At WAR with Pedro

Somebody asked me this question on Twitter: If I could have any pitcher from any time pitch one game (say a seventh game of the World Series or the ubiquitous “pitching for your soul” scenario”) who would I choose?

I immediately typed: Pedro. 1999.

This is always my fallback position. Back in the Trivial Pursuit days, my mother would guess “Babe Ruth” on pretty much every sports question. She has actually become much more knowledgeable about sports, in part because of this mess of a blog, but back then it was always “Babe Ruth,” even on, you know, billiards or horse racing questions.

And that’s how I am with Pedro Martinez’s 1999 season. Any baseball pitching question can be answered, somehow, by: Pedro, 1999. I would actually like to answer ALL questions that way. When I go fill up gas, and the little pump screen asks: “Cash or Credit” I’d love to be able to type in: Pedro, 1999.

Anyway, the choice lit up the Twitter lines with the expected objections — what about Bob Gibson in ’68 or Randy Johnson in 2001 or Walter Johnson in 1913 or Sandy Koufax in 1965.* You could make an argument for those and a couple dozen more — Carlton in ’72, Gooden in ’85, Grove in ’31, Hershiser in ’88, Mathewson in ’08, on and on.

*Am I the only one who gets kind of annoyed when people put some sort of finality stamp at the end of their opinions? You know what I mean by finality stamp — someone will not just say “Sandy Koufax in 1965 was quite sprightly.” No, they will say something like “Koufax. 1965. End of story.” Or: “Gibson. 1968. The end.” Or: “Carlton. 1972. Period.” Or: “Old Hoss. 1884. Goodbye.”

What are these emphatic termination words supposed to achieve? I mean YOU put those words there, right? I didn’t miss some mediator coming in and ending declaring your viewpoint supreme, did I? It’s not like you pulled Marshall McLuhan out of nowhere to confirm your opinion … YOU confirmed your opinion. How does that mean anything? Is this like the Internet equivalent of taking off your shoe and clomping it on the table like a gavel? Stop doing that. It’s stupid. Period. End of story. Goodbye.

Anyway there was one alternative to Pedro 1999 suggestion that I found interesting for a completely different reason.

The suggestion: Pedro in 2000.

This post is not actually about Pedro Martinez, not specifically, but about WAR. As I assume everyone reading this blog knows, there are two prominent variations of the statistic “Wins Above Replacement.” There is Baseball Reference WAR. And there is Fangraphs WAR. Best I can tell when it comes to everyday players, the two systems are fairly similar — any real variations on players’ totals probably comes down to how defense was calculated.

But the two calculate pitcher’s WAR differently and this might be seen mostly clearly in Pedro Martinez’s 1999 and 2000 seasons.

Martinez made the same number of starts and threw roughly the same number of innings both seasons, which is helpful comparison purposes. In 1999, Martinez threw 213.3 innings. In 2000, he threw 217 innings.

The other numbers, though, are quite different:

1999: 23-4, 2.07 ERA, 5 complete games, 1 shutout, 160 hits, 313 Ks, 37 walks, 9 homers.
2000: 18-6cfi, 1.74 ERA, 7 complete games, 4 shutouts, 128 hits, 285 Ks, 32 walks, 17 homers.

OK, before diving in, here is what Baseball Reference WAR says:

1999: 9.7 WAR
2000: 11.7 WAR

So Baseball Reference has Pedro’s 2000 season worth two extra wins.

Here’s what Fangraphs WAR says:

1999: 11.9 WAR
2000: 9.9 WAR

And it’s almost precisely reversed — Fangraphs has Pedro’s 1999 season worth two extra wins.

Obviously both seasons are all variations of awesome and we’re just picking between Reese’s Peanut Butter Cups and Mini Reese’s Peanut Butter Cups. And you probably know exactly why the two Pedro Seasons are calculated differently but let’s go step-by-step here.

Baseball Reference WAR values the 2000 season more because Pedro Martinez gave up fewer runs and fewer hits. It’s a bit more complicated than that, but that’s at the crux of things. Martinez’s ERA+ in 2000 was 291, which is the record for a season. In 1999 his ERA+ was merely an otherworldly 243.

So that’s at the heart of Baseball Reference’s process — Pedro Martinez gave up 11 fewer runs in 2000 (largely because the league hit an almost unbelievable .167 against him, 38 points less than the year before) and that meant it was a clearly better season.

Fangraphs WAR, meanwhile, doesn’t deal with ERA. It deals with the three things that Fangraphs believes a pitcher can control: Strikeouts, walks and home runs. In 1999, Martinez struck out an obscene 13.2 batters per nine inning (just behind Randy Johnson’s 2001 record) and he hardly walked anybody and, perhaps most overlooked, he gave up NINE HOME RUNS all season. Both of these seasons were smack in the middle of the Selig Era, when home runs flew like confetti, and to give up nine homers all year …

Well, let’s look at the top five in the AL that year in homers per nine innings:

1. Pedro, .380
2. Mike Mussina, .708
3. Freddy Garcia, .805
4. Omar Olivares, .831
5. Jamie Moyer, .908

That doesn’t look very close, does it?

Anyway, of the three things at the heart of the Fangraphs process he did two of them (strikeouts, home runs allowed) better in 1999 and the third (walks) was more or less a wash. So that’s why Fangraphs thinks 1999 was a clearly better season.

What makes this cool, though, is that it’s a great way to decide exactly which kind WAR speaks loudest to you. Which season do YOU THINK is better? If you think the 2000 season was better, then you are probably a Baseball Reference person. If you think 1999 — you’re Fangraphs.

I asked Tom Tango what he thinks and, as usual, he came up with an interesting way of looking at things. Looking at it another way, the question in play is this: How much control do you think a pitcher has on balls hit in play — yes, we’re crossing back to the famous BABIP (Batting Average on Balls In Play).

