Phillies kwERA

Mike Fast put up a great article at The Hardball Times, highlighting the leaders in kwERA. What is kwERA? Fast explains:

A few years ago, Guy M and Tom Tango came up with a neat toy called kwERA that works a little bit like FIP, except that it ignores a pitcher’s home run rate. It’s a measure of strike zone dominance.

The formula is kwERA = [5.30] – 12 * (K-BB)/PA.

I hadn’t heard of kwERA before, so I tossed some numbers into a spreadsheet to see what the Phillies looked like.

Note that I used TBF (total batters faced) as opposed to the PA in Fast’s column. It didn’t change anything, but since TBF is one of the columns, it makes it easier to follow. Also note that I used the full season stats for Roy Oswalt and J.A. Happ, including their starts with their new teams. Lastly, Fast initially used 5.40 as a constant, but notes in the comments that it is league-dependent. As such, the constant I used instead is 5.30.

Ryan Madson 4.57 21.7 93 4 25 0.04 0.27 0.23 2.59
Roy Halladay 2.17 178.0 705 21 158 0.03 0.22 0.19 2.97
Jose Contreras 3.55 38.0 159 12 40 0.08 0.25 0.18 3.19
Roy Oswalt 3.53 135.0 549 36 124 0.07 0.23 0.16 3.38
Cole Hamels 3.56 139.0 586 46 138 0.08 0.24 0.16 3.42
Brad Lidge 5.09 23.0 103 14 28 0.14 0.27 0.14 3.67
Chad Durbin 3.43 44.7 189 15 40 0.08 0.21 0.13 3.71
Average 3.97 67.7 283 19 53 0.07 0.19 0.12 3.89
Joe Blanton 5.86 106.0 469 26 72 0.06 0.15 0.10 4.12
Jamie Moyer 4.84 111.7 460 20 63 0.04 0.14 0.09 4.18
Antonio Bastardo 5.11 12.3 54 9 13 0.17 0.24 0.07 4.41
Nelson Figueroa 3.46 26.0 104 9 15 0.09 0.14 0.06 4.61
Kyle Kendrick 4.37 127.7 534 34 64 0.06 0.12 0.06 4.63
David Herndon 4.08 35.3 156 10 16 0.06 0.10 0.04 4.84
Danys Baez 4.95 36.3 167 20 20 0.12 0.12 0.00 5.30
J.C. Romero 2.59 24.3 111 20 17 0.18 0.15 -0.03 5.62
J.A. Happ 4.03 22.3 82 15 10 0.18 0.12 -0.06 6.03

For you visual learners (click to open a larger version):

Those pesky Sabermetrics keep finding ways to make Madson look good and Happ bad! After all, Happ is a “Cliff Lee lite” and Madson doesn’t have a closer’s mentality.

Overall, though, the findings shouldn’t really surprise you. ¬†Halladay and Oswalt are at the top and Baez and Romero are at the bottom. Lidge may be the other surprise given that his control (5.5 BB/9) has been such an issue, but he averages more than a strikeout per inning pitched. Additionally, kwERA ignores home runs and Lidge has given up five of them in 23 innings of work and his HR/FB is at a whopping 18%.

kwERA doesn’t differ a great deal from other ERA estimators such as SIERA, as the differential ranges from -0.27 (Kendrick, 4.90 SIERA) to +0.46 (Figueroa, 4.15 SIERA). But I still thought it was interesting to look at — hopefully you did, too.

(inb4 “kwERA? More like kwEERA! LoLz!1!1!!!”)

Leave a Reply


This site uses Akismet to reduce spam. Learn how your comment data is processed.


  1. Tangotiger

    August 05, 2010 08:54 AM

    PA is plate appearances. TBF is total batters faced. PA = TBF. Some of us think it’s silly to have two different names for the exact same thing. I prefer to stick to PA.

    In any case, ideally, the denominator should be PA – IBB – SH. But of course, that’s barely going to change anything in almost all cases.

