The Seven Sub-2.00 FIP Seasons

For my first post on High Heat Stats, I’m taking a shortcut and recycling something I wrote for Replacement Level back in April.  It’s no more or less relevant today, but I think it’s worth a little more discussion than it got in my dark corner of the blogosphere.

Bonds 73, McGwire 70, Sosa 66…
Gibson 1.12, Gooden 1.53, Maddux 1.56…
Hornsby .424, Sisler .420, Williams .406

There are single season numbers like the home run totals listed above that the average baseball fan has committed to memory (even if some fans would prefer to remember Maris 61, Ruth 60, Ruth 59…). There are others, like Gibson’s 1.12 ERA and Williams’s .406 batting average, that may not actually be all-time records, but if we look past the deadball era, they stand out so far above more recent accomplishments that we can’t forget them.

Read the rest to see the numbers you should really know.

One thing we as fans seem to have lost during the SABR revolution is the ability to memorize truly meaningful numbers. Sure, Gibson has an amazing ERA, but how much of that came from pitching from a high mound in a cavernous stadium in a pitcher’s era? Williams hit .406, but why do we care so much about a number that ignores walks (for the record, his OBP that year was an absurd .553).

Perhaps this is all for the best. There’s nothing wrong with memorizing numbers that blew our minds decades ago and recognizing at the same time that those numbers don’t evaluate those players’ true abilities as well as metrics developed since.

Still, it seems to me there are a few numbers that modern fans should pass down to our kids the way .406 and 1.12 and 61* were passed on to us. Fielding-Independent Pitching is a better way to evaluate a pitcher’s true talent than ERA. By including only strikeouts, walks, and home runs allowed, it focuses on the outcomes most within a pitcher’s control, ignoring all the variables that can happen once a ball is put in play.

Who had the best single season FIP of all time? It seems to me that we should all know that. Even without memorizing leaderboards, you might guess correctly, as the season and the pitcher who accomplished it are equally legendary in SABR circles. You may be surprised, though, to know that, since 1920, when Babe Ruth essentially invented the home run, and today, there have been just seven seasons in which a pitcher has qualified for an ERA title and kept his FIP under 2. Here they are, in ascending order of greatness:

Hal Newhouser, 1946 (1.96)
This was the second of Newhouser’s consecutive MVP seasons. His 1.94 ERA didn’t match the 1.81 he put up in ’45, but in 292 2/3 innings, he did strike out 275, while walking just 98 and giving up only ten homers.

By today’s standards, Newhouser was quite the workhorse in ’46, starting 34 games and appearing in relief in three others, but those numbers were in line with many of his prime seasons and far shy of the 313+ innings he pitched in 40 games (36 starts) the prior year. Based on volume, one can make a solid case that Newhouser was better in ’45, when he struck out fewer and walked more, but gave up just five homers and, thanks in part to a .259 BABiP (batting average on balls in play), yielded the same number of earned runs and four fewer runs in total.

We should note that Newhouser’s 1945 numbers were compiled during wartime, when many of the great players were overseas. In ’46, Ted Williams, Joe DiMaggio, and others returned to action, but Newhouser kept mowing them down as if they were scabs crossing the picket line.

Tom Seaver, 1971 (1.93)
Another Hall of Famer in his prime (Seaver was 26, Newhouser was 25), Seaver was utterly dominant for the ’71 Mets. In 286 1/3 innings, Seaver struck out 289 and walked just 61. He did give up 18 home runs, but they came in a more homer-friendly environment than Newhouser’s. Seaver wouldn’t match that .57 HR/9 again until ’75, by which time his strikeout rate had plummeted.

Remarkably, hitters fared even worse against Seaver when they put the ball in play, averaging just .263 and being stranded almost 86% of the time when they did reach base. His 1.76 ERA was the second best in the decade.

Sandy Koufax, 1965 (1.93)
At least until the late ’90s, Koufax’s run from 1962 until he retired after the ’66 season was the most dominant in the game’s history. In ’65, he threw a stunning 335 2/3 innings (he started 41 games and appeared twice more; any guesses why he retired at 30 a year later?). His 382 strikeouts are one short of Nolan Ryan’s 20th century record. He walked just 71 hitters, which at 1.9 per nine innings was the second lowest of his career. Had he not given up 26 home runs, this might be the best season of all time.

I’d be remiss not to mention that the National League batted just .246 as a whole in ’65, ten points lower than Newhouser’s AL in ’46 and three points lower than Seaver’s NL in ’71. Koufax also pitched in a ballpark that suppressed scoring by eight percent, so the odds were in his favor.

Sandy Koufax, 1963 (1.85)
In ’63, Koufax pitched 311 innings, averaging nearly eight innings per start. He struck out 306 and walked just 58, an even better walk rate than he’d have in ’65, and gave up eight fewer home runs. The league batting average and Dodger Stadium’s park factor were the same as in ’65. His 1.88 ERA was right in line with his FIP and well ahead of his 2.04 mark in ’65. This was less a factor of his BABiP (a fantastic .238, but still worse than ’65′s .234) and more a factor of his strand rate. He left 82.9% of baserunners on the bases in 1963, scattering what few hitters he allowed on base more effectively then he would two years later.

Bob Gibson, 1968 (1.77)
This one is commonly cited as the greatest pitching season ever. Everything went right for Gibson in ’68. He pitched 304 2/3 innings despite making “just” 34 starts (in case you don’t have a calculator handy, that rounds up to nine innings per start). He struck out 268, which is not quite Koufaxian, but does represent nearly eight batters per nine innings, and walked just 62. Only 11 batters, or less than one every three games, made home run trots against Gibson in ’68. Furthermore, luck was on his side, as the league hit just .230 on balls in play against him, which helps explain the .65 difference between his ERA and his FIP.

This time, we can’t help but note that the National League hit just .237 in 1968, the lowest mark for either league in baseball history. Scoring was so hard to come by that the mound would be lowered after the season and has never been raised to the pre-’69 height since. Gibson was truly remarkable in ’68, but his 1.12 ERA could not have been accomplished at any other time.

Dwight Gooden, 1984 (1.69)
Here’s the first guy on our list who isn’t in the Hall of Fame and likely never will be. Try telling someone in 1984, though, that the best pitcher they’d ever seen would have no compelling Hall case almost 30 years later. Gooden was a 19-year-old rookie in ’84, and made just 31 starts, but he pitched 218 innings and struck out 276 helpless hitters. His 11.39 K/9 are better than anyone we’ve seen on this list so far. His walk rate, at 3.01/9, is also the highest yet, but he gave up just seven homers in a more challenging run scoring environment (the league hit .260 in ’84 and Shea Stadium played neutrally).

Batters hit .296 on balls in play against Gooden, a truly average mark, but he struggled to strand runners (72.5% left on base). This, combined with the highest walk rate in the early part of his career, left him with a still-impressive, but perhaps not legendary 2.60 ERA. In the following year, Gooden shaved more than a point off his ERA, despite a 25% reduction in his strikeout rate and six more home runs allowed. He did this by walking four fewer hitters in almost 60 more innings. He also cut his BABiP by 27 points and increased his strand rate to a historic level (86.9%). One might argue that the reduced strand rate shows growth from a player who had finally entered his 20s, and that his much improved ERA was not all a factor of luck. Either way, both seasons were historically great (his FIP in ’85 was 2.13, and his 9.0 fWAR in ’85 tops his 8.6 in ’84).