If you think a pitcher has COMPLETE CONTROL over balls in play then you will naturally think that Pedro was a better pitcher in 2000 when his ERA and hits allowed were much lower. The BABIP numbers could not be more stark.

– In 1999, despite his dominance, Pedro allowed a .325 batting average on balls in play — which was actually the FIFTH HIGHEST in the AL.

– In 2000, Pedro allowed a .237 BABP — which was the LOWEST in the AL.

So, if you believe a pitcher has complete control over balls put in play then you will believe that Pedro Martinez learned a whole lot between the end of the 1999 season and the beginning of 2000.*

*I believe it is this Pedro gap, by the way, that helped inspire Voros McCracken to come up his theory about pitchers not having control of balls hit in play.

OK, but if you think a pitcher has ZERO CONTROL over balls in play, then you will definitely believe that Pedro was a better pitcher in 1999 and was just a whole luckier in 2000 (or had a team that played much better and smarter defense, which is in a way the same thing for a pitcher).

What Tango says — and I concur — is that it’s likely neither absolute is true. It’s likely that pitchers do not have complete control on balls hit in play, and it’s likely that pitchers are not entirely powerless.

“Since reality is somewhere between the two … we get into our conundrum: must we take a 0/100 approach to everything we track?” Tango asks. “Or, can we start to give partial credit? … No one likes the idea of partial credit, because it implies a level of precision that we can’t possibly know.”

Tango comes down closer to the side that a pitcher has limited control over balls in play. I again agree. I think there will still be studies and thought experiments that get us closer to that relationship between pitching and defense, but right now I lean just a touch more to the Fangraphs side. I think Pedro pitched a little bit better in 1999 than he did in 2000. That 313-37 strikeout to walk ratio is just absurd. Those nine home runs allowed, wow. I don’t think he was a full two wins better. But one game — we’re talking one game — I’m taking that Pedro Martinez in the middle of the Selig Era who didn’t let the ball in play much, who always kept it in the ballpark and who was good for 13 outs a game on his own.

And Tango? Well he says Baseball Reference and Fangraphs give us the extremes … and the answer, almost certainly, lies in the middle. And this is why Tango, when looking at Baseball Reference WAR, at Fangraphs WAR will split the difference.

This would make Pedro’s 1999 and 2000 seasons almost EXACTLY EVEN.

Which, if you think about it, is a good way to end this. Period.

120 thoughts on “At WAR with Pedro

  1. George Evans Light

    Once played trivial pursuit with a Canadian college classmate. The Canadian sports edition. So we each had automatic answers to speed things up. I got Gretzky he got La Fleur for Hockey I got BC Lions he got the Argos for Grey Cup et cetera et cetera et cetera.

    Reply
  2. Christopher Tice (@Ctice90)

    Your 1999 and 2000 Pedro stats are off. His 23-4 record was in 1999 and his 1998 record was 19-7 and his 2000 record was 18-6.

    Otherwise, great article. I love Pedro and would read about him all day. He is the answer to the original question. PERIOD.

    Reply
  3. hack

    babip is a function of how well the pitcher executes the game plan. the fielders are positioned as though the pitcher will locate his pitches. when he locates according to game plan the batters will usually hit to where the fielders are. a high babip must be a function of missing location and batters being able to hit it where they ain’t. a low babip must be a function hitting location and forcing the batters to hit it where they’ve been predicted to hit it. the pitcher has quite a bit of (but by no means total) control of babip. at least that’s my theory.

    Reply
    1. SB M

      I find this hard to believe, since one of the key observations about BABIP is that it varies significantly from year to year for the same pitcher. Are we really to believe that a pitcher is very good at executing his team’s game plan one year, but the next year he’s awful at it, and the year after that he’s great at it again?

      Reply
      1. hack

        i might put that down to batters and teams making adjustments from year to year. then pitchers making adjustments in response. and so on.

        Reply
      2. Pat

        hack (apt name), please read Voros McCracken’s piece before you opine on what BABIP means. Short answer, you have wandered really far off the res, here.

        SB M, you may be interested in Tom Tippett’s further exploration of the BABIP trends identified by McCracken. One upshot is that certain pitchers do in fact demonstrate a durable ability to prevent hits on balls in play, but that the effect is really small compare to the larger truth of what McCracken identified. (Also, maybe just me, but it looks like a lot of the guys who can prevent hits on balls in play are also supreme strikeout artists, meaning pure velocity and location are doing both at once.)

        http://www.baseballprospectus.com/article.php?articleid=878
        http://207.56.97.150/articles/ipavg2.htm

        Reply
  4. PhilM

    Woody Allen appreciates the reference. . . .

    Season by season, I can understand the Fangraphs approach. The thing I can’t get over is the career evaluation of Carl Hubbell. It’s WAY too low (period, end of story), and I can’t figure out if it’s the context of low walks) or what — so I just defer to Baseball Reference’s calculation.

    Either way, I’m pretty sure your soul would be safe with either incarnation pitching for you.

    Reply
  5. cass

    Unfortunately, Baseball Reference WAR for pitchers has a weird adjustment based on the defense of the players. It basically assumes that the fielding was at the same levels for all pitchers on a team and adjusts accordingly. But the samples are far too small to rely on that, so it has some quirky results. If fielders performed worse when a given pitcher started, which happens all the time, bWAR will blame the pitcher for that. It’s similar to how two pitchers on the same team might get a different level of run support making pitcher wins inaccurate for judging pitcher performance.

    Luckily, there is a stat out there that does what Joe claims that bWAR does – it’s RA9-WAR on FanGraphs. It’s based just on runs allowed, nothing else. Some combination of fWAR and RA9-WAR probably yields the best answer.

    Reply
    1. NRJyzr

      Reading this piece, I also thought of RA9-WAR.

