Notice: Trying to get property 'display_name' of non-object in /var/www/html/wp-content/plugins/wordpress-seo/src/generators/schema/article.php on line 52
keyboard_arrow_uptop

This is admittedly something of a data dump article, but judging by some of the Bat Signal questions we get, I think it’s a useful one to consolidate everything in one place. Plus, it’s another opportunity to take our awesome, shiny new toy out for a test drive, and that’s always fun.

For those uninitiated, I highly recommend you take a few minutes to get initiated. Our statistical wizardry department, led by Harry Pavlidis, Jonathan Judge, Robert Arthur, Dan Turkenkopf, and Gregory J. Matthews, created two extremely useful metrics for evaluating pitchers in cFIP and DRA. Here’s a link to the original piece unveiling those two beauties, along with the follow-up piece introducing the subsidiary stat, DRA-, which we’ll be talking about today. Both should be required reading in your summer school class.

For our purposes in evaluating “buy” and “sell” targets for fantasy purposes, the important thing to understand is that DRA as a raw metric assigns “only assigns the runs that a pitcher most likely deserved to be charged with” to the pitcher’s tab. It’s a look at the actual performance of a pitcher to date. DRA- standardizes that performance around a standard scale of 100, where performances below 100 are better than average, and the farther above 100 you wander, the farther below league average you’ve performed.

Where DRA- incorporates earned and unearned luck into the picture, cFIP is the most accurate measure unveiled to date for evaluating “true talent” performance independent of factors beyond a pitcher’s control. DRA-, in other words, is how we evaluate the earned results to date, while cFIP is how we project the likelihood of a pitcher’s future performance.

I would be remiss if I didn’t include a disclaimer that comparing these metrics isn’t exactly an apples-to-apples exercise on account of their modeling techniques (DRA- does not utilize a forced standard deviation while cFIP does), but for what we’re doing here they are more than close enough for jazz.

So let’s take a look at some of the pitchers who see the most significant deviation between their DRA- and cFIP numbers. First, here’s the top ten pitchers who’ve posted the biggest discrepancy between their DRA- and cFIP thus far in 2015. I’ve filtered the data set to a minimum of 40 innings pitched thus far to keep relievers out of the club, and then I set the cut off for both sides of the coin at league-average (100 cFIP) projection.

Greener Pastures Ahead

NAME

IP

ERA

FIP

cFIP

DRA

DRA-

cFIP – DRA-

Stephen Strasburg

45.3

6.55

4.00

100

7.84

190

-90

Travis Wood

47.7

4.72

4.71

91

5.44

148

-57

CC Sabathia

77

5.38

4.31

91

5.96

141

-50

Mat Latos

49.7

5.44

3.12

97

5.46

145

-48

Jon Lester

79

3.99

3.55

95

5.44

139

-44

Anibal Sanchez

82

5.16

4.23

95

4.68

129

-34

James Shields

87.7

3.59

4.17

85

4.87

118

-33

Jorge De La Rosa

47.7

4.91

3.75

93

4.12

121

-28

Gio Gonzalez

71.3

4.42

3.11

96

4.72

123

-27

Matt Shoemaker

68.7

4.85

4.41

95

4.7

121

-26

Strasburg is probably the biggest question mark in all of fantasy land thus far in 2015. His stuff appears intact, but between nagging injuries, bizarre command regression, and a changeup that has more or less disappeared from utility, Strasburg’s been an unmitigated disaster. Drafted an average of 24th overall as the fourth pitcher off the board, his present value checks in 238th among starting pitchers and 887th overall per ESPN’s player rater. The good news, as such, is that cFIP sees a pitcher with a league-average performer’s kit whenever he makes his way back off the DL. That’s still nowhere near the level of elite production managers paid for on draft day, but it would be a marked improvement upon the tire fire of a season he’s posted so far.

CC Sabathia is an interesting name. He continues to get killed by the longball, as late-career CC Sabathia is wont to do, but his underlying peripherals actually suggest a pitcher deserving of a much better fate than his topline numbers have suggested thus far. His 91 cFIP is tied with fellow old-timer A.J. Burnett at 25th among starters, and while the home park isn’t likely to aid him in mitigating his gopheritis, I like him as a relatively cheap target to acquire for above-average production in the latter half of your rotation.

