WAR for Knuckleballers

As you probably know, there are multiple ways to calculate WAR, with our version (sometimes referred to as fWAR) and Baseball-Reference’s version (or rWAR/bWAR, depending on who is citing it) being the two most popular. And, as you probably also know, they can be quite different when it comes to evaluating pitchers. Two years ago, I wrote a couple of posts (Part One, Part Two) explaining why we constructed our pitcher WAR the way we did, and what we see as the pluses and minuses of both systems. Neither system is perfect, and neither system works in every situation, but we feel that using a FIP-based approach to evaluating pitchers is more transparent and more effective for the majority of pitchers in baseball.

But, there are pitchers who routinely outperform their FIP, and the way we do pitcher WAR will systematically underrate those pitchers. The primary group of pitchers who post results that are better than their BB/K/HR rates would suggest are knuckleballers.

Tom Tippett — who now works for the Red Sox — wrote about this extensively back in 2003, so this isn’t a new idea. When Voros McCracken released DIPS 2.0, he adjusted the system to account for the fact that knuckleball pitchers didn’t really fit into the normal trend. Even before R.A. Dickey ever started throwing the pitch, knuckleball pitchers were a well established exception to the norms of FIP. And that means that you don’t want to use standard “fWAR” for pitchers to evaluate Dickey, or any other knuckleball specialist.

Because there are exceptions to FIP, we do not simply display only FIP-based WAR here on FanGraphs. Back in August, we rolled out a few stats that we put under the umbrella of Fielding Dependent Pitching, helping to bridge the gap between a pitcher’s WAR based on his FIP and what his WAR would be if we just evaluated him by runs allowed. Not surprisingly, Dickey’s RA9-wins (which is just WAR based on runs allowed instead of on FIP) over the last three years are significantly higher than his FIP-wins, which is what we call WAR for pitchers here on FanGraphs. By FIP, Dickey has been worth +10 WAR over the last three years. By runs allowed, he’s been worth +15 WAR.

Because Dickey throws the knuckleball, and we already know that knuckleball pitchers outperform their FIP, you should lean more towards RA9-wins for Dickey than FIP-wins. Pitchers like Dickey are why we present both options here on FanGraphs. There simply isn’t one system that works perfectly for everyone. No matter what kind of calculation you use, there are going to be guys who don’t fit perfectly into the model. We added RA9-wins to help show the differences between the two systems, and to give you an alternative way of valuing pitchers who don’t fit perfectly into the assumptions that FIP makes.

You shouldn’t just use RA9-wins for any pitcher who outperforms his FIP, as often times, that’s simply the product of good teammates or some good luck, but you should also know that FIP doesn’t work for every pitcher. Specifically, it doesn’t work for knuckleball pitchers. Eno Sarris helpfully created a custom leaderboard of every pitcher in baseball history who has been noted to throw a knuckler, and as a group, their combined RA9-wins total (1,181) was 206 wins higher than their FIP-wins total (975).

This isn’t just an R.A Dickey or Tim Wakefield thing. Knuckleball pitchers induce weak contact that leads to consistently lower than average rates of hits on balls in play. Because of this, we strongly suggest that you use RA9-wins to evaluate a knuckleball pitcher, not FIP-wins. Dickey has outperformed what our standard WAR suggests, and he’s likely to do so going forward as well. He’s a known exception to the rule. If you want to quote Dickey’s FanGraphs WAR, you’re better off using RA9-wins than FIP-wins. For pitchers like Dickey, it’s a better representation of his real value.




Print This Post



Dave is a co-founder of USSMariner.com and contributes to the Wall Street Journal.


65 Responses to “WAR for Knuckleballers”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Dan Rozenson says:

    Good insights. No stat is perfect.

    Vote -1 Vote +1

  2. Matt says:

    Thanks Dave. I was the guy that asked about this in the chat, this is really helpful!

    Vote -1 Vote +1

  3. Colin says:

    Why wouldn’t I just add about 20ish percent to his war win total?