      FWIW, Pedro’s RA9-WAR numbers for the season in question:
      1999: 10.0
      2000: 12.2

      Such as it is.

      In part due to the odd babip number for Pedro in 99, I tend to just say “Pedro at his peak” for such answers.

      Reply
      1. 18thstreet

        I wonder if the Red Sox defense had a great year in 1999, also. Trot Nixon was, at times, a really great defender (until he wasn’t). And the 2000 club had a rotation at third base of Wilton Veras and Manny Alexander. Jose Offerman was a year older in 2000 (as were we all, but that man did not age well.

        Someone (someone who who isn’t me) could look at the putouts on balls in play during Pedro’s starts in 1999 and 2000 and see if there is any discernible pattern.

        My guess? Manager’s decision.

        Reply
  6. invitro

    Great post, and I’m convinced, for now. There’s gotta be a way of getting the %luck in BABIP within like 20%, it seems. I don’t know anything about BABIP results. Are there any pitchers that have an exceptionally high (or low) one over a number of years? Or hitters? What about SLG on balls in play for pitchers?

    Reply
  7. Craig from Az

    To me, baseball reference WAR tells you who had the better season. Fangraphs WAR *might* do a better job of predicting who might have the better NEXT season. You might say Pedro’s 2000 season was only better because he got lucky. So what? It’s still a better season.

    Reply
    1. jposnanski Post author

      But this is the argument, isn’t it? What Fangraphs might argue is that Pedro was the better pitcher in 1999 but the RED SOX had the better/luckier result in 2000.

      Tango offered another example. What’s the better pitched game?

      Clemens 20K, 0 walk, 1 run game.
      Armando Gallarraga’s imperfect game: 1 hit (not legit) 0 runs.

      You could argue that Galarraga’s had to be the better game because he allowed 0 runs. But did he PITCH the better game? Such is why the argument, I think, goes pretty deep.

      Reply
      1. BobDD

        So Fangraphs will say Clemens, and BR will say Gallarraga? But pitcher A is the one that I would expect to more likely continue his dominance.

        Reply
      2. Matthew Clark

        In this case I actually go Galarraga. He didnt just pitch a perfect game. He pitched a 28-out perfect game. That last out must have been the hardest out he ever got.

        Reply
      3. Spencer

        For the purposes of the question “one game for your soul” I think you want Clemens to replicate his performance rather than Galarraga.

        Clemens struck out 20. That’s leaves only 7 batters with a chance to get a hit. You’re in pretty good shape at that point.

        Galarraga, on the other hand, struck out very few batters (I forget the amount, 4 maybe?) That’s well over 20 balls in play. Way too many for me to be comfortable with my soul.

        Reply
    1. Darrel

      When I read the term Selig Era it just about made me throw my laptop through the window. I love Joe but how he continually rationalizes, forgives, sucks up to(pick one) those who actually took steroids and then throws a not so subtle dig to the guy who happened to be the commissioner at the time infuriates me. Bonds, Clemens, Sosa, McGwire et al knowingly, willingly, purposely cheated the game and Joe wants to canonize them while disparaging an old man in a suit for their behavior. Let’s put the majority of the blame where it belongs which is on the players.

      Reply
      1. DjangoZ

        I’m with you. Joe is probably my single favorite writer in the last few years, but this part really drives me crazy. Oh well, everyone has their hangups. My only explanation is that he loves baseball so much and to really grapple with the PEDs would derail almost any story he would write about this era. So instead he makes fun of it and tap dances over it.

        Reply
      2. Spencer

        @Darrel

        I think Joe’s stance is that it’s tough to blame individual players for trying to compete in an era when a significant number of their peers were on PEDs.

        At that point you blame the culture and rules of the sport (PEDs weren’t tested for)

        That falls on Selig and the owners. They turned a blind eye to what was going on in the name of profit.

        You can disagree with it but you can’t claim that Joe is being inconsistent on this.

        Reply
        1. Darrel

          At the risk of violating the period, end of story annoyance Joe mentioned earlier I have only two words for those who want to blame Selig and the owners for not instituting drug testing in the 80′s or 90′s. Those two words: Don Fehr

          Reply
  8. Steve Adey

    So take a survey of some fans who never remember statistics. Give them a choice of those two seasons and ask the question: “Which one pitches for your soul?”

    Reply
    1. Dave

      I’ve never had the time to sift through what data are available, but it’s seemed to me that BABIP, while good for first order knowledge, doesn’t cover the whole question and that pitchers do have some control, a good bit of control in fact, as to whether the hit ball is, say, a blooper to between short and left or a rope to left. Whether the former ends up as a hit or an out will depend on where the fielders where playing, how good a jump they have, where the sun is, and how fast they are. The rope on the other hand, while still dependent to a degree on those factors, is less so. Thus, what makes the hit a blooper or a rope? The batter obviously–how he normally swings at a ptich in that general area and whether he changed anything on that particular swing. But, a lot depends on exactly where the pitch was. If it was intended to miss the plate, break late say, by an inch or so and it’s there, it’s more likely in my mind to be a blooper–fat part of the bat isn’t on it (more many, many batters). If however it misses over the plate by a inch or more, then it’s more likely to be hit with the fat part of the bat and be a rope. And, it also depends on whether the batter was fooled, by whatever amount, by the location and/or the speed, or its lack.

      A guy like Maddux in my mind, who had great command of where his pitches went, changed his speeds, and whose pitches had lots of movement would more likely end up with the blooper. Pedro at his peak would be the same.