Pedro Martinez, 1999 (1.39)
Have I mentioned that FIP is not park- or era-adjusted? Think about this for a minute. Pitching in Fenway Park in 1999, Pedro Martinez was .30 points better in FIP than any other pitcher in the history of the game. Even before we adjust for his home park and the hitters he faced, Martinez’s FIP in 1999 probably exceeded the next best season ever by a more impressive margin than any other single season record in any stat.

Let’s get to the numbers.
213 1/3 innings.
313 strikeouts (13.2/9)
37 walks (1.6/9)
9 home runs (0.36/9)

Batters hit .323 when they put the ball in play against Martinez that year. That’s a combination of horrible luck and a field that inflates batting average like few others (it played nearly run-neutral in ’99, due in large part to suppressing homers, but it still ranks tied with Tiger Stadium in ’46 as the friendliest offensive environment on our list). Despite the bad luck, bad fielding, and tough park to pitch in, he led the majors with a 2.07 ERA.

It gets better. In 1999, the American League hit .268, the highest average since 1937. Manny Ramirez hit .333/.442/.663 for the Indians. Edgar Martinez hit .337/.447/.554. Even John Jaha hit .276/.414/.556. Twelve American Leaguers hit 35 or more home runs. Yet, in this environment, Pedro Martinez struck out 8 1/2 batters for every one he walked and 35 for every one who homered off him.

After Pedro’s 1.39, the next best FIP in the American League was Mike Mussina’s 3.25. Aaron Sele was third at 3.85.

The following year, Pedro saw his strikeout rate dip by a batter and a half per inning, while walking just five fewer hitters. He gave up almost double the home runs in just four more innings. Despite all this regression, he turned in a 1.74 ERA.

I urge you all to commit the number 1.39 to memory. I guarantee you you’ll never have to replace it.

115 thoughts on “The Seven Sub-2.00 FIP Seasons

  1. 1
    James Smyth says:

    Great post!

    That 1.39 FIP is INSANE

    Also, with Bob Gibson’s 304 2/3 innings in 34 starts, he was only four outs shy of averaging nine innings per start

  2. 2
    Ed says:

    Welcome Bryan! Nice first post!

    I’m going to use this as an opportunity to ask a very tangentially related question:

    In 1968 the Cleveland Indians had two starting pitchers (Tiant and McDowell) with sub 2.00 ERAs. Does anyone know if any other post deadball era teams have accomplished this feat?

    • 3
      Phil says:

      Not that I can see, just scrolling through league leaders. It’s rare enough when there are two pitchers in the entire league sub-2.00.

      • 6
        Richard Chester says:

        The PI shows that Tiant and McDowell in 1968 represent the only occasion of two pitchers from the same team with sub-2.00 ERA.
        Here are the years in which there were two or more pitchers in the majors with sub-2.00 ERA.
        1985: Gooden, Tudor
        1972; Tiant, G.Perry, Nolan, Carlton
        1971: Blue, Wood, Seaver
        1968: Bolin, Gibson, John, McDowell, Tiant, McClain, McNally
        1966: Koufax, Peters
        1964: Chance,Horlen, Koufax
        1943: Lanier, Chandler
        1933: Hubbell, WEarneke

        • 16
          Ed says:

          Thanks Phil and Richard! It’s amazing that this feat isn’t talked about more. I’m an Indians fan and even I wasn’t aware of it until a few weeks ago when I stumbled upon it.

        • 20
          Hartvig says:

          RC- Obviously you are talking about post 1919, correct? Because in 1908 you wouldn’t even crack the top 10 with an ERA of 1.702.

        • 75
          Howard says:

          Does Warneke’s 1933 ERA round down to 1.99. Because it’s officially listed at 2.00.

          • 76
            Howard says:

            According to my quick and very basic calculations Warneke’s 1933 ERA is actually a tiny tick above 2.00 so he would come off the list. Someone better at math may well find that I’m wrong.

          • 80
            Richard Chester says:

            Warneke’s ERA was 2.00466 which rounds off to 2.00. When I ran the PI I incorrectly set ERA =< 2.00 instead of 1.99. Sorry about that.

            Perhaps I should have realized something was wrong when I mis-typed his name.:-)

          • 83
            bstar says:

            Richard, what is Warneke’s first name?

          • 85
            Richard Chester says:

            Warneke’s first name is Lon. He became an NL umpire after his playing career ended.

      • 27
        bstar says:

        It’s even rarer for one pitcher to post two consecutive seasons of sub-2.00 ERA. In fact, past the deadball era, there are only 3 pitchers to do it:

        Hal Newhouser 1945-46
        Sandy Koufax 1963-64
        Greg Maddux 1994-95

        The first two are on Bryan’s FIP list but Maddux is not due to his non-extreme strikeout totals.

  3. 4
    JDGentile says:

    This was a wonderful read. Thanks for this, Bryan

    At some point in the near future Fangraphs is likely to remove IBB from their FIP formula and create constants for the two separate leagues, considering one has pitchers hit and the other doesn’t. At this point, Pedro’s FIP will actually go lower. So commit to 1.39… for now.

  4. 5
    RJ says:

    So essentially Adam Dunn is a walking litmus test of pitching excellence?

    PS Good read, but wouldn’t have minded a quick explanation of how FIP is calculated. Easy enough to google, but still.

    • 51

      Fair enough, RJ. Still learning my audience here. And yes, Adam Dunn is a walking vacation from the immense power of luck in baseball.

      • 68
        RJ says:

        It’s quite alright, everyone else seems to have understood it just fine 🙂 Like I said though, good post. I’m relatively new to all this so I always appreciate learning new stats.

  5. 7
    John Nacca says:

    “Martinez’s FIP in 1999 probably exceeded the next best season ever by a more impressive margin than any other single season record in any stat.”

    The single season record for most wild pitches is 83 by Mark Baldwin in 1889. The next highest total is 63 by Tony Mullane in 1884.

    Yeah I am grasping at straws a little!!!!

    • 21
      Hartvig says:

      Depending on where you put the cutoff, Chief Wilson’s 36 triples in 1912 is 10 more than anyone since two league play began. If you move the cutoff point back to putting the pitchers mound at 60′ 6″ it’s still a substantial lead of 5.

      • 52

        I’ll debunk these two in one shot, although I love that you guys are digging for a better answer. John, I’m not partucularly impressed by wild pitches thrown at a time when the catcher might not have even worn a glove, let alone a mask. Hartvig, the triples record is amazing, but there’s a lot of context (ballpark construction, acceptance of swinging for the fences) that goes into a number from 1912. Pedro put up the best FIP ever, by a significant margin, when most contextual factors were working against him.

        • 65
          Artie Z. says:

          I wouldn’t go that far. Pedro had one contextual factor that someone like, say Lefty Grove, didn’t have in his favor – batters (for the most part) not caring whether or not they struck out or made an out by making contact.

          There’s no way that Pedro puts up this FIP number in 1931. His walk rate is great for 1931 (it’s good for most any time) but the walk rates are about the same across 1931 AL and 1999 AL (probably more similar if you remove pitcher hitting from 1931 AL).

          League run scoring is pretty close between the 1931 AL and the 1999 AL (5.14 in 1931 and somewhere around 5.2 in 1999 – not sure how to adjust for interleague play). I think this, in part, leads to the closeness of the 1931 and 1999 FIP constants.