    Vote -1 Vote +1

  4. Kyle says:

    “but you should also know that FIP doesn’t work for every pitcher”

    and slightly changed from how you worded it:

    “(pitchers who) induce weak contact lead to consistently lower than average rates of hits on balls in play”

    Please preach the heck out these things to your readers

    Vote -1 Vote +1

    • Jim says:

      That second part of your post there is one of the reasons most stat people consistently underrate Mariano Rivera. A .262 BABIP over 1200+ innings isn’t just luck.

      Vote -1 Vote +1

  5. MikeS says:

    Knuckleball pitchers induce weak contact that leads to consistently lower than average rates of hits on balls in play.

    Without doing any research, I wonder if this sort of thing could apply to guys like Mark Buehrle who seem to outperform their FIP as well.

    Vote -1 Vote +1

    • Marc says:

      Well, Buehrle is also a world class defender.

      Vote -1 Vote +1

    • derp says:

      Someone ran through Buehrle’s career defense and determined, if calculated accurately, that it spits out a runs saved/9 number that when compared to an average defender at his position is almost exactly the same as his career ERA-FIP separation. It’s a hell of a coincidence, or it explains it perfectly.

      Vote -1 Vote +1

  6. Nick says:

    I still wouldn’t give anything up for him.

    -32 Vote -1 Vote +1

    • Joe says:

      A fine post, but stop posting it in every article please.

      +27 Vote -1 Vote +1

    • Ben says:

      Agree with Joe, thrice is enough.

      Vote -1 Vote +1

    • TKDC says:

      The Blue Jays finished 4th every year 2008-12 and have won fewer than half their games in that stretch. Now, they might be the best 4th place team for all or most of those seasons, but lets not oversell it.

      Vote -1 Vote +1

    • Jaysfan says:

      @TKDC you honestly believe the roster they Jays have right now will finish behind one of the red sox or orioles? I get that saying playoffs and division champs automatically is a stretch, mostly because of the rays, I’m not a believer in the Yanks at all this year, but suggesting their doomed tofinish fourth again is level 1 trolling.

      Vote -1 Vote +1

    • TKDC says:

      @Jaysfan,

      I’m not sure if you are beta-trolling or something here, but I was responding to: “best team in baseball not to make the playoffs like Jays teams of the recent past”

      By my count, finishing in 4th place kinda precludes being the best team in baseball to not make the playoffs, considering the 3rd place team in the East (always the Yanks, Sox, or Rays) each of those years also missed the playoffs.

      My post was clearly looking back, not forward. The Jays have an excellent opportunity to do well this year, but for some reason I think if the Jays make the playoffs, Jays fans will try to spin it into some continuation of a great team finally making the playoffs when really what you probably had (again, I’m talking about 2008-12) was an average to above average team with no chance to compete due to the strength of the AL East.

      Vote -1 Vote +1

      • vilifyingforce says:

        Whatever you were replying to must have disappeared, I don’t see any reply talking about the Jays being the “best team in baseball not to make the playoffs like Jays teams of the recent past” in the story or the comments.

        Vote -1 Vote +1

  7. Slats says:

    People just don’t trust knuckleballers to be able to sustain their success. Right or wrong, that’s the story.

    Vote -1 Vote +1

    • Phantom Stranger says:

      I agree the market largely undervalues knuckleball pitching. If Dickey had been a conventional pitcher with the same track record, his deal with the Blue Jays would have likely been higher per year.

      What the Blue Jays did was pure Moneyball, acquiring talent undervalued by the rest of the market.

      Vote -1 Vote +1

      • Dan says:

        But that’s the problem, right? Dickey is hard to value. He’s had 2 good, and one great season at ages 36-38. Many knuckleballers have aged well, but Dickey is a different breed because of the velocity at which he throws the pitch. I have no doubt that he can pitch in the majors for the next 3-5 years, but it is reasonable to wonder if he can sustain this level of performance.