      While BABIP tends to be very similar for all pitchers (well, not Matt Cain and some others who tend to sit around the 260 stat most of the time and not for the guys we don’t remember as they got shelled in their 3 innings of cups of coffee), I’ve wondered whether “slugging percentage” shows a lot more spread than BABIP (It’s not slugging percentage because home runs wouldn’t be part of it) and whether the percentage of hits that are line drives show a lot more spread. Some day maybe I’ll get the time to pull a bunch of this together. I throw it out because maybe a reader has that type of time to study it.

      Reply
  9. AlbaNate

    Every time I look at Pedro’s insane FIP of 1.39, I think it must be a typo. I mean, how can it possibly be so low? I can’t really choose between the two, but I lean ’99 because of that FIP.

    Jamie Moyer: It’s hard to believe that Mr. I-Broke-Robin-Roberts-Record was actually one of the best pitchers at suppressing Home Runs in 1999.

    Reply
  10. bl

    When in doubt, always Vote for Pedro.

    My Trivial Pursuit default answer for any “English Actor” question was David Nevin.

    Reply
  11. John

    I have never understood the argument for Pedro’s ’99 as better than his ’00.

    To recap:
    In the season I prefer (2000), he gave up 20% fewer hits, and a few fewer walks. He actually gave up fewer earned runs, even though he allowed nearly double the home runs. If you didn’t hit it out of the park, you didn’t score. Since one of the things the pitcher can’t control is the dimension of his park, I don’t hold that against him. In fact, he only gave up 36 extra base hits all year, basically one extra base hit per game pitched. In ’99, he gave up 47 base hits.

    Pedro in 2000 was better than Pedro in 1999, and as a Yankee fan, I can’t believe he didn’t win the MVP.

    Reply
    1. BobDD

      because 29 starts

      that’s why some choose the other pitcher’s Joe named over Pedro as best season

      Pedro has the best rates; Gibson and Johnson substantially more games and innings. Like Joe said, there are good arguments on both sides which is what makes it such an enjoyable debate. Like Pedro ’99 vs ’00.

      Reply
      1. bl

        but in a single game scenario number of starts is irrelevant. all that matters is likelihood of dominance in that one game.

        Reply
        1. Ian R.

          Sample size matters when you’re trying to predict future performance, even if it’s future performance in one game. Gibson’s 300-inning sample from 1968 has more predictive value than Pedro’s 200-inning sample from 1999.

          If we think along those lines, the best argument for picking 1999 Pedro might actually be the existence of 2000 Pedro. It clearly wasn’t a fluke because he went ahead and was just as dominant (albeit in different ways) the next year.

          Reply
      2. Ian R.

        A somewhat related point: Gibson and Johnson didn’t just start more games; they averaged more innings per start. Certainly that factors into the decision – would you rather have a starter who’ll give you 7 shutdown innings and leave the last two to your bullpen, or a guy who might not be as dominant on a rate basis but will take you into the ninth?

        Reply
        1. bl

          Points well taken. Still, I have to choose Pedro. No player ever made my baseball soul happier. And what’s the point of saving the soul if it’s not happy.

          Reply
      3. NRJyzr

        Based on conversations I’ve had elsewhere….

        One potential reason folks mention Koufax and Gibson is the not-insignificant number of people who cannot grasp the idea that a 1.74 ERA in 2000 is better than a 1.12 ERA in 1968.

        Context eludes them.

        Reply
        1. Elfego Slaughter

          It’s fall of ’99 and I’m playing in a golf club championship ..I’m in a twosome the other being an ex San Diego Padre pitcher not a year or so removed due to arm injury where there was a favorable($) ruling given him. In essence I’m 50yrs. old and he’s a 20 something with deep pockets ‘n deeper talent.
          As luck would have it we get to walk the fairways together to talk baseball. I let him know in no uncertain terms that in my time the ‘SHOW’ was Koufax ‘n Clemente(the show being something that resembled the Lakers). I’m somewhat removed from current knowldede of baseball ,but proceeded to ask him who the best pitcher in the bigs is today? He confidently responded with “Pedro… everything he brings moves”.
          I proceeded to tell him it appears basketball and football are improving,but baseball is declining…the baby boomers were the last generation to play baseball almost to exclusion and that we dwarfed everything before and after with monstrous nos. of participants. Baseball peaked during my twenties was an inherent conclusion and the ‘bigger,stronger faster scenario was not applicable in this game of skill. Was Pedro better than Koufax? Quite possibly! Was there anything that resembled Clemente in ’99 or ’89 or ’79 or ’69 or even 2009? I simply never saw it.

          Reply
    2. Richard Aronson

      I completely agree. The argument about pitchers allow home runs (as the basis for Fangraph’s WAR) basically says “pitchers allow batters to hit the ball hard”. Well, hard hit balls are likelier to be singles, doubles, and triples as well as homers. So if a pitcher suppresses homers, they are suppressing all solid contact. Pedro in 2000 had the best pitching year ever by ERA+, and by a huge margin. So I would go 2000.

      Reply
      1. Mr Furious

        Absolutely. Both years are Pedro at his peak, but those two performances in particular are among the most memorable in his career, and not to be overlooked in the same incredible 1999 season is Pedro coming out of the bullpen in the decisive ALDS Game 5 tied 8-8 in the 4th and pitching six shutout innings, striking out 8 Indians to carry the Sox on to the ALCS.

        1999 has the clear “domination intangibles” edge over 2000.

        Reply
  12. Geoff

    What are these emphatic termination words supposed to achieve? I mean YOU put those words there, right? I didn’t miss some mediator coming in and ending declaring your viewpoint supreme, did I? It’s not like you pulled Marshall McLuhan out of nowhere to confirm your opinion … YOU confirmed your opinion. How does that mean anything? Is this like the Internet equivalent of taking off your shoe and clomping it on the table like a gavel? Stop doing that. It’s stupid. Period. End of story. Goodbye.

    This is actually way less annoying to me than people who add “Just sayin’” to the end of their comments for some unknown reason.

