FanGraphs Fantasy Baseball


RSS feed for comments on this post.

  1. Finally a write-up about this! I noticed this too a while back. As a result I’m always quick to look at pitcher’s opposing hitter contact rates when trying to see how legit the K/9 is…

    Comment by Sean — March 31, 2011 @ 4:52 pm

  2. Cool article! I never thought about the fact that a struggling pitcher will have more chances to strike guys out. And the two tables of pitchers you used are just far too convincing…

    Comment by Ameer — March 31, 2011 @ 4:55 pm

  3. Thanks Mike. The charts would have benefitted from including the actual K/PA rates, but this is an excellent article.

    Comment by JohnnyBigPotatoes — March 31, 2011 @ 5:36 pm

  4. Well, at least John Axford’s K/9 was good today.

    Comment by nosferatu — March 31, 2011 @ 5:37 pm

  5. Excellent article, Mike.

    Comment by RotoChamp — March 31, 2011 @ 6:00 pm

  6. By the way, I use for K%. StatCorner has the stat too, but they haven’t updated their leaderboards in years.

    Comment by Mike Podhorzer — March 31, 2011 @ 7:02 pm

  7. I have to say, this is a pretty brilliant analysis. It’s clear, definitive, and very interesting. I really appreciate this piece and it has definitely made me think about a relationship I probably would never have noticed (or at least would not have noticed for a very very long time). Thanks!

    Comment by B N — March 31, 2011 @ 11:14 pm

  8. Thanks, appreciate the feedback! I am a big fan of SwStr% on FanGraphs, so I look at that a lot and compare it to K/9. Shields had the lowest SwStr% of his career last year, yet the highest K/9. Hmmmm

    Comment by Mike Podhorzer — March 31, 2011 @ 11:32 pm

  9. Great article!

    I realized this a long time ago and have been using it instead of K/9 for a while. Now is there a way to use this to create a more accurate FIP or xFIP type metric? Applying the same concept to BB/9 would also be a good idea, no?

    Comment by UncleCharlieVT — April 1, 2011 @ 9:22 am

  10. Awesome article.

    In the second group (where K/9 is lower than K/PA), I think for most of those pitchers its not a issue of BABIP being related luck. Rather, when you look at that list of names (Hallady, Lee, Oswalt, Haren, Sabathia, Wainwright) you are seeing a lot of the top pitchers in baseball. On the other hand, with the other group you are looking at a 3rd tier of fantasy pitchers (Jackson, Beckett, Shiels, Niese)

    Comment by Nick Tenaglia — April 1, 2011 @ 9:59 am

  11. Very nice article. I just went and redid all of my calculations for tomorrow’s draft using K:TBF, BB:TBF, and HR:TBF.

    Comment by Blue — April 1, 2011 @ 10:40 am

  12. Why do FIP and xFIP used Innings Pitched in their formula then? I mean, obviously someone has thought of this before but the formula is still the same? Is there something better about Innings Pitched or would it be possible to create a better predictive metric using Batter Faced?

    Comment by Briks — April 1, 2011 @ 10:47 am

  13. I seem to recall an article by TangoTiger where he did a breakdown by replacing IP with PA and adjusted the coefficients within FIP accordingly and found it improved FIP slightly, but not enough to rehash the tool for his liking.

    I don’t fully subscribe to the “easier, therefore better” mantra there, as I firmly believe that any increase in value to the statistic outweighs the added difficulty in understanding it.

    For scaling K/PA or BB/PA, I think the general idea is to multiply by 38 (the average, or so, number of batters faced in a typical game).

    Another interesting side-effect of using PA as the denominator is for K/BB. Denominators get an unhealthy weight naturally. Lets just say you have 20K vs 5BB = 30K/5BB = 6 K/BB. To increase it 1, you need 5K, but to decrease it by 1, you only need 1BB. A much better indicator would be K/PA – BB/PA, giving you the “difference efficiency” per PA. Again, multiply by 38 if you wish to scale to a 9-inning game.

    All-in-all, PA is always the better option, because it removes all (?) external forces inherit in IP (defense, BABIP luck, etc). For me, funnily enough, it’s also more intuitive to think of everything in the case of a single batter, rather than an inning. Don’t know why, just is.

    Comment by mattinm — April 2, 2011 @ 11:03 am

Leave a comment

Line and paragraph breaks automatic, e-mail address never displayed, HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Close this window.

0.148 Powered by WordPress