Speaking of the longball, I wrote a little while ago about James Shields’ struggle with addiction to it, and that unfortunate reality has continued. His 1.64 homers-per-nine is far and away the highest of his career and the first time since 2010 he’s allowed even one-per-nine. It’s been an odd feature of an otherwise outstanding season, especially given his home park (where he’s allowed half of his 16 bombs). He’s a top-20 pitcher by cFIP, though.

Gio Gonzalez and Matt Shoemaker represent a couple of my favorite “buy-low” guys for the middle of your rotation. I wrote up Gio at greater length in last week’s Starting Pitcher Planner, and the moral of the story was that he’s a rediscovered changeup away from moving right back into the range of performance you’d expect from him. Shoemaker, meanwhile, has been turning more to his excellent splitter lately, which is good news for his whiff rate. Add in a stabilizing home-run rate, and the window to acquire his services on the relative cheap may be closing.

Danger, Will Robinson

NAME

IP

ERA

FIP

cFIP

DRA

DRA-

cFIP – DRA-

Shelby Miller

84.7

2.02

3.38

104

1.86

37

67

Chris Young

52

2.25

3.61

101

1.76

51

50

Michael Wacha

77

2.45

3.49

100

2.29

52

48

Mike Pelfrey

70.7

3.18

4.07

118

3.63

70

48

Aaron Harang

83

3.04

3.84

105

2.87

59

46

Mike Bolsinger

48

2.25

3.02

102

2.33

60

42

Adam Warren

69

3.78

4.5

121

3.29

79

42

Anthony DeSclafani

77.7

3.36

3.64

112

3.19

73

39

Jordan Lyles

49

5.14

3.79

113

3.47

77

36

Jordan Zimmermann

77

3.74

3.08

110

4.01

76

34

Ah. Shelby Miller. Way back in October I talked about Miller’s introduction of a two-seamer down the stretch, noting it had some game-changing potential given its similar velocity to his four-seamer but sharply different movement. I was skeptical, however, noting that he hadn’t shown much comfort with the pitch and that his curveball had gotten slower and loopier to the point where he didn’t have any great secondary offerings to back up the gas. Well, fast-forward to the first two-plus months of 2015, and the winker’s become his second-most popular pitch, while he’s worked in his cutter with much greater frequency as well. He’s made some very legitimate improvements to get where he is. That said, he’s getting lucky on balls in play even adjusting for his groundball tendencies, his homerun rate is half his career mark, and he’s still not striking out a ton of guys despite marginal improvement in his swinging strike rate. Managers shouldn’t expect him to fall off the table, and neither does cFIP, but he looks more like a slightly below-average starter going forward than the top-15 option he’s produced as thus far.

As if fantasy managers needs confirmation to be skeptical of Mike Pelfrey’s run this year, he projects to one of the larger swoons on this list. His tire fire of a start last time out, while extreme, is much closer to what cFIP projects going forward than the 2.28 ERA he’d been rocking over the 11 starts leading up to it. The Twins are one of the least-shifting teams in baseball and have one of the worst defensive right-sides in baseball by FRAA, so Pelfrey’s better-than-average BABIP on grounders is an unlikely aberration to date. Given how few whiffs he generates that’s a bunch of bad news for his future prospects if more of those grounders start finding grass instead of gloves.

On the one hand I’m inclined to say numbers be damned, ride Mike Bolsinger! He’s made some notable mechanical tweaks to get more direct to the plate, adding extension and generating better plane on his pitches over the last month and change. I’m glad cFIP doesn’t project a scenario where Bolsinger falls off the table, as I think he should be able to produce average-ish numbers going forward and I like him as a deep league rotation piece.

On the other hand, if you take my feelings on Bolsinger and turn them exactly upside down, that’s what I see when I look at Adam Warren. He ran off one of the ugliest (and, as a bitter Red Sox fan, most obnoxious) stretches of five consecutive quality starts you’re ever going to see before failing to make it out of the fifth in his last turn. He’s had to adjust and move away from his four-seamer lately because it was getting pounded, and we’ll see if the adjustments he’s made (more sinkers and sliders) yields fruit. I’m not buying the profile, though.