  2. Drew

    August 05, 2010 09:07 AM

    So Lidge doesn’t suck? I feel better now, thanks! You need to do more graphs that make Lidge look better than he is.

  3. Aaron H

    August 05, 2010 09:28 AM

    Drew, I think what this stat (and other advanced pitching metrics like SIERA, xFIP) is trying to show is how the pitcher is truly performing when results that are beyond their control (such as luck with balls hit in play, HR per flyball rate, etc.) are taken out, leaving only what they can truly control (strikeouts, walks, and ground balls). In Lidge’s case, this still shows that he has a big area for improvement, since his walk rate is so high, but that he’s still pitching decently-he just has an unlucky HR rate now.

    In another case-if you looked at Cole’s ERA from last year and this year, you’d say he sucked in ’09 and is good in ’10. However, I think his kwERA, which would ignore his bad luck in ’09, would show that he’s essentially the same, talented pitcher that we’re seeing right now.

  4. Aaron H

    August 05, 2010 09:30 AM

    But of course, this doesn’t prove that Lidge should still be the closer-Madson is clearly a better pitcher than him, no matter what the “closer mentality” advocates say.

  5. Tangotiger

    August 05, 2010 09:42 AM

    I would not say that this stat says anything about how lucky or unlucky, or skilled or unskilled, Lidge is with the HR.

    The ONLY thing this stat says is that if you focus only on BB, K, and HBP, that Lidge is doing alright.

    This is no different than looking at someone who has an OBP of .400, but those being all singles, and saying the hitter is doing pretty good. He’s just being pretty good at one part of being a ballplayer.

    Same with Lidge. He’s done ok with his K and BB. It takes no position on how he’s doing on HR.

  6. Rob

    August 05, 2010 10:59 AM

    In what way is this stat more helpful than a k/bb?

  7. Tangotiger

    August 05, 2010 11:45 AM

    Differential is better than ratio. I demonstrated on my blog a few months ago.

  8. KH

    August 05, 2010 11:46 AM

    Bill, sometimes you are quite merciless lol. Its better to just ignore the poor people that think Jay Happ is anything but a 4th or 5th starter and that Madson could never be a quality closer. Good stuff as always.

  9. JR

    August 05, 2010 12:19 PM

    You’re stats are fun for discussion purposes, but some people take these way too seriously. So let me get this straight, Brad Lidge is a better pitcher this year then Ubaldo Jimenez (kwERA of 3.76). Who knew. Can we make the trade today!

  10. CH Phan

    August 05, 2010 12:39 PM

    Hahaha – I felt like Drew when I looked at Lidge’s stats. I had a sudden vision of an old-time, Warner Bros. Bugs Bunny baseball cartoon. Lidge would be tanking a Phils win at the bottom of the 9th, while Bugs screams into the camera BUT THE STATS SAY HE DOESN’T SUCK! Then we’d hear the crowd booing & the final music.

    The camera would pan back on Bugs leaning against a tree outside the stadium, baseball cap cocked back off his giant ears, smacking away on a large carrot, making that wise-acre face and saying, “Numbahs, numbahs, numbahs. That’ll lo-in ya.”

    Eh, I’m just saying that maybe we should take everything in moderation.

  11. Phylan

    August 05, 2010 12:40 PM

    That’s not an indictment of the statistic, it’s both of you drawing incorrect conclusions from it.

  12. Phylan

    August 05, 2010 12:41 PM

    *grossly misappropriates a metric*

    Wow guys stats are pretty dumb, shucks!!

  13. Steve

    August 05, 2010 02:54 PM

    I like sabermetrics as much as the next guy, but even I have to laugh occasionally when new statistics are invented. At this point, you can use statistics to show anything you want.

  14. zfg

    August 05, 2010 03:03 PM

    I don’t think anybody ever said “pitcher X is better than pitcher Y because his kwERA is lower!” It’s just another tool for analysis. You wouldn’t use a hitter’s batting average or walk rate as a standalone metric, but it IS something that can be useful to you.