        Vote -1 Vote +1

  8. Rippers says:

    “We’re using ERA- for Dickey because knuckleballers have a long exhibited history of being an exception to FIP, by the way.”

    fWAR is FIP based, for those who are uninformed

    Vote -1 Vote +1

  9. JS7 says:

    A 38-year-old knuckleballer posting the best K/9 of any knuckleballer in history is truly special.

    Vote -1 Vote +1

  10. Old School says:

    Tim Wakefield will always be better.

    Wakefield: 200 career wins.
    Dickey: 61 career wins.

    -32 Vote -1 Vote +1

  11. Matty Brown says:

    Can’t fucking wait to watch him throw those Knuckle-balls this season for my Jays!

    Vote -1 Vote +1

  12. Dan says:

    no analysis into every knuckleballer the last 10-20 years and how their stats hold up to FIP and RA9 analysis? Dickey might be “outperforming” his FIP more than other knuckleballers in which case you might expect him to regress more towards how other knuckleballers have performed relative to FIP.

    Saying Dickey has outperformed it for 3 years by a decent amount isn’t really relative to anything (small sample size and all) without an analysis of how all other knuckleballers have fared.

    Vote -1 Vote +1

    • Kyle says:

      3 years isn’t a good enough sample size?

      Vote -1 Vote +1

      • Dan says:

        not for a pitcher sample size of 1.

        There are dozens of pitchers every season that have outperformed their FIP the previous 3 seasons. Most of them will not significantly outperform their FIP in year 4.

        Vote -1 Vote +1

      • cass says:

        Do you have a source for that? As far as I’ve read, a pitcher who outperforms his WAR for three years in a row is absolutely more likely to outperform it in the 4th year.

        Vote -1 Vote +1

  13. channelclemente says:

    Do you see the ‘poor contact’ factor play out in BABIP as well?

    Vote -1 Vote +1

  14. Johnny A. says:

    Awesome to see a 38-year-old suddenly post such historic stats. Remember just about a decade ago we saw this kind of thing often. Then something changed and players started aging quicker. Nice to see Dickey bucking that trend

    Vote -1 Vote +1

  15. Chris says:

    How about this wild stat known as ERA? It’s obviously context dependent but it’s hilarious how peripheral stats are seen to be more valuable than actual runs allowed. Peripherals are better for predicting future performance but ERA is still all that truly matters for a pitcher.

    -5 Vote -1 Vote +1

    • DCN says:

      “Better for predicting future performance” is pretty important. The question is whether there are some pitchers for whom future performance is better predicted by ERA.

      Vote -1 Vote +1

    • bookbook says:

      If you’re going that route, how about RA? I think I read somewhere that the unearned runs each pitcher allows are not only largely the pitcher’s “fault” but also are material to final game results.

      Vote -1 Vote +1

    • joser says:

      No, runs allowed (whether they’re earned or not) matter, and they matter to the team. ERA has a big defensive component (the same pitcher in front of a better defense will post a lower ERA) which makes ERA a team stat. A team stat that traditionally got awarded to individual players because, traditionally, we didn’t know how to parse out the defensive components. Now we do and — surprise, surprise — the defensive-independent stats are better predictors of future ERA than ERA is.

      (This is similar to RBIs on the batting side: runs are all that matter to a team, and RBI is a team stat that traditionally got awarded to individual players even though they weren’t responsible for writing out the batting order or getting guys on base ahead of them; now that we have other, better ways to measure offensive production we can regard RBI, like ERA, as just a very flawed non-predictive stat with a lot of lovely tradition grown up around it.)

      Vote -1 Vote +1

      • Natman says:

        Ok see as a stats guy the assertion by @josser is just odd. These other measures are not antiquated simply because new measurements are used. In fact, these traditional stats have meaning and value, especially when viewed in larger context around the newer stats. And the new stats might hold some flaws that these old stats might very well shed some light on.