It’s been a strange trip for Jordan Zimmermann owners so far this year, and the bad news is that our projection system sees a bumpy road ahead. He’s given back all the whiffs he gained last year and then some, which makes at least some intuitive sense because those gains were tied to an increase in whiffs generated with a four-seam fastball he threw 70 percent of the time last year, and that made for an extremely volatile profile. He’s lost some hop on that pitch, and it’s driven a loss of three percentage points off his aggregate whiff rate. He’s no longer getting ahead at an elite rate, and his usually-inflated line drive rate is inflated all the more. He looks on the surface like a buy-low guy, in that his performance hasn’t matched his draft price and by some standard metrics like BABIP and strand rate it looks like he may be getting lucky. But his pitch performance and batted ball data suggest that’s not necessarily the case, and I’d avoid overspending to acquire his services with expectations that he’ll revert to top-25 form.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
timber
6/16
Jordan Lyles had season-ending surgery earlier this month, so I would imagine that most owners have already dumped him.
markpadden
6/16
Is there any evidence that cFIP is a better rest-of-season ERA predictor than DRA? The original article shows that DRA is actually more accurate (in predicting future runs) than FIP for all sample sizes tested. I don't see where anyone has tested DRA vs. cFIP, however.
BuckarooBanzai
6/16
Yep. Rather than try to parse it out for you I'll instead point you to Jonathan Judge's primer at Hardball Times:

http://www.hardballtimes.com/fip-in-context/
markpadden
6/18
Thanks. Frankly, I'm not impressed with cFIP for a whole host of reasons (http://www.hardballtimes.com/fip-in-context/#comment-61647). The author simply showed that it correlates y2y with *itself* much better than other metrics. He never performed a test to measure avg. error vs. rest-of-season ERA, which is what fantasy owners care the most about. The one in-season test he did do (correl. of each metric to RoS RE24/PA) showed cFIP with essentially no improvement over something like SIERA -- which is troubling since his model was created (overfit, IMO) using the same data he used for testing, while SIERA et al. did not have that same advantage (for years 2011-2014).

Most importantly, he never tested cFIP vs. DRA. So it just seems like a curious choice to assume that cFIP-DRA will reveal the best buy-low or sell-high fantasy pitchers, until/unless more work has been done to show that cFIP actually predicts ERA accurately out of sample.
johnjmaier
6/18
I am not saying he is an ace but doesn't Mike Fiers fall into the greener pastures camp? Unless I am misunderstanding some concepts he should be in the top 10 no? Is Fiers at least a #3 going forward? Granted it would help if the Brewers were better fielders.
BuckarooBanzai
6/19
Fiers definitely has some separation, as his results have been below-average to date and cFIP has long projected him to be a great play on account of the K/IP and K:BB components. I'm bearish. I just recently wrote him up in a Starting Pitcher Planner, and I'll refer you to that, as not much has changed since:

"...his 18.2 pitches-per-inning is tied for the least efficient rate among all starters, and coupled with the 29th-ranked offense by TAv “supporting” him, he’s a genuinely poor play in win-based leagues. He’s also given up more than his fair share of hits, and opposing hitters have earned their bases by making the hardest contact they’ve made against any starting pitcher in baseball thus far. It’s been enough to offset a walk rate that has whittled away to next-to-nothing over his past few starts, but he’s a WHIP liability as well."

He's also poor at holding baserunners. The whiffs are nice, and keep him relevant in most formats. But I wouldn't overpay for him as a buy-low or anything.

http://www.baseballprospectus.com/article.php?articleid=26487
johnjmaier
6/26
Thanks for the reply.
irishguy8
8/11
8/11 Update:
On the "buy" guys"
- Strasburg has been great, when healthy.
- Latos has been good too, despite his last start.
- Lester has been pretty good. And he struck out 14 the other day.
- Gio has been great.
- Shoemaker has been good until his last start.
- Wood, who cares.
- CC has maintained his bad levels.
- Sanchez has been bad.
- Shields has been off and on, better lately.
- De La Rossa has been off and on, with some stinkers thrown in.

So it's 50/50 I guess. I haven't looked at the Danger guys.