    Reply
    1. BobDD

      I was willing to be very emphatic on my top 50 list when I turned it in, but as time goes on I am learning to de-emph at a rapid rate. Damn accountability!

      Reply
    2. Chris H

      I think “just sayin’” means “I might be wrong about this, in fact I probably am wrong about this, or at least it’s certainly debatable, but I’m not going to engage in discussion about it.” Which I agree is a pretty darn annoying sentiment.

      Reply
  13. Pete Grossman

    What I’ve seen elsewhere before: the latter half of 1999 and the first half of 2000. That “season” has the best of it all.

    Also: what will it take to get you to use ERA- instead of ERA+? It took me a bit to get used to it, but after converting I can’t imagine going back. It’s so much more clear and useful a term. Instead of some odd percent better than league average, you know exactly how it compared to league average. ERA- is 40? 40 percent of league average. Done. Doesn’t have any problems with crazy multipliers when the ERA of a guy gets very low. If you have an ERA of 0.50 when the league is at 5 (say you’re a reliever who had a dominant year) then your ERA+ would be 1000. What does. That even mean? No idea. ERA-? 10. Perfectly clear.

    Reply
    1. Spencer

      @Pete

      Both numbers say the exact same thing

      Era+ is 1000. That means ten times better than league average

      Era- is 10. One tenth of league average.

      They’re just inverted. Era+ is more common and more familiar

      Reply
      1. Pete Grossman

        Phil largely captured my point. The idea is that ERA is a stat that we are familiar with being best when it’s as small as possible. ERA+ makes it best to be as large as possible. And the ERA- figures simply offer a clearer interpretation of what actually happened. If a player has an ERA- of 28, I know exactly what this means. But what then is the corresponding ERA+? I can’t do that accurately I’m my head. A little less than 400, I guess? But is that guy really an additional 60% better than someone who has an ERA- of 33? That doesn’t make sense to me. He’s slightly better, and both are awesome. But ERA+ inflates those differences.

        ERA+ isn’t bad, but there’s a reason that Fangraphs writers and others have switched to using ERA-, if they use one of these numbers in their writing. Yes, it’s not as familiar to money. I said as much. I simply think it’s much better and clearer, especially given that ERA is best when it’s lowest. At least everyone can agree on that.

        Reply
        1. Brett Alan

          Well, my take on it is that I don’t so much want to compare ERA+ to ERA as I want to compare it to OPS+. I honestly don’t know how fair a comparison it is, but to me when, for example, I was deciding whom to vote for in Joe’s ill-fated BR Hall of Fame, one of the things I looked at was how good those numbers were. 100 means you were average, 130 means you were considerably above average. If, like, Pedro, you had a career ERA+ of 154 and five seasons above 200, well, WOW. (Only Mariano has a better career ERA+, and obviously a reliever isn’t a fair comparison here, not to deny that Mo’s numbers are great.) To me those big numbers make sense, and to me again it’s useful that it at least seems to be comparable with OPS+.

          As long as I’m posting, might as well comment on the article itself…I agree with Joe. I can totally buy that pitchers don’t have nearly as much effect on balls in play as we might once have though. But basing WAR ONLY on Ks, BBs, and HRs seems to me a bridge too far. It just strikes me as too limited. (Consider, for example, that it seems to really undervalue knuckleballers.)

          Also, as a grammar teacher I just have to complain about the headline for the NBC version of this article reading “Who Do You Choose?” I choose he, of course!

          Reply
    2. PhilM

      This points up what is confusing about ERA+ (which I nonetheless really like and use in a lot of my calculations): the comparison element. ERA of 0.50 isn’t “ten times better” or “1000% better” than average. 100% better than average would be an ERA of zero — it’s totally (100%) better than surrendering runs. Other measures (hitting home runs, e.g.) are essentially limitless, so Babe Ruth can be 250% or whatever above the league average, but ERA can’t go below zero: at most it can be “100% better.” It’s a problem with the denominator: an ERA+ of 200 is actually 50% better than average [(200-100)/200] — 2.25 is 50% below (“better than”) 4.50. So an ERA+ of 1000 is (1000-100)/1000 better than average, or 90%. ERA- of 10 conveys this 90% level more directly.

      Reply
  14. Frank Blackman

    Your number crunching makes a compelling case that Pedro may well have been the best ever for an entire season. But that is not your initial premise. How can you be certain he could have handled the pressure of a Game 7, since there are no examples to prove that argument?
    And it is not just speculation that the pressure is exponentially greater. It has to be similar to the pressure that ultimately separates competent setup men from great closers.
    Has a pitcher faced a more difficult situation than Sandy Koufax did in 1965? All the Dodgers asked him to do was come back on two days rest to pitch a Game 7. He responded by throwing a three-hit shutout to win the World Series.
    That amazing performance would seem to provide the answer to your question.

    Reply
    1. cass

      What separates competent setup men from great closers is skill. What separates great setup men from great closers is very little.

      There’s a stat called leverage index, though, if you’d like to measure this stuff. In general, great setup men make great closers. An inning is an inning.

      Reply
    2. murr2825

      I have always wondered about that “pressure” question. The normal pitcher that reaches the pinnacle of a Game Seven World Series start doesn’t get there in a spaceship, brought in from another planet. He has likely faced “big game” pressure situations in Little League, Babe Ruth, high school, college and even trying to make his first big league roster in spring training.

      In all those previous situations, he was pitching what felt like the biggest game of his career. Sure the crowds and the limelight were not the same (excepting something like the finals of the LL World Series) but the pressure he felt inside his far less emotionally mature self was every bit as daunting.

      My point, I guess, is that by the time even a 24-25 year old pitcher toes the mound in a game seven, he has long ago figured out how to stay cool and within himself, or he wouldn’t be there.