          Pedro’s HR rate, while insane in 1999, is near the league average in 1931 – with lower HR totals you can’t have an extreme outlier on the downward side like 1999 Pedro because HRs allowed are bounded by zero. Basically, he’s not going to improve that by much even if you put him in 1931.

          And that leaves strikeouts – and there’s no way Pedro K’s 8.5 guys per 9 IP in 1931. The culture of baseball just wouldn’t allow it. If I look at the top 500 single season K/9 rates in MLB history, there are 6 through 1905 (a lot of Rube Waddell – although one is at #500 so he’ll probably fall off soon) and then 0 from 1906-1940. Johnny Vander Meer cracked the top 500 with his 8.03 in 1941 and then Newhouser and Feller cracked 8.4 in 1946. In the mid-1950s people like Herb Score, Sam Jones, and then Drysdale and Koufax came on the scene and we started seeing K/9 rates over 10.

          I think an underappreciated aspect of the TTO is that hitting culture plays a role. With players like Joe Sewell striking out once a month in 1931 (it was a down year for Joe), Pedro doesn’t approach his K totals in 1931 unless you think that he was VASTLY better than someone like Lefty Grove. I’m more inclined to think that major league hitters have a choice – a choice to swing from the heels all day, every day or a choice to try to put the bat on the ball. In the 1930s the choice was overwhelmingly in favor of putting the bat on the ball, and it doesn’t matter if you’re Roger Clemens or Walter Johnson or Lefty Grove or Randy Johnson or Koufax or Pedro or Bob Gibson – if a major league player really just wants to make contact, and a manager is choosing those players who make contact, it’s unlikely you are going to average a strikeout an inning.

          A simpler way to put it is this – there’s no manager in the 1930s who would allow Michael Bourn, with his 9 HRs through today, to have struck out 138 times. Michael Bourn, if you transport him back in time, probably wouldn’t play the same way because that’s not how baseball was played.

          All that said, Pedro’s season is great. If it’s not the best pitching season ever (and I think it is independent of whatever FIP says), it’s close. But he had help in that batters stopped caring as much about whether or not they struck out.

          • 66
            Artie Z. says:

            Factual error: Not sure where I got the 8.5 K/9 for Pedro in 1999. There’s certainly no way he has a K/9 rate of 13.2.

          • 87

            Artie, you’re absolutely right. Though to be fair, I did say “most contextual factors”, hitters’ willingness to strike out was a huge advantage for Pedro in ’99. From a run prevention standpoint, ’99 in the AL was about as tough as it gets, but from a FIP minimization standpoint, it’s been tougher.

            If I were to guess what a graph of league average FIP looks like over time, I’d assume it mirrors the same graph for ERA (low through ’19, high around ’31, low in ’68, peaking again around ’00), only with a gradual downward adjustment overlaid to reflect the rise of the strikeout.

            Great point.

          • 100
            John Autin says:

            Artie, perhaps I’m misreading you, but it seems that you’re presuming a fairly constant relationship between raw SO rate and FIP.

            If that’s not the case — if the SO component of FIP is more of a relative than a raw rate — then the contextual difference you cite between 1999 and 1931 becomes almost moot. Whether the league SO rate is high or low, there are individual outliers — for 1924-26, Dazzy Vance averaged 7.5 SO/9 against a league average of 2.8 — and I presume that it is their “outlier-ness” more than their raw SO rate that gives them an edge in FIP.

          • 108
            Artie Z says:

            Not sure where this is going to show up – it’s in response to John’s comment 100.

            John – the FIP constant C looks to be tied to run scoring, not K rates. The quick and dirty calculation is that Lefty Grove would have had to strike out about 320 batters in order to post a FIP under 2. This is with giving up 10 HR and 63 BB+HBP in 288 2/3 IP. This is with an ERA+ of 217 and a league leading ERA of 2.06. Grove was 2nd in k/9. His actual FIP is around 3. In order for anyone to strike out 320 batters (and over a batter per inning) in 1931 it would have taken a massive overhaul of how batters approached the game. If I take Lefty Grove and give him the same non-strikeout stats in 1999 I’m betting that Lefty strikes out more than the 175 he did simply because batters took a different approach during that era – and his FIP improves even if he does NOT improve to the average average K/9 rate in 1999 (his K/9 rate in 1931 is 5.5, the average in 1999 is 6.2). Let’s give him 6 K/9, or 192 K in 1999 – reasonable for Lefty Grove right? His 1999 FIP becomes 2.97 (down from 3 in his actual 1931 year) despite now being below average in K/9.

            Look at Dazzy Vance in 1924. 308 1/3 IP, a 2.16 ERA, a 7.65 K/9. He had 86 BB+HBP and allowed 11 HR. He almost doubled the K rate of the next player in the NL, and he was about 2.5 K/9 above the best AL pitcher (Walter Johnson had 5.12). His constant term is a little better than Grove’s at 3.03, but that’s because run scoring was around 4.5 for the 1924 NL. He struck out 262 that year, and his FIP is 2.64.

            When you start making comparisons across years it gets really interesting. This is kind of an additional point and not a direct response to John’s comment.

            Seaver, Koufax, Gibson, Newhouser, and Gooden really make the list because their FIP constants are low – they don’t have vastly better rate stats than than some other guys. Look at it this way – if Bob Gibson put up the exact same numbers in 1984 that he did in 1968 (I mean he pitched 304 2/3 innings, allowed 11 HR, struck out 268, and had 69 BB+HBP … heck, give him a 1.12 ERA) his FIP would be over 2. It’s nothing that Gibson did, it’s just that more runs were scored in 1984 so the 1984 constant term is greater than the 1968 one. If Gibson put up the same numbers in 2001 his FIP would have been 2.44. If Randy Johnson put up his 2001 numbers in 1968 (which really doesn’t seem all that unreasonable if you stuck a 6’10” lefty on the Dodger Stadium mound in 1968) his FIP would have been 1.46. If you are thinking “Well if these guys were in the league then the constant would change”, then just swap out Randy Johnson’s numbers for Bob Gibson’s and vice versa.

            FIP, in an odd sort of way, penalizes people who put up great numbers (like Randy Johnson) in big offense years. Look at Maddux in 1994 – 4 HR and 37 BB+HBP in 202 IP. He struck out 156 for a K/9 of 7. He can’t crack the level of 2. Heck, give him 220 Ks, which puts him at about 9.8 which was the NL leading figure in 1994. Maddux STILL (in a season with 4 HR and 37 BB+HBP and 220 K in 202 innings) can’t crack two because of the constant value.

            It makes Pedro’s 1999 FIP all the more amazing. If you put him in 1968 his FIP would be under 1. Now, comparing across eras, does that make any sense? Putting Pedro’s 1999 in 1968 (when we might more reasonably expect him to put up those numbers) shouldn’t make his FIP lower, should it? Or am I missing something?

          • 109

            John and Artie, thanks for clarifying. I’ll admit I was under the impression that the FIP constant was always 3.20, and that I didn’t realize Fangraphs wasn’t giving me FIP numbers before 1920 back in April. If my point had been ‘these are the seven best pitcher seasons ever’, I’d have failed miserably (though I think I got the first one right).

            In reality, I think there’s still some value to using FIP to identify great seasons, and I like the idea of committing some amazing SABR numbers like Pedro’s 1.39 to memory (which was my point), but you guys have helped me understand that most of the numbers on the list above are the great fielding-independent pitching seasons from years where scoring was down league-wide.