  15. Phylan

    August 05, 2010 04:00 PM

    Yes, you can use statistics to “show anything you want,” if a) you’re a liar and b) the person you’re citing them to has no critical thinking abilities.

  16. bb24

    August 05, 2010 04:29 PM

    I’m totally on-board with a lot of sabermetrics stuff and the idea behind taking a more analytical look at the game than traditionalists do. That said, don’t you ever feel like the thousands of bloggers doing this keep inventing new stats that are just *slightly* different but don’t tell us anything new?

    That is, as Bill’s conclusion shows, the only thing this stat does is confirm exactly what we already know from other (easier to understand) stats, and what most people could have guessed just from watching.

  17. Bill Baer

    August 05, 2010 04:29 PM

    Oh cool, Tango stopped by while I was gone. I didn’t have a chance to get an autograph, dad gummit.

    As pointed out, kwERA only shows that Lidge has been good in one limited perspective. But he does have a 3.65 SIERA. Despite losing a lot of stuff on his fastball and slider he has still been able to get strike three at a pretty high clip and pitchers that do that usually enjoy more success. Lidge hasn’t had as much success because he’s been a bit unlucky in a small sample of innings and he has had very bad control not only in terms of walks but simply falling behind hitters and getting himself in fastball counts.

    Per Baseball Reference, Lidge has only been ahead in the count (0-1, 0-2, 1-2) with 27 of 103 batters faced. When the batters are ahead, they have hit for a 1.129 OPS; when the count is even, the OPS lowers to .737; and when Lidge is ahead, they only hit for a .556 OPS.

  18. kmart

    August 05, 2010 05:52 PM

    How is the constant determined? It seems like a pretty arbitrary number to use.

  19. CH Phan

    August 05, 2010 05:52 PM

    As anticipated, Phylan’s sense of humor is MIA. I didn’t say: “Wow guys stats are pretty dumb, shucks!!” Didn’t even imply it. In fact, I typed out the word: moderation. but nevermind.

    Hm, maybe what I meant was: Wow guys who ONLY believe stats are pretty dumb, shucks!!

    Lighten up.

  20. Scott G

    August 05, 2010 10:07 PM


    The matchup of J.C. Romero vs. strike zones came out as expected again.

    Charlie Manuel’s job should be taken when he allows Romero to face a RHB. If he can’t see two steps ahead tonight to see that the Marlins were just planning to burn Chad Tracy, then he doesn’t deserve to be a major league manager. I’d rather see anyone in the bullpen pitch to the opposite handed batters before Romero faces a righty. I’d even rather see Lidge face a LHB.

  21. euphronius

    August 06, 2010 11:01 AM

    Why cant Bastardo have Romero’s job.

  22. euphronius

    August 06, 2010 11:02 AM

    Oh wait is Bastardo hurt?

  23. SJHaack

    August 06, 2010 12:11 PM


    I imagine that the constant was determined by finding a number that made this one fall in line with ERA so that if you’re already familiar with that number you don’t need to do a mental jump shift to decide what kwERA is “good”.

  24. kmart

    August 06, 2010 06:03 PM


    That’s kinda what I thought, I was just curious though.

  25. David

    August 08, 2010 02:57 AM

    I’m not sure K-BB on its own is the best way to handle this stat – Cliff Lee has only given up 9 walks, but he’d need 175 strikeouts (over 10 per 9IP) for a kwERA under 2.

  26. Brian

    August 09, 2010 04:58 PM

    Just tried to read this article. It gave me a huge headache. I guess I’ll go watch some baseball WITH MY EYES and determine how good these guys are. As far as Madson goes, if you watch the games, when he pitches in the 8th his stuff is devastating and when he comes into a save situation he throws BP.

Next ArticleThe Transformation of Cole Hamels