      Reply
    3. NRJyzr

      While it’s not Game 7 of the World Series, we *do* have Pedro’s performance in an elimination game, 1999 ALDS Game 5. Six innings of no hit relief on one day of rest isn’t too bad, especially not when it’s done with the tiny strike zone pitchers had in 1999. ;)

      Reply
      1. gcuzz

        And what about Game 7 of the 2003 ALCS vs. the Yankees? While not a World Series game 7, you can argue that the pressure in that game was equally as big as any World Series game when you consider the teams involved, the craziness and sheer intensity of some of the previous games in that series, and what was at stake in terms of staring down 85 years of heartache for a franchise.

        Reply
      2. Herb Smith

        I’m glad someone brought up that Pedro relief performance. I’m a Cub fan who has repeatedly watched Kerry Wood’s 20K one”hit” performance (which Bill James’ “Game Score” stat ranks as the greatest 9-inning pitching performance of all-time)…but even I will admit that Pedro’s hitless relief game was the most unreal, dominating pitching show of all-time.

        Or, at least that I’ve ever seen.

        Reply
    4. Jeff A.

      Well, though it obviously isn’t the same pressure as a WS game 7, Pedro’s performance at the 1999 all star game is pretty compelling evidence that in a highly visible, much touted stage against the game’s elite hitters, he was up to the task.

      Reply
    5. Bill Caffrey

      First of all, you’re arguing from a sample size of one, which is just silly.

      But beyond that, the fact that Koufax was on 2 days rest adds to the degree of difficulty, but not the pressure. In fact, I’d argue it decreases the pressure, since if Koufax goes out and gets bombed it can easily be written off as “2 days rest was just too much to ask.”

      Reply
    6. Mr Furious

      Pressure? In 1999, Pedro came into Game 5 of the ALDS and pitched six innings of shut-out RELIEF!!

      Not saying that outdoes Koufax, but it also answers the question, I think.

      Reply
    1. BobDD

      just wow!

      His Slg% against in 2000 was only .259
      His RA/9 was only 1.82
      and his WHIP was an all-time best 0.737

      so those numbers support 2000 being his best year
      ultimate dominance for rate stats anyway
      only possible better would have to be outside MLB

      Reply
  15. tombando

    IT’S JACK MORRIS MORON HE PITCHED TO THE SCORE BETTER THAN ANYONE THIS SIDE OF DAVID LEE ROTH.

    YOU KNOW IT.

    Reply
  16. Rick Rodstrom

    Period. End of story. Goodbye.

    That actually explains my dislike of WAR in general, whether bWar or fWar.

    Other stats are stats—accumulations of things like wins or hits or RBIs. And other stats contain very simple ratios, hits compared to at-bats, or runs compared to innings pitched, or what have you.

    Now you may prefer some stats over other stats, but they are all limited in some way. They provide food for argument, but they don’t offer conclusions.

    WAR makes conclusions. Based on a complex series of judgment calls as to how much value to assign to every facet of the game, WAR has the audacity to say, this guy is the most valuable. This guy is the best player in the game. This guy is the best player in history.

    Period. End of story. Goodbye.

    Reply
      1. Rick Rodstrom

        Thank you. Perhaps i should have said, other stats are just stats, while WAR is not so much a stat, as it is a conclusion. By accepting WAR as the definition of excellence, you are accepting one persons biases about what constitutes value and undermining your own independent judgement. It’s going beyond what a stat can legitimately do, because it is addressing an issue with so many variables that assigning numbers to every possible contingency is impossible, yet it pretends to be something solid, absolute, definitive, scientific. This whole blog sometimes becomes an argument not over baseball, but over WAR—that is, over one man’s opinion of what constitutes value. WAR, rather than starting a discussion, consumes the discussion, and more often than not, ends the discussion.

        When Tom Tango says about BAPIP

        “Since reality is somewhere between the two … we get into our conundrum: must we take a 0/100 approach to everything we track?” Tango asks. “Or, can we start to give partial credit? … No one likes the idea of partial credit, because it implies a level of precision that we can’t possibly know.”

        Now mind you, this is about a measurable event—a ball put in play. Overall value contains so many variables that they don’t even have names, much less numbers.

        It’s like ‘The Meaning of Life—the stat (TMOL)”. John Doe’s life has a TMOL value of 9.4. Jack Smith’s life has a TMOL value of 5.3. Therefore, John Doe’s life is more valuable than Jack Smith’s life.

        That, anyway, is my opinion. You may disagree. But at least it’s my opinion, and not piggy-backing on somebody else’s.

        Reply
        1. JimV

          Actually WAR (and those that calculate it) makes absolutely no conclusions – it is simply a statistician’s way of providing additional context to the pure counting stats – a more advanced formula than hits/at-bats for example. It is people who don’t understand that all stats are “99% accurate 95% of the time” who end up drawing less than ideal conclusions.

          Stats, when used to draw conclusions, always have uncertainty factors which need to be thought about before making any conclusions. In all cases, you need to use more than one (advanced and counting stats included) to make more reliable statements – and even then there is no foolproof way to make a 100% accurate assessment like “who is the best”.

          Reply
          1. DM

            JimV,

            Well said. A metric is not the same as a conclusion (“Excuse me, officer. It seems my metric has gone rogue and generated a conclusion all by itself”). A metric such as WAR provides information. In this case, it helps summarize estimated contributions across multiple dimensions of the game. People can use it to help them analyze.

            I feel like a broken record, but it’s not intended to be a definitive answer in and of itself. No one involved in its creation has claimed that, as far as I’m aware. It’s an approximation, it’s an estimate.

            Maybe it needs a new name. Maybe instead of translating runs of various types into wins, we could leave it expressed as an estimate of runs generated and runs saved through hitting, pitching, baserunning, and defense, and we could call it “Estimated Net Run Contribution”. We could even eliminate the damn decimal. We could have a big marketing campaign complete with signs showing WAR with a big “X” through it.