          • 110
            bstar says:

            The FIP constant varies to ensure that league average FIP = league average ERA.

            Past the deadball era, the lowest constant value is 1955’s 2.368.

            The highest mark ever is 3.501 in 1930.

            For 2012, the FIP constant currently sits at 3.097.

            Here’s the year-by-year list:


      • 63

        If I ever get the opportunity to be a General Manager, the triples and stolen base records will be broken.

        • 98
          Jonas Gumby says:

          Not if I get there first. AND, the wild pitch record will also fall, as my first order of business will be to acquire Ankiel, and reconvert him.

  6. 8
    no statistician but says:

    Deciding that FIP is a better way to discover a pitcher’s “true talent” seems rather arbitrary, but well in line with the way the game has evolved from a multi-dimensional pastime into one that is obsessed with the Three True Outcomes. What if a pitcher’s talent lies in getting batters to hit slow ground balls or lazy flies? I realize that, yes, fielding those BIPs is usually up to other players, but does this mean that the only capability worth having or caring about is that which is played out in a head to head confrontation? I can’t buy it, but that’s pretty usual with me.

    So while I agree that, yes, these were great seasons, I don’t agree that FIP is all that makes them great, or that there aren’t other seasons in the same general range of greatness that derive from a broader balance of skills.

    • 9

      Greg Maddux might agree.
      But so would Fausto Carmona.

      • 14
        bstar says:

        So would Tim Hudson and Tom Glavine.
        And R.A. Dickey and Phil Niekro.
        And Gaylord Perry and Jim Palmer.
        And Billy Wagner and Joe Nathan.
        And just about every deadball pitcher who toed the rubber.
        And most sinkerballers, Vaseline-ballers, and sandpaper-ballers.
        Or basically anyone who’s m.o. is inducing weak contact.

        The biggest outlier of all is Mariano Rivera. A 2.21 career ERA, but only a 2.75 FIP. Are we supposed to believe that Mo’s true talent level is that of a 2.75 ERA pitcher? I just can’t buy that.

        • 43

          I agree, bstar.
          The 7 seasons above are certainly among the all-time greats.

          But I can’t buy into FIP as the definitive pitcher’s stat.
          A definitive pitcher’s stat cannot show preference to a specific pitching style. Pitching is an art as well as a science, and every man has his own approach to bettering the man with the bat.

          And pitchers are not victim or beneficiary of their defense or ballpark. They are an aspect of it.
          Baseball is about adjustments, and part of what makes an excellent, well-rounded pitcher is the understanding of what is behind him and the ability to find the advantage.

          A list of the greatest seasons has to include Pedro 2000.

          Pedro in ’99 blew people away, yes.
          Pedro in ’00 was just unhittable.

          His SO/9 went down 1.4
          And his H/9 somehow went down 1.5 (only Nolan and Tiant better)
          Consider that.

          The Sox’ defense improved somewhat.
          18 less errors, roughly same range factor.

          That’s one less error every nine games.

          And it didn’t seem to benefit Pedro.
          He had 7 unearned runs is ’99.
          Only 2 in 2000.

          8 more homeruns?
          Is a homerun more exclusively a pitcher’s fault, compared to a line drive in the gap?

          Using FIP to make the case that these were the greatest seasons of alltime is going too far. The seasons that mathematically describe this combination of three metrics, that’s it.

          But using the same result-specific stat to compare power pitchers and pitch-to-contact guys is like saying okay, Infinite Jest and Jitterbug Perfume – which was the better book?

          • 54

            Voomo et al, I wouldn’t say that FIP is “the definitive pitcher’s stat”. There’s probably no such thing, and if there ever is, it will probably be some blend of a FIP-based WAR and FDP outcomes (which would reward volume and park/era effects). I *am* saying that FIP is at least as descriptive, and certainly more predictive, than ERA. A lot of fans know about Gibson’s 1.12, and they shouldn’t necessarily forget that number, since it carries some meaning, but I think Pedro’s 1.39 is more meaningful.

            As to Pedro ’99 vs. 2000, each was amazing in its own way, much like Gooden’s ’85 and ’86. I called out ’99 because I created this list starting with FIP, rather than starting with what I thought was the best season and choosing numbers to justify it. I think it’s possible that a slightly improved defense was more responsible for the reduction in his ERA than anything he did better the second year, and that’s what FIP tells us.

    • 11
      JDGentile says:

      Maybe I accidentally skipped over the part where the author said “FIP is all that makes them great” and “the only capability worth having or caring about”?

      • 31
        no statistician but says:


        And where do I say that I’m attributing these points of view to Bryan O’Connor? I’m arguing against the view that it’s a mistake to focus excessively on, let’s call it, the One True Stat, whether it is WAR, ERA+, OPS+, HR, BA, or the National GDP. I don’t believe in the philosopher’s stone, either.

        • 55

          no stat, I get that there’s a community that seems to be putting a lot of emphasis on all-in numbers like WAR and VORP, but I’ve never heard a worthwhile baseball person say that we should focus on One True Stat. To me, the SABR revolution was about recognizing that we can do better than the stats that were developed 100 years ago and don’t tell us much about the way baseball games are won today. The more stats available, the better informed we can be (unless one of those stats is saves).

          • 77
            no statistician but says:


            We’re on the same page.

            1) Your choice of a first post was meant to be a little provocative, I’m sure, and I sometimes comment to . . . stimulate response, let’s say.

            2) But if you watch this site long enough, you will see comments that indicate, perhaps thoughtlessly, that WAR in particular is the arbiter supreme. Who deserves the MVP or the Cy Young? It’s easy, the guy with the highest WAR, and those benighted voters of the past who didn’t understand, for example, that Williams was twice the player DiMaggio was in 1947, were obviously engaged in a vendetta or overcome with terminal stupidity, NY bias, or a lack of insight into the mighty power of the base on balls. Even I was too young then to remember that particular season, but I’ve read a number of comments over the years about what happened, and just now, when I look at the actual vote and see that, not just Joe D but Joe Page got far more 1st place tallies than Williams, and George McQuinn got the same, I have to wonder if there isn’t something a little hollow about Williams’ performance that year, his high WAR notwithstanding. Maybe he should have gotten the award, I don’t know, but the people who were there watching didn’t think so in large numbers. My point: it is very difficult to pound just a tiny grain of skepticism into the granite heads of true believers.

            And I suppose this is another of those comments meant to provoke response. I’ll tell my wife to duck.

    • 13
      birtelcom says:

      The problem, nsb, is that there is little evidence of a consistent “talent” for getting hitters to hit slow ground balls and lazy flies. The great knuckleballers have been able to do it, and a very, very few pitchers, but for the vast majority of pitchers, it’s the FIP that tends to be more constant from year to year and from team to team as opposed to he non-FIP elements of ERA. Those non-FIP elements tend to jump around from year to year in a way that suggests they are heavily reliant on the whims of chance and varying defensive performance. Modern sabermetrics doesn’t look to FIP to be contrary or perverse, but only because that’s where the evidence takes it.

      • 22
        Hartvig says:

        Robin Roberts was remarkably successful at inducing F-8’s for a number of years but he also led the league in strikeouts a couple of times during that stretch as well.