        For example, yeah RBIs are indicative of being thrown into a spot that has some degrees of chance. However, getting that hit or run or flyball or whatever has meaning in a baseball game. And a guy who can hit a flyball out is better than a guy who weakly hits one back to the pitcher or a drawn in infield. That “RBI” then has some meaning for the individual player. Granted there are other statistical measures for “clutch” etc., but I’m not so sure these stats truly capture what RBI is trying to capture.

        ERA, yeah, its rough because there is judgment and skill behind a pitcher. But man oh man, everyone bemoans the pitchers who are better than their FIP (hello Matt Cain). I’d argue a John Lannan fits here too. Seriously, this article goes to that problem too. So getting rid of a stat for one that has problems is silly.

        Vote -1 Vote +1

      • David says:

        Natman, I think you’re clinging to the comfort of older stats because you have context for them, but Im gonna go with Josser on this one, RBI has no value, and when I say no value, I am not exaggerating. It only appears to have value because of some degree of correlation to things like RBI rate, ISO and wOBA/wRC+

        As for ERA vs FIP, you bring up Matt Cain, which is a fabulous example of a pitcher with a relatively normal pitch repertoire consistently outperforming the ERA that his FIP says he should generate (not outperforming his FIP, FIP is FIP) However, when we look deeper into his peripherals, the reasons become evident. Yes, Matt Cain does appear to have a better than expected BABIP which is probably due to the scouting reported and pitch F/X supported theory that he gets exceptionally good late movement on his pitches, particularly his fastball, but the biggest difference comes when we look at his home vs away HR/FB% which pretty clearly shows that his home ballpark (where he has pitched just north of 50% of his games) helps him drastically limit his homers allowed while giving up the same amount of fly balls. This gives us very good evidence that if Matt Cain pitched in, say, Fenway, he wouldnt have nearly the same career ERA – FIP split. We arent “getting rid of stats because they dont work for one person” we are using all of the other data we have to try to figure out why it doesnt work for that one person.

        Vote -1 Vote +1

      • YanksFanInBeantown says:

        The home run rate being lower in his home park only helps his FIP. What you’re saying only accounts for his xFIP split.

        Vote -1 Vote +1

  16. DCN says:

    What about other pitchers who routinely outperformed WAR (mainly location-based guys known as “smart pitchers”)?

    Among them

    Tim Hudson (11 out of 13 years, 0.36 gap over 2,682 innings)
    Mark Buerhle (11 out of 13 years, 0.32 gap over 2,789 innings)
    Tom Glavine (16 out of 21 years (9 out of 9 after 2000), 0.41 gap over 4413 innings)
    Matt Cain (7 out of 8 years, 0.38 gap over 1536 innings)
    Mariano Rivera (13 out of 18 years (13 out of 16 since moving to the bullpen), 0.54 gap over 1218 innings)
    Catfish Hunter (13 out of 16 years, 0.31 gap over 3449 innings)

    Can we identify common traits for these guys? Do they belong to a few different groups?

    Vote -1 Vote +1

    • bookbook says:

      Good pitchers with lame fastballs? I wonder if this comes down to survivor bias in some form or fashion. (i.e. only slow fastball guys with the luck to outperform their constituent parts get to pitch for 12 or 18 or 21 years in the majors.)

      Or maybe there’s unusual movement on their pitches.

      Vote -1 Vote +1

      • Phantom Stranger says:

        It’s very interesting you bring up unusual movement. I have always suspected a spitballer would produce abnormally low BABIP numbers, much like the knuckler. Someone should investigate Gaylord Perry’s numbers and possibly other cases of known users.

        Most people think Whitey Ford was cutting the ball most of his career.

        Vote -1 Vote +1

      • Wil says:

        Well I wonder if for Hudson and Glavine if it has to do with their K numbers. Neither is a strike out pitcher but both are effective. It could be that guys who are “contact pitchers” get shafted by WAR simply because of the way FIP weighs K/9 and BB/9.

        Vote -1 Vote +1

      • Matthias says:

        The survivor bias, I think, is a lot stronger we realize. I would not be surprised if it popped up in scenarios like this.