            Yeah…..ENRC….That’s the ticket

          2. Rick Rodstrom

            I only wish what you were saying was true, but in practice, WAR is rarely used that way. For instance, in his post on Albert Pujols, Joe Posnanski asks “How many times in baseball history has there been someone who was not only the game’s best player but CLEARLY the game’s best player” by using WAR to determine this. And frankly, I’m not so sure that those players were clearly the best, though they are certainly in the conversation. Like, I think Stan Musial was a better player than Ted Williams (an opinion shared, incidentally, by Bill James, who ranked Musial higher than Williams). I’m not so sure that Willie Mays was better than Mickey Mantle, or even Hank Aaron. Again, you can argue with me, but Joe’s post takes these positions as a given, based on WAR. It’s a stat that lends itself to this way of thinking (you can especially see WAR absolutism in the yearly MVP arguments). Used in small doses, WAR provides an interesting perspective, but more often than not, it ends up being abused as a trump card that invalidates other arguments.

            Like, there’s a little known algorithm that some people use to determine which of Pedro’s seasons was better: they watched the games.

          3. Chad

            Rick-

            Where exactly did James rank Musial higher than Williams. I know it’s not in the NHBA. Musical is one of the best baseball players of all-time, but I see no valid argument anywhere for him being ranked higher than Williams.

          4. Adam

            On page 396 the “Bill James Historical Abstract” (paperback 1988), James wrote, in rating Musial highest among left fielders for peak and career value over Williams, “Look, I am not saying anything at all negative about Ted Williams. . . . But if I had to choose between the two of them, I’d take Musial in left field, Musial on the basepaths, Musial in the clubhouse, and Williams only with the wood in his hand. And Stan Musial could hit a little, too.”

          5. DM

            Hi Chad,

            I can actually your question to Rick…..he’s right about James (at one time) ranking Musial above Williams, but you have to go back about 30 years, to the ORIGINAL Historical Abstract from 1985. In that publication, Bill James had Musial edging out Williams as the #1 left fielder at all time, both in peak value and in career value. Musial #1, Williams #2.

            In the New Historical Abstract from 2000, he reversed that, and had Williams (#7 overall) as the #1 left fielder, with Musial (#10 overall) as the #2 left fielder.

            In essence…..I would conclude that he has very little to separate them in his mind. In the original ranking, he gave Musial an edge in the field, in the clubhouse, and on the basepaths, and Williams an edge in hitting, and basically said if he had to choose, he’d take Musial. I assume in the New Historical Abstract 15 years later, I think he went more or less with the results of his Win Shares system.

          6. DM

            Rick,

            Just as another follow up on WAR and your belief in how Joe’s using it…….

            If Joe really believed in the absolute nature of WAR, his top 100 list would more closely resemble the all-time WAR leaders. In othe words, Blyleven would be top 40, Nap Lajoie would be top 25. But, he has them way further down the list than that. Phil Niekro is 33rd in bWAR. I’m guessing he’s not going to make it at all. Craig Biggio was 134 in bWAR, but he made Joe’s top 100. And many, many more examples……

            Does he leverage the information in WAR? Absolutely. But he’s clearly taking other things into account when coming up with his rankings.

          7. JimV

            One more comment from me regarding Rick’s comments about Joe’s use of WAR for the greatest player over 10-year periods. Given the amount of analysis required to go through every 10-year period starting in 1900, if you were using several different stats, you would still have to find some way of balancing those stats across different periods/positions, and even then there isn’t a 100% accurate way of doing that. Since we are merely intending to provoke discussion (as the whole top-100 list is supposed to do), and show that Pujols was, for a period of time, clearly the best player in baseball (in Joe’s opinion since that is also what this list is about), only WAR provides a simple (albeit approximate) mechanism to do that.

  17. Jason Cline (@jasoncline13)

    Nothing to do with anything, but, I’ve been catching up on Game of Thrones. I hear these arguments over Fangraphs WAR vs BB Ref WAT, ’99 vs 2000, and I hear Littlefinger ane Varys debating/arguing. It’s cracking me up. Pedro, in either of these years, would be a fine choice in the ‘pitching for your soul’ game. Personally, I’m leaning toward Fangraphs.

    Reply
  18. Steve

    Pedro had the best pitched game I ever saw. It was September 10, 1999, against the Yankees. He hit the first batter (who then got caught stealing), and then had 17 strikeouts and zero walks for the rest of the game. Chili Davis closed his eyes and hit a solo home run in the second inning, but that was it for the Yankees. Pedro threw a one-hitter, facing only 28 batters. I never saw such domination. It was as if Pedro was throwing a whiffle ball, the way it moved. The Yankees (with a World Champion lineup) were utterly helpless against Pedro that night. Really, they should have just sat in the dugout all night. I believe some of the Yankees said they had never seen such dominant pitching before. Here is the box: http://www.retrosheet.org/boxesetc/1999/B09100NYA1999.htm

    Reply
    1. Rico Petrocelli

      I was at that game
      It was unreal. Second deck, the place es rocking. I am a sox fan and lovingit. The stands were full of Dominicans who were dancing in the aisles, more and more with each strikeout. I think Yankees had the lead on that long Davis homer so there was pressure too. A big night.

      Reply
      1. Paul Zummo

        Yet another person who was at that game. I am a Mets fan, so basically watched the game as an impartial third party observer, though for some reason was rooting for the Yankees and actually tired to get the Yankee fans to start pepping up.

        My other memory of that game is spending an unholy amount of money on beer (this was three months after I graduated college). That might explain my rooting for the Yankees that night.

        Reply
    2. Ken

      I remember watching that game on TV. I was and am a huge Yankee fan, but that’s one of my favorite games of all time. Never seen anything like it.