  7. 10
    birtelcom says:

    Koufax, Gooden and Newhouser all burned out quite early, and Pedro was done somewhat early too. Is it possible that super-extreme FIP performance is unusually risky for a pitcher’s long-term health? Is it possible that a pitcher may survive longer if he lets his fielders pick up more of the burden (albeit running the risk that those balls in play will sometimes fall in for hits?). Certainly we’re looking at an extremely small sample of players here, and I’m not reaching conclusions or advocating any position here, just throwing out an hypothesis. It’s undoubtedly true that pitching itself is an inherently risky business and any sample of pitchers, of whatever style, will have its short-career casualties. But your list does have some of the most famous abbreviated-career guys ever, which does invite some curiosity.

    • 48
      nightfly says:

      That theory is part of why Doc Gooden talked, in spring training 1987, of changing his approach and pitching more to contact… he wanted to save the wear and tear on his arm by getting guys earlier in the count.

    • 56

      More strikeouts + more walks = more pitches. Since the pitch count is a relatively new phenomenon, a lot of high-K pitchers probably burned out in the past because they were asked to make more pitches than other pitchers who threw a similar amount of innings. Then again, Nolan Ryan and Stave Carlton and Roger Clemens struck out a lot of guys and issues some walks and lasted forever.

      The Twins might agree with your hypothesis though.

  8. 12
    Jason Z says:

    Great read Bryan. Thanks.

    In regards to Koufax retiring young, I recall some scenes
    from Ken Burns’ documentary on Baseball that were of Koufax’s
    retirement press conference.

    He said that he didn’t know the long term effects of cortisone
    shots and pain killers. But that he had a long time to live and
    wanted to be healthy.

    Koufax had to undergo a grueling routine that involved icing that
    arthritic elbow after every start.

    He said that he was high much of the time he pitched those last couple
    of years and just didn’t want to continue.

    Roger Clemens possible comeback has me thinking about a Koufax story
    from spring training at Dodgertown in the mid 80’s.

    Koufax who was at Dodger camp as a special instructor dominated the
    Mike Marshall and Greg Brock led Dodgers to such a degree that the
    players were giddy with the prospect of Koufax making a comeback.

    He never did, but that story always left me wondering, here is a guy 20
    years retired and yet he dominates.

    Makes me believe that if Clemens does come back, he will be successful.

  9. 15

    A wonderful debut, Mr. O. I’m trying to come up with words for Pedro. I’m just not sure we realized what we were actually seeing at the time.

  10. 18
    Ed says:

    A few near misses:

    Steve Carlton: 2.01 in 1972

    Luis Tiant 2.04 in 1968

    Those were the two closest misses I could find. Anyone got others?

  11. 19
    Ed says:

    Here’s an interesting article from Beyond the Boxscore that computes a “normalized FIP” based on league averages. If you weren’t in awe of Pedro before, you will be now:

  12. 23
    bstar says:

    “Fielding-Independent Pitching is a better way to evaluate a pitcher’s true talent than ERA.”

    But is it better than ERA+, or RA9 after adjusting for team defense and strength of opponent?

    Here’s Rally(Sean Smith) summarizing why he doesn’t bother with FIP:

    “…Here’s the way I look at it: Take a pitcher, let’s call him Ricky. I know what league, park, opponents faced, and in front of what defense Ricky pitches in. Taking all of that into account, I estimate that a replacement level pitcher would give up 5.5 runs per game. If Ricky allows 5.25 per game, then he performed barely better than replacement level. That’s the question I am interested in, and the answer is satisfactory to me.

    If Ricky does this while striking out 200 and only walking 35, that means nothing to me about his value last year. Though it probably we can expect better results in the future…”

    In other words, Rally sees FIP as a predictive tool for the future but not a useful one for evaluating seasonal performance.

    Even Tom Tango, FIP’s inventor, has said recently on more than one occasion that it’s probably best to average FIP and ERA+ instead of just using FIP, and, although he prefers FIP for single seasons, that ERA+ is better over the long term(career).

    • 34

      In other words, Rally sees FIP as a predictive tool for the future but not a useful one for evaluating seasonal performance.

      Exactly the way I feel about it. Since I’m mostly interested in baseball history, I only use br/rWAR. I’ll peek at Fangraphs if I see my team just acquired someone.

    • 35

      I agree. From a historical perspective, runs allowed are more important to me, especially since over time luck tends to even out.

      • 59

        I’m certainly not a “FIP is right and ERA is worthless” guy, but I think you guys are selling short the descriptive value of FIP. FIP tells us what a pitcher did. ERA tells us what a pitcher and his defense did. Some elements of FDP (fielding-dependent pitching) are more attributable to the pitcher than the fielders, and we’ll be better able to evaluate pitchers when we can start to isolate these factors, but so much of FDP is fielding and sequencing and dumb luck that I don’t feel comfortable calling it the pitcher’s fault.

        Here’s my main beef with ERA: if Zack Greinke induces a weak grounder that Erick Aybar bobbles, ERA lets him off the hook for that runner and any runs that wouldn’t have scored had he made that play. But if Coco Crisp leads off the next inning with a gapper to left-center, but Mike Trout saves a triple with a diving catch, ERA doesn’t put Crisp on third and decide if he would have scored if the A’s were given another out to work with. Using RA mitigates this issue, but there’s still so much noise in there that wasn’t necessarily the pitcher’s fault.

        • 74
          bstar says:

          Here’s one of my many beefs with FIP, Bryan. Pitcher A gives up three doubles off the top of the wall, a foot from being a home run. Two runs have scored, there’s a man on second, yet according to FIP nothing has happened even though no defense in the world could have prevented these doubles. Pitcher B gives up three home runs hit in the same spot that manage to pass over the outfield wall by a few inches. According to FIP, all three of these batted balls are 100% the pitcher’s fault.

          So doubles high off the wall=100% pitching independent
          HR that barely clear the wall=100% pitching dependent

          Why are FIP literalists OK with this when it’s obviously wrong?

          singles, doubles, triples, groundouts, and flyouts = 100% pitching independent. Do you really believe this, Bryan?

          Try this exercise. Watch very intently every at-bat of the next baseball game you have the time to really scrutinize. Come up with a number of batted balls, whether hits or outs, that actually had anything to do whatsoever with the quality of the defense. Cans of corn, infield pop-ups, sharply-hit singles and doubles that Willie Mays could not have caught need not apply. I think you’ll be quite surprised at how low the number actually is.

          We simply can’t call all singles, doubles, triples, ground outs, and fly outs “luck” or “100% pitching independent”. That’s not baseball. It’s FIP baseball, which resides in a fantasy world.

          • 78
            Howard says:

            Any stat can be made to look ridiculous if it is scrutinized enough. A hitter can drive in 200 runs w/a .000 BA. Or he can bat 1.000 w/o ever getting on base. More to the point, a pitcher can give up an infinite # of runs but if they all happen after a two out error those runs don’t count in his ERA.

          • 89

            Good call, bstar. I should have been ready for this. Your example is every bit as damning of FIP as mine was of ERA. I certainly never said that I believe FDP outcomes are 100% pitching independent. I just believe they’re closer to 100% than to 0.

            Countless studies (here’s one done by Bill Petti and expanded by Tangotiger: have concluded that FIP is a better predictor of year 2 ERA than ERA or just about any other metric. If that doesn’t speak to true talent, how is it true?

            I acknowledge that there are FIP-beaters out there and that there always will be. I’m leaning toward Cueto or Dickey (the low-RA guys) over Kershaw or Gonzalez (the low-FIP guys) for NL Cy Young this year, though that may change. I just think that ignoring empirical data about FIP theory because it doesn’t feel right when you watch games is counter to SABR’s “search for objective knowledge about baseball”.