        We say that Matt Cain has made it past the “lucky” stage now, 7 years into his career. But a coin will land heads 7 times in a row with a probability of 1/128. There have been many, many pitchers that began careers over Cain’s time frame. That’s not an insignificant probability.

        There’s obviously the magnitude of his over-performance that should be accounted for. I was only hoping to use simple probability to shed light on the possibility of the survivor bias, not to attack Matt Cain, specifically.

        Vote -1 Vote +1

      • Nivra says:

        It’s not a coin flip that you’re doing 7 seasons in a row.
        The probability that a pitcher outperforms his FIP to the extent that Matt Cain does each year, is pretty small, and is nowhere close to 50%.

        Fangraphs doesn’t have an ERA-FIP stat, so I’ll use BIP-wins as proxy. In 2012, Cain was 4th out of 88 pitchers for BIP wins. Assuming a Uniform distribution (which probably isn’t true, it’s most likely normal), he has a 4.5% chance of being 4th or higher.

        In 2011, he was 8th out of 94 or 8.5%. In 2010, he was 4 out of 92 or 4.3%. Since I’m lazy, I’m just going to average all three seasons and say a player has a 5.8% chance of posting BIP-wins that high from random fluctuation.

        The probability of doing this 7 years in a row?
        0.000000002

        Now, for this to be Matt Cain being an extreme example because so many pitchers have entered the league, we would have to have seen about 10 million pitchers enter the league over the last 7 years.

        And this is using a uniform distribution. If I used a normal distribution, the probability would be even lower.

        Vote -1 Vote +1

    • bstar says:

      Absolutely we should not use FIP/fWAR for guys like Hudson/Rivera/Glavine.

      But what about those pitchers who underperform their FIP also on a consistent basis? This creates even crazier differences between the players’ career rWAR/fWAR. Here’s some examples:

      Player name rWAR/fWAR
      John Lackey 22.3/35.7
      Josh Beckett 31.7/40.1
      Carl Pavano 14.9/25.7
      Javy Vazquez 40.2/55.1
      Liv Hernandez 21.7/35.4
      Derek Lowe 30.4/48.4

      Derek Lowe’s fWAR is almost 60% higher than his rWAR. fWAR paints Lowe as a guy who’s knoccking on the door of the Hall of Very Good, and I’d like to think we all know better than that (same for Vazquez, really). These differences are actually more striking than the WARs of pitchers who outperform their FIP, mainly because fWAR is usually a tad higher than rWAR overall.

      If Fangraphs is finally going to fix their WAR for knuckleballers, they need to do the same for all pitchers who consistently out- or underperform their FIP. fWAR for these pitchers has little value.

      Vote -1 Vote +1

  17. Shoe strings says:

    The beauty of a knuckleball in super slo mo should be worth an extra win.

    Vote -1 Vote +1

  18. Tom says:

    WAR isn’t fair to use on jews, either. Their stinginess is skewed by the multicollinearity when you build a regression equation — it seems as though being cheap is highly correlated with having a huge fucking schnozz, which overestimates the weaknesses of a jew pitcher, so, yeah, you know…

    WAR is anti-semitic. probably constructed by some black dude that was screwed over by the 20% interest rates charged by some guy named Joseph Steinowitz in NYC.

    -42 Vote -1 Vote +1

  19. shthar says:

    It’s tough to turn down good pitching.

    But I think he’s gonna do worse in Toronto.

    Everything’s gotta be perfect for a knuckle baller. And this guys going to a new team, new league, new country even.

    We’ll see.

    Vote -1 Vote +1

    • ZenMadman says:

      Yeah, but most importantly, the weather has to be perfect for the knuckle ball. There is no weather in a dome. Toronto plays 55% of their games in a dome.

      Vote -1 Vote +1

      • shthar says:

        So he’s gonna be pitching in an environment he’s rarely pitched in before?

        I don’t see that as a plus.

