 3.61 xFIP versus lefties (165 batters) and 3.26 xFIP versus righties (110 batters). Sorry Orioles fans clinging to old Bedard jerseys; Chris Tillman is not the stud or the sleeper we thought he was. Ditto on Brad Bergesen, who I once had a fantasy man crush on several years back. After 10 years, you should not be fooled by Jason Marquis. He tends to start things off well with new teams, but it always ends badly. Has Doug Fister been ol’ reliable for you thus far? Don’t expect it to persist, as he’s more likely to take his hand and slap your fantasy team with it in the future. I’ve shaken off my preseason (Phil) Coke addiction, and what of former top Twins draftee Phil Humber? 2.85 ERA, 3.77 FIP looks nice for something you plucked off the waiver wire for a stream that never seemed to end in a drop, but lackluster strikeouts plus league-average WHIP plus poor ERA prospects equal trade toss in to get a better deal done. Finally, Livan Hernandez is not even worth mentioning.

Next time out (this upcoming Monday), we’ll look at EXTRA, actual ERA and actual tRA to date. Until then, as always, leave your love/hate in the comments below.

Print This Post
Jeffrey Gross is an attorney who periodically moonlights as a (fantasy) baseball analyst. He also responsibly enjoys tasty adult beverages. You can read about those adventures at his blog and/or follow him on Twitter @saBEERmetrics.
Sort by:   newest | oldest | most voted
Will Hatheway
Will Hatheway
Jeffrey – Cahill was a bargain this year because of his projected luck last year. Now his babip is still low, if not insane, but his lob has gone from lucky-if-sustainable to crazy. Either way, not only does it seem that his GB% in general can help here, not to mention park and defense, but do you think that the extreme drop of his two major weapons (2-seamer and change fall 4 inches more than average) has something to do with his apparent luck (in which case perhaps it is somewhat sustainable)? I don’t know if there is anything to… Read more »
jeffrey gross
jeffrey gross

Keep in mind that in the 2011 run environment (as well as the 2010 run environment), hits/runs/outs are “on a different scale” then they used to be. There are 40+ pitchers with 1+ starts with ERAs under 3.00, and the league average ERA/FIP/xFIP are half to three-quarters a run lower than it was 2002-2009. Same with WHIP. A 1.31 xWHIP is about league average here.

jeffrey gross
jeffrey gross

You can catch more on xWHIP and my take on some buy low pitchers by clicking here: http://bit.ly/kgWVT9. That link is a recording of my Monday appearance on Fantasy Phenoms’ online radio show. I’ll be back on this upcoming monday as well.

jeffrey gross
jeffrey gross
@Will Cahill’s 76.5% LOB% last season was a bit high last season (~73% MLB Average), but not unsustainably so. I’d agree that his current 85% mark is not sustainable, and that he’s in line for some ERA regression heading forward, but he’s a solid top 24-36 pitcher (SP3). Cahill’s not really as much of a strikeout artist as he was in his first handful of starts (even though he was in the minors, Cahill apparently changed his delivery or something according to what I’ve read to be more control/groundball oriented pitcher). Nonetheless, Cahill does have average control, average whiff stuff,… Read more »
Dave Studeman
Dave Studeman

Sabermetricians have shown that once the ball is in play, the outcome of the at-bat, unless the result is a home run, is largely out of the hands of the batter.

I don’t think this is true.  It’s largely true for pitchers, but batters have a significant impact on their BABIP, particularly over several seasons’ data.

Jeffrey Gross
Jeffrey Gross


My statements on xWHIP on the Fantasy Phenoms show might better explain what I meant to say, but I 100% agree with your statement. All that I am trying to convey is that once the ball is put into play by the batter, a lot of what happens next is up to him. For the pitcher, once the ball is out of the hand, it is then up to the batter and fielder for the ultimate outcome.

Jeffrey Gross
Jeffrey Gross

*a lot of what happens is up to the fielder

Jeffrey Gross
Jeffrey Gross

The hitter controls the directionality, strength of contact and angle of trajectory as much as he can, but it’s the fielding that either converts a ball in play into an out or not. The batter does what he can to avoid it, but 70+% of the the time, usually, the fielding says “no”

Jeffrey Gross
Jeffrey Gross

Maybe I should change that sentence to read:

“In rejecting batting average as baseball’s most popular metric, Sabermetricians have shown that once the ball is in play, the outcome of the at-bat, unless the result is a home run, is substantially less within the control of the batter than once believed”

Dave Studeman
Dave Studeman

The batter does what he can to avoid it, but 70+% of the the time, usually, the fielding says “no”

True, but that’s irrelevant to the question of “control.” Control can best be identified through sustained variance against the average.

I don’t mean to nitpick, but I don’t agree with your last sentence either.  Sabermetricians “rejected” batting average because they felt that OBP and SLG were better indicators of offensive worth.

I think you’re best served by just focusing on the pitching side of your equations, cause I think you’re misrepresenting the batting side.

Jeffrey Gross
Jeffrey Gross

Yeah, I think I’m just going to delete my intro paragraph here….

Hommy Tanson
Hommy Tanson

Is HR/OFFB% park factor data publicly available anywhere?

Jeffrey Gross
Jeffrey Gross

@Hommy Tanson,

Sadly, I do not believe so. I have private HR/FB% index factors that I use, though that can be misleading in parks like Oakland or Wrigley that have a lot or very little foul territory.

Jeffrey Gross
Jeffrey Gross

There is this, but it’s getting stale each season:

Hommy Tanson
Hommy Tanson

Thanks! Very interesting article by the way. It seems very likely that eFIP will prove an even better predictor of ERA than xFIP. One more thing, which I can probably figure out with a little thinking of my own: using the spreadsheet, how would you go about computing a pitcher’s home eFIP? Would you simply multiply the park factor by two?

Hommy Tanson
Hommy Tanson

Sorry, subtract 1 from the park factor, multiply by 2, add 1?

Jeffrey Gross
Jeffrey Gross

@Hommy Tanoson,

Thanks. The calc automatically calculates total eFIP. Park factor uses a 1/2 step because only half the expected games are at home, neutral road.
(1-((HR_PF-1)/2+1) is how you apply it