          • 92
            bstar says:

            OK, Bryan, but is it a better predictor than ERA+ or RA9 adjusted for defense, park, and quality of opponent?

            I’m aware that FIP is slightly more predictive of next year’s performance than raw ERA. But is that really enough to call it “a better indicator of true talent”, especially when so many different types of pitchers outperform their FIPs on almost a yearly basis? To me, your link simply suggests that FIP is a good predictive tool of future performance, which is what I think it should be used for.

            Are we to believe that Tim Hudson’s “true talent” is actually half a run worse every single year of his career than his ERA suggests? What type of pitcher does a RA9-based metric under or overrate? None.

            Tango tells us that FIP and ERA will eventually converge, but I gave plenty of examples above that suggest they don’t always do that.

            And i DO think actually watching the games should be a part of sabermetrics, absolutely. How many erroneous conclusions have been reached about FIP and BABIP because people weren’t watching the games to see if these conclusions actually made sense?

            Don’t the recent FDP metrics introduced by Fangraphs strongly suggest that FIP alone was missing the mark? I read in your article about it (great read, by the way) that Fangraphs almost did change their fWAR calculations; I wish they had.

            Here’s the link to Bryan’s article about the new Fangraphs FDP metrics:


          • 94

            Re: bstar’s #92:
            If you agree that FIP is more predictive than future ERA, how is it not a better indicator of true talent? What makes a guy with a lower FIP likely to allow fewer runs next season than a guy with a higher FIP if he’s not more talented?

            I think you’re exaggerating the number of FIP-beaters. You named several outliers, and there are several more. But there are thousands of pitchers throughout baseball history whose ERAs have fluctuated from year to year despite relatively consistent FIPs. Are they better pitchers in the years when their ERAs are lower, or are they beneficiaries of some things working in their favor?

            Loving this thread, by the way. Thanks for linking to my FDP piece.

  13. 24
    deal says:

    How much of the 1999 Red Sox defensive issues were due to Nomar Garciaparra?

    I doubt it was due to him. A look at their roster and it is hard to see that team in the playoffs, yet they won 92 games.

    What was the FIP for Saberhagen?

    • 60

      Saberhagen had a 3.30 FIP and a 2.95 ERA in ’99, so the defense worked well for him. That’s only 119 innings though, so it doesn’t tell us much about the relative strengths of the defense behind the two pitchers.

  14. 25

    Great first post, Bryan. I only hope I can be nearly as good in my debut.

    It’s pretty much unquestionable to me that Pedro had the greatest peak of any pitcher in history. In fact, Mariano Rivera is the all-time leader in ERA+ at 206, and of course, he racked up that incredible number by pitching 1200+ innings over one and two-inning stints.

    From 1997 to 2003, Pedro’s ERA+ was 213 over 1400+ innings (199 starts, 2 relief appearances). Amazing.

    • 26

      By the way, Sandy Koufax’s best one-season ERA+ was 190, in 1966, his final year.

    • 28
      bstar says:

      Good luck with your debut, Dan. I think it’s also pretty clear Greg Maddux has the second greatest peak in history. Over a seven-year period(same as Pedro) from 1992-1998, Maddux’ ERA+ was 190 over 1675 IP.

      The similarities between these two peaks continue. Right in the middle of those seven-year periods, in years 3 and 4, Pedro and Maddog had their two best seasons. The ERA+’s over those two years are bizarre in their similarity:

      Pedro 1999-2000 265 ERA+
      Maddux 1994-1995 265 ERA+

      Among those 14 player seasons, you’ve got 9 ERA titles, 10 ERA+ titles, 9 WHIP titles, 7 Cy Youngs, and 13 top-5 Cy Young seasons.

      • 33

        That’s pretty amazing too, bstar. I guess we probably shouldn’t overlook Walter Johnson’s 185 ERA+ from 1912-1919 (8 seasons). Not that ERA+ should be the only criterion, but it’s a nice snapshot.

        I was expecting Randy Johnson to be in the discussion, but not when we’re looking at 7-8 year runs. Same goes for Koufax.

    • 69
      Lawrence Azrin says:

      How few years can “peak” be – two? three? four? five?

      Without looking anything up, here are some other great peaks through the 20th century:
      – Cy Young: 1901-03
      – Christy Mathewson: 1903-05
      – Walter Johnson: 1910-1918 (or 1911-12 within that)
      – Grover Cleveland Alexander: 1915-17
      – Lefty Grove: 1928-1932 (or 1929-1931 within that)
      – Carl Hubbell: 1933-36
      – Bob Feller: 1938-1940
      – Robin Roberts: 1950-55
      – Sandy Koufax: 1962-66
      – Bob Gibson: 1967-70
      – Tom Seaver: 1969-71
      – Dwight Gooden: 1984-84
      – Roger Clemens: 1986-87, 1990-92
      – Greg Maddux: 1993-96
      – Pedro Martinez: 1997-2003 (or 1999-00 within that)

      Brett Saberhagen might have made this list if they told him every years of the 80s was an “odd” year…

      Just a fun exercise, no attempt at definitive completeness.

  15. 29
    Phil Gaskill says:

    > the National League hit just .237 in 1968, the lowest mark for either league in baseball history

    Not quite, sorry. (a) The NL’s BA was .243 in 1968, BTW. (b) The AL’s BA the same year was .230, which I *think* might have been the lowest in history. (c) The AL’s BA each of the two *previous* years (1966 and ’67) was less than .237. I didn’t have time to look up a lot more seasons; I did try 1908, but both leagues hit .239 that year, as did the AL in 1972..

    • 38
      Richard Chester says:

      That .230 BA in 1968 is the lowest. It’s easy to find. Go to BR, click on seasons, locate League Index at the top, click on American League Batting and then click on BA twice to sort in ascending order.

    • 61
      • 67
        Phil Gaskill says:

        You were reading the wrong column. I find that particular chart to be very poorly presented, and I can only sympathize with anyone misreading it. Here’s the 1968 line:

        .230 (.23011) .243 (.24313) .237 (.23669) 1968

        The order of the columns is, unfortunately: AL, NL, MLB, Year.

        I didn’t even have to look the AL figure up, BTW, if you’d like additional “proof”: I was there (lived in Washington, D.C. from ’67 to ’74; was even there for the last .500 Washington team before this year), and I well remember that “disgraceful” .230 league batting average. (This entire paragraph to be read with a smile. 😉

        • 70
          Lawrence Azrin says:

          I distinctly remember at the end of the 1968 season that some writers called Yaz’s Al-leading .301 BA a “disgrace”, particularly in New England, where his .301 BA/ 23 HRs/ 74 RBI was seen as quite a comedown from his Triple Crown-winning previous year.

          I guess they never heard of OBA, OPS, or times on base, all of which he led the AL in. Not to mention OPS+, runs created, Adjusted Batting Runs and WARP. That is in large part due to his MLB-leading 119 walks.

  16. 32
    Andy says:

    This is a pretty awesome debut post, Bryan. Great work.