        Vote -1 Vote +1

      • ZenMadman says:

        Well, he did okay pitching in the dome in tampa: 1H, 1R, 0ER, 0BB, 12K, 106 pitches, 79 strikes. It was basically a perfect game except for two balls booted by David Wright.

        I don’t think Dickey will require much adjusting to the dome.

        Vote -1 Vote +1

  20. Bab says:

    Would it be persnickety to ask about wind pattern differences between Dickey’s old stadium and new stadium…?

    Vote -1 Vote +1

  21. Fatbot says:

    In the zeal to heap as much love on Dickey as possible on Fangraphs (second only to Greinke), there continues to be logical inconsistencies. This article concludes “It’s not your traditional approach — nothing about this knuckleball pitcher is “.

    Now here we have Fangraphs lumping in Dickey with all other knuckleballers. Well, which is it? Is he different or the same?

    If he’s different — if he’s truly throwing a faster, less moving “knuckle” pitch he’s controlling instead of an actual knuckleball (that’s defined by its unpredictability) — we need to throw out the knuckleball comparisons and view him with other offspeed specialists, and not make these articles with sweeping comparisons with past knuckleballers he’s so different from.

    Or, if he truly is yet another knuckleballer in the tradition of knuckleballers, then you need to examine why, after generations of knuckleballers, is Dickey so much better? Perhaps, is there something about today’s MLB players that the knuckle is more devastating against them (hint: see record-breaking K% over past 5 years no past knuckler pitched against)?

    Today’s post-steroid era batter approach is contrasted here. Girardi, a guy who didn’t play that long ago, approached with patience and “take it to right field”. Now today’s typical player who doesn’t give a rat’s ass if he strikes out, Rick Ankiel: “To me, I’m going to go up there and take a hack in those three swings.”

    I was totally wrong in my anti-Dickey stance of thinking the clock would strike midnight on him in 2012. As long as MLB is dominated by hitters with the latter approach, I will continue to be wrong about Dickey as he will continue to pitch like an ace.

    Vote -1 Vote +1

    • shthar says:

      I think there’s always a built in sympathy for knucklers.

      Not so much the underdog effect. More the, I can’t throw a hundred miles an hour, but I could do that! effect.

      Vote -1 Vote +1

  22. Erik says:

    I would think we might solve these disparities by using pitchfx eventually.

    It seems logical to me that pitchers who throw softer will have lower BABIP’s.

    Pitchers who throw fewer fastballs, super slow pitches like knuckleballs, or pitch down in the zone should all do better than expected.

    Vote -1 Vote +1

  23. gcil says:

    I raised this issue about 2 years ago before I knew anything about advanced stats. I had just started getting into it, and I questioned the use of BABIP for certain pitchers, and in turn their FIP and WAR.

    I was treated to the usual “you just don’t understand” BS from the sabre guys I was talking too, but I thought what I was saying was as clear as day. Pitchers who threw certain pitches, or pitched in certain areas (high and tight, low and away, or on the very inside of the plate) would induce weak contact, and would have a naturally lower BABIP.

    I just randomly pulled 3 guys off the top of my head, Moyer, Glavine, and Maddox. I checked on baseball reference and low and behold they had lower BABIP’s than the league average by about .010. I then went on to look up knuckleballers, and it was even more extreme than that.

    I finally came across Tom Tippet’s article, and I was confused because it was from 2003. Obviously someone else thought about what I was saying way before, and had proved it to be true. Yet no one had come up with a different or revised stat for showing this.

    I don’t believe this phenomenon applies only to knuckleballers, but also to players with exceptional changeups, inside movement, and the prototypical “crafty lefty” nibblers. I’m sure there are other instances as well that it applies as well. I think looking into building a second WAR, or different stat all together for pitchers might be a good idea.

    Vote -1 Vote +1

  24. froston says:

    I made a rough aging curve for prominent knuckle ballers using RA9-Wins. Thought people might be interested.

    http://postimage.org/image/hjllo9cej

    Sample was small, basically just the 10 knucklers with the most games started + Dickey and minus Cicotte since he pitched a century ago.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current ye@r *