  17. 36
    Brent says:

    OK, so I have a question about FIP that I am sure this community can answer, one that has always bothered me. In some ballparks in MLB there are places a ball can be hit that a fielder could never field it, yet it is not a HR. I am not talking about screaming liners in the gap or right down the line here, I am talking about balls that are not reachable by any human being, no matter where they play. We exclude HRs from FIP because no one can catch them, why not exclude what I will describe as Monster Balls (high enough off the Green Monster that no one can reach it), although that isn’t the only park with such a phenomenon (just the most famous one). Is it just that the Monster Balls are too statistically insignicant to matter or too hard to count?? Could the ignoring of Monster Balls be slightly changing the stats for BoSox pitchers, at least their home stats, and anywhere else that such a wall exists?

    • 39
      JDGentile says:

      HR’s are included in FIP. Monster Balls, and any other type of Ball in Play, are not. Maybe you mean BABIP?

      • 40
        Brent says:

        I guess I wasn’t clear. A ball that hits 35 feet up on the Green Monster (or 20 feet up, for that matter), is just as “out of play” for purposes of recording an out as a HR, and, logically, in any formula trying to calculate that, should be treated the same way, shouldn’t it?

        • 44
          JDGentile says:

          Ok, I understand now. You wouldn’t want to charge the pitcher the run value of a HR for those Monster Balls, though, but the run value of a double? Interesting idea…

          It would be largely insignificant, I imagine, but for pitchers who spent half their IP at Fenway it could be meaningful.

          • 46

            Yeah, but some of those balls would be flyball outs in other parks. Isn’t xFIP about as close as it gets to park-adjusted FIP? Well, at least, in that it adjusts the pitcher’s home runs allowed based on fly ball rate.

          • 47
            JDGentile says:

            If that’s your concern, Dan, then just use FIP- or FIP+ which would adjust for park (and league incidentally which plain FIP doesn’t do).

          • 79
            bstar says:

            But xFIP is even more faulty than FIP, because it assumes that no pitcher has HR-suppression skills. Matt Cain has proven that wrong, much to the chagrin of many a Fangraphs writer.

            Even Tom Tango has said, paraphrasing, “After a few years, you can basically throw xFIP out the window”.

  18. 37
    birtelcom says:

    Hitters facing Pedro in 2000 had an overall slash line of .167 BA/.213 OBP/.259 SLG/.473 OPS (an OPS+ of 18!). Compare that slash line to that produced thus far in 2012 by the ninth spot in the batting order in the NL (mostly pitchers and pinch-hitters): .176 BA/.226 OBP/.248 SLG/.474 OPS. So a team facing Pedro in 2000 was producing on offense what an entire lineup, top to bottom, of NL ninth-spot batters would produce today.

    • 45
    • 62

      It’s true- mMost AL teams used lineups of nine pitchers against Pedro that year. Batters couldn’t figure him out, so managers thought pitchers might have a better shot.

    • 71
      PP says:

      When you put it that way, does anyone else remember James paraphrasing Stephen Jay Gould by saying the reason Pedro might be so good is due a few minor things that add up exponentially (or something like that)? There has to be some explanation for a year like 2000, and the others.

      • 72
        Lawrence Azrin says:


        Yes, I do. Was it in the NBJHA?

        • 84
          e pluribus munu says:

          p. 865
          (That’s all I have to say, but the site won’t let me submit a reply that short – what encouragement to old windbags like myself!)

          • 86
            bstar says:

            That’s currently in storage for me, can you quote a few pertinent passages if you’ve got the time, epm or somebody else?

        • 88
          PP says:

          Right, found it in vol 2 (still pissed I threw away vol 1). The analogy was based on Gould’s saying the reason the human mind was so much more complex than a monkey’s was due to some small advantages in their brains that makes humans capable of doing “vastly more.”

        • 91
          e pluribus munu says:

          My edition is just one volume (2001):

          (More or less quoting): “7 factorial is 5040; 10 factorial is 3.6 million. . . Stephen Jay Gould once expressed the thought that, when we finally understand the difference between the mind of man and the mind of a monkey, it will turn out to be something simple like this – that man’s mind is not vastly different . . . but is capable of vastly more because some small advantages create enormous differences by making combinations with one another and with other parts of the mind. I think of that in connection with Pedro. How can he be *so much* better than other pitchers? His fastball is good, but there are 20 or 50 people in the league who throw just as hard. His curve isn’t better than anyone else’s, his control isn’t. But he is vastly better in toto because he has some additional factors – his ability to change his arm angle, to change speeds on all pitches without losing control – which interact to make geometric combinations.”

  19. 53
    Nick Pain says:

    I don’t imagine pitchers like Pedro and Maddux will have any trouble getting into Cooperstown. It seems to me pitchers have been seriously underappreciated in recent balloting. The only predominantly starting pitcher inducted this millenium was Bert Blyleven, and we all know what an effort that was. Guys like Kevin Brown, David Cone, Orel Hershiser, and Brad Radke haven’t even sniffed The Hall with pretty stellar numbers.

    • 81
      Howard says:

      Three of those guys don’t have overwhelming #s and the fourth, Brown, is suspected to be a PED user. Had they avoided injury there is no doubt Cone and Hershiser would be in and if he hadn’t retired early and there were enough sabermetric guys voting Radke would have had a shot.

  20. 58
    Jacob says:

    Great post, loved the tension of the countdown (even though I kinda knew the answer).

    Just wanted to say here that I remember every game of Pedro’s ’99 season like yesterday. I was such a huge fan of that man, I’ve never rooted for anybody like I rooted for him back then.

    • 82
      Howard says:

      I know what you mean. I still remember Ron Guidry’s 1978 season like that. I saw every one of his starts that year except for the two weeks my mom and dad forced me to go to Italy with them. Parents……

  21. 90
    e pluribus munu says:

    I love this post, but I don’t agree with placing undue weight on FIP. It isolates a certain type of capability, one which is extraordinary, but I don’t believe that great pitching should be understood in this way. So I’m with nsb and birtelcom in my general thoughts.

    The rules of the game give pitchers a set of tools; fielders are among the most useful. Measuring a pitcher according to his ability to achieve success by excluding those tools celebrates one model of pitching at the expense of others, and, as noted, the model has too often been self-consuming.

    Maddux seems to me to offer a more sustainable model. I think the results he achieved in ’94-’95 are ultimately a more impressive combination of baseball skill and art, since they left him with enough gas in the tank for 200 additional wins. (Of course, he had the advantage of knowing that at least one of his fielders would always be the best in the game at his position.) For me, the ultimate perfect game would not be 81 straight strikes, it would be 27 straight first-swing dribblers – to the mound, if Maddux is on it.

    That said, I need to go light the incense at my Koufax shrine, say ten Hail Pedros, and thank Bryan for a great post.

    • 93
      bstar says:

      Regarding Maddux, this is why I’ve never taken much stock in Game Scores. It too heavily weighs strikeouts but does not capture the efficiency of pitching a two-hit shutout on less than 85 pitches but only a few K’s. There needs to be some sort of efficiency rating added to Game Scores to fix this.

      • 96
        e pluribus munu says:

        That’s a good point, bstar. On the other hand, I’ve always liked Game Scores because unearned runs count. I think ERA, for all its virtues, is probably a less accurate reflection of pitching quality than RA, for all its injustice. The Game Score formula of two virtue points for every injustice seems a Solomonic balance – but, as I assume you’d agree, only a fool would submit to an injustice just to put a couple of K’s in the bank.

  22. 99
    John Autin says:

    Bryan — Welcome, and congrats on your first HHS post!

    Although I am not entirely aboard the FIP bandwagon, I enjoyed this post because I, too, mourn the fact that sabermetrics has yet to generate any iconic numbers.

    One thing that’s not clear to me about FIP is illustrated by Hal Newhouser’s 1945-46 seasons. The latter one made your list — even though Prince Hal’s ERA (and RA) were better in ’45, and his HR/9 was much better — mainly due to his 8.5 SO/9 in ’46, almost 40% higher than his 6.1 in ’45.

    However, the AL K rate from 1945-46 went up from 3.4 to 4.3 SO/9. Newhouser in ’45 had about 1.8 times the AL average SO/9; in ’46 it was about 2 times. So his relative K rate in ’46 was only about 11% higher than in ’45. Does FIP address differences in league context?

    Also, I’m curious where the best years of Dazzy Vance rank in the FIP pantheon. In 1924-25, especially, Vance lapped the field in SO/9, and had good control plus relatively few HRs.

    • 102
      birtelcom says:

      According to fangraphs, while Dazzy’s ERA went from 2.16 in 1924 to 3.53 in 1925, his FIP was a much more consistent 2.64 in ’24 and 2.69 in ’25. He led the majors in FIP by a large margin in both seasons (Carl Mays was second in 1924 at 3.19 and Pete Donahue was second with 3.23 in 1925), but was 17th in the majors in ERA in 1925. For the two seasons combined, Vance just ekes out a lead for best ERA in the majors (with a 2.79 to Eppa Rixey’s 2.82 for the combined 1924-25), but is dominant in FIP (2.66 to Dolph Luque’s second place 3.29 for the combined seasons).

    • 103
      bstar says:

      JA, yes the constant in the FIP equation is figured so that league average FIP = league average ERA every year. Since scoring dipped from 1945 to 1946, ’46 was a better year to put up a good FIP. In fact, if you weed out the dead-ball era years, I find 5 of Bryan’s 7 seasons listed to be in the top 9 all-time lowest FIP years in history:

      Here’s the 9 lowest FIP seasons after 1919 based on league average:

      1. 1968 (Gibson is third on Bryan’s list)
      2. 1972
      3. 1967
      4. 1943
      5. 1946 (Newhouser is seventh)
      6. 1963 (Koufax is fourth)
      7. 1971 (Seaver is sixth)
      8. 1942
      9. 1965 (Koufax is fifth)

      I wonder if FIP+ (or the Fangraphs-preferred FIP-) would have been a better measure of FIP greatness.

      I tried to come up with a list of the seven best FIP- seasons of all-time, but I’ll be damned if I can figure out how to finagle the Fangraphs’ single-season leaderboards.

      John, just go to Fangraphs and type in Dazzy Vance in the search box to get his FIP totals. It’s listed on his home page.

      • 105

        The top seven FIP- seasons since 1919 (on the fangraphs pitching leaderboard, choose the advanced tab, choose your seasons, and click the “split seasons” box:

        1. Martinez ’99 (30)
        2. Martinez ’00 (46)
        3. R.Johnson ’95 (46)
        4. R.Johnson ’00 (46)
        5. R.Johnson ’04 (48)
        6. Gooden ’84 (49)
        7. Martinez ’03 (49)

        I think there’s some evidence here that (1) FIP is a repeatable skill, and (2) it’s easier to blow away the league-average FIP in an era when Ks are up and HRs are up but you’re not giving them up. The top 60 FIP- seasons include only two before 1959: Newhouser in ’46 and Harry Brecheen in ’48

  23. 101
    tag says:

    I think the reason few know iconic SABR numbers is because, basically, they’re derivatives. I mean, a good chunk of the general population can tell you what Apple’s stock is trading at (or near), but only an options trader can tell you what AAPL December 13 calls are going for. Even though SABR numbers may represent a closer estimation of player performance and are “better,” they remain esoteric because they are calculated from more common and familiar stats that retain their hold on the public mind.

    My question with regard to FIP is a logical one. If missing bats is a skill pitchers possess because of their velocity, variations on it, movement and control, don’t we logically also have to conclude that the ability to induce batters to contact the ball less than optimally through the same means is also a skill, whether it shows up easily in statistics or not? Now I imagine this second skill would be a far more rarefied one because it would stem more from the control/change of speed/movement elements (and mastery of them would have to be off-the-charts) than from the velocity element, but after watching guys like Fergie Jenkins, Rick Reuschel and Greg Maddux most of my life, I’m pretty certain it exists. And I think some recent research is verifying that it does.

    I think the reason for Maddux’s awesomeness was that he was able to control late movement like perhaps no one else in history.

    And I think the explanation for Pedro’s awesomeness is that he combined the best virtues of both the velocity and superior control pitchers. He blended Gibson’s/ Seaver’s heat with the Maddux/Reuschel outside-corner-at-the-knees and in-on-fist pinpoint precision. That’s gonna make for some freaky numbers.

    • 104

      Tag, I agree that the ability to induce weak contact is a skill. The problem is that weak contact doesn’t correlate directly with outs, as strikeouts do (with the exception of the rare wild pitch). So a sinkerballer may keep the ball on the ground, leading to fewer homers (which would show up in FIP) and fewer line drive extra base hits (which wouldn’t show up in FIP). Striking out hitters is a repeatable and effective skill. Inducing weak contact is a skill some pitchers can repeat that may or may not lead to run prevention, depending on various factors outside the pitcher’s control.

      As I’ve said many times in this thread, FIP isn’t perfect. I just like it more than ERA.

      • 107
        tag says:

        Bryan, I agree completely, and if my post hadn’t become so long I would have raised that very point. Any time the ball is put in play good/bad things (depending on POV) can happen and lead to runs. But I think a pitcher has to weigh that risk against the risk (in the short term) of throwing too many pitches in attempting to get strikeouts and subsequently tiring/being pulled in favor of a bullpen pitcher who may not be near his equal and (in the long term) of blowing out his arm before his time.

  24. 111
    Brooklyn Mick says:

    In 2012 no starter comes close to sub-2.00 FIP. The search criteria needs to be narrowed to a minimum 60 innings pitched to find a sub-2.00, and there lies Aroldis Chapman with a 1.41 FIP.

  25. 112

    I know this if off topic but I’m looking into starting my own weblog and was wondering what all is required to get set up? I’m assuming having a blog like yours would cost a pretty penny?
    I’m not very internet savvy so I’m not 100% certain. Any recommendations or advice would be greatly appreciated. Cheers

  26. 113
    Marylyn says:

    Magnificent goods from you, man. I have understand your
    stuff previous to and you are just too excellent.
    I really like what you have acquired here, really
    like what you are stating and the way in which you say it.
    You make it entertaining and you still care for to keep it sensible.
    I can not wait to read far more from you. This is really a terrific web site.

    my page: air impact wrenches 3 4 (Marylyn)

  27. 113
    Marylyn says:

    Magnificent goods from you, man. I have understand your
    stuff previous to and you are just too excellent.
    I really like what you have acquired here, really
    like what you are stating and the way in which you say it.
    You make it entertaining and you still care for to keep it sensible.
    I can not wait to read far more from you. This is really a terrific web site.

    my page: air impact wrenches 3 4 (Marylyn)

  28. 115

    Hey! Someone in myy Mysppace group shard this site with us sso I came to
    take a look. I’m definitely loving the information. I’m book-marking and will
    be tweeting this to my followers! Terrific blog and suhperb design annd style.

Leave a Reply

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