      Reply
      1. Steve

        The Yes Network broadcasts the Yankees games, and they sometimes show old games during rainouts and days off, but they only show games when the Yankees win. I wish they would show some of the classics when they lose. That September 10, 1999 game, man oh man, I still talk about it to this day.

        Reply
  19. Stephen

    It’s been hinted at above, but what I’d want to know is Pedro and what relief pitcher?

    Martinez averaged about 7 1/2 innings per start in those two years. Koufax is ’65 averaged over 8, Gibson in ’68 a whopping 8.9 innings per start.

    Not trying to take anything away from Martinez, who was of course great, but the games are 9 innings. In ’99 Pedro started 29 games and was pulled before the eighth inning began 14 times–just under half of the starts. Obviously part of that is the expectation for pitchers of the time and the way his manager chose to use him–but I don’t know that we can give him credit for things he wasn’t generally asked to do. For me, the bottom line is he did not generally pitch full games, and I want someone pitching for my soul who can–or at least, who did.

    So if it’s a 7-inning game, Pedro makes a lot of sense. Otherwise, sorry, I need to know who’s behind him.

    Reply
      1. Stephen

        Are you saying that Martinez would have pitched deeper into more games if he’d been pitching from a higher mound? What’s your evidence for that? Gibson had exactly as many CGs in ’69 as he did in ’68, when the mound was higher. Niekro and Jenkins increased their CG totals from ’68 to ’69. It’s not that simple or straightforward.

        Anyway, you didn’t answer the question. Remember, half the time, they took Martinez out before the eighth inning began. This game is for your immortal soul. So: Martinez and what relief pitcher?

        Reply
  20. Ross

    Great post. One thought on how to split the difference is to say that Pedro was a bit unlucky in ’99 and say that he should have gotten at least league average BABIP. I know it’s arbitrary, but I would be interested in seeing someone smarter than I calculate expected ERA, ERA+, and bWAR if he had that league average.

    Reply
  21. Herb Smith

    From what I know, BABIP tends to reward good hitters, and the good ones will consistently put up high BABIP (even though they’ll fluctuate quite a bit from year to year).

    However, there seems to be no rhyme or reason to BABIP for pitchers. As noted above, Pedro had back-to-back magnificent, historic seasons. He was likely an almost identical replica of himself from 1999 to 2000. Yet, in one year his BABIP was about the WORST in the league, and the other year, it was the absolute best.

    Does that make sense? The only “trend” that I’m aware of (in tracking pitcher’s BABIP) is that knuckleball pitchers DO have consistently low BABIP. Which makes sense. But it contradicts the extrapolated sensical thing: that guys like Maddux and Pedro would seem to have superb BABIP. And they do not.

    Reply
    1. Pete Grossman

      That’s not entirely the case – there are a few other indicators that lead to a lower BABIP. One is their ground ball/fly ball tendency. Fly ball pitchers tend to have a lower batting average against, since fly balls make more outs, but they also make more runs. In terms of run production it’s basically a net wash, but it does have an effect on BABIP. A corollary to that is that some pitchers seem to have an additional ability to generate in field fly balls, and this further reduces their BABIP.

      The last factor is something that Mike Fast wrote about years ago in his excellent (it’s basically required reading for a SABRist, I’d say) analysis of the small amount of Hit f/x data that he had access to. You can find it here: http://www.baseballprospectus.com/article.php?articleid=15532. Basically pitchers with higher strikeout rates tend to have a lower average horizontal speed of batted ball against them, and lower batted ball speeds lead to fewer hits. Pitchers then do have some measure of control over the type of contact that batters get off of them. Maybe this leads at least to lower slugging percentages against them.

      Reply
  22. Herb Smith

    Here’s an interesting thought about the guy who said “Koufax, 1965. End of story.”

    1965 was only his THIRD best year, according to Baseball-Reference WAR. And his ’63 and ’66 seasons are quite a BIT better. I think that many fans, when talking about Koufax, can’t help but to include his World Series performance in ’65 in their narrative…the Yom Kippur stand, the 4-hit shutout in Game 5, and the 3-hit shutout only 2 day’s later, in Game 7.

    Now, if we’re including the post-season, then we might have a few other great pitcher seasons to add to our list. Suddenly Christy Mathewson’s 1905 looks pretty good: 31-9, 1.28 ERA, led the majors in WAR, and threw 3 consecutive shutouts in the World Series.

    Tough to top.

    Reply
  23. Richard Aronson

    For what it’s worth: Koufax had two WAR seasons better than Pedro’s 1999, and neither was 1965. I’m picking Koufax in 1966 because it was his best ERA+ season, and more importantly, he had 27 complete games, and as many shutouts as Pedro had in 1999 and 2000 combined (Koufax had 5 in 1966, and a mind boggling 11 in 1963). Pedro might give you a better 7 innings but Koufax is likelier to give you all nine. Plus, Batman.

    Reply
    1. NRJyzr

      Koufax may not have been likelier to give nine innings if he pitched to the postage stamp sized strike zone that was called fifteen years ago. :)

      For the Koufax fans, it may be worth noting his road/home splits; except for his final season, he appears to have been somewhat mortal away from Dodger Stadium.

      Reply
  24. Bothrops Asper

    Tom Tango also showed that the talent/luck “stabilization” point for BABIP is est. 3700 BIP. So if we want to know how much skill a player has, we just need to regress accordingly. The question isn’t if pitchers hat some control over BABIP, but what the sample size needs to be to find it amongst all the noise.

    Reply
  25. John Gale

    I’m not sure if I was the first one who suggested Pedro 2000 on Twitter, but since I can’t find anyone else, I’ll go ahead and take credit for inspiring this post. I’m fairly certain this is the first (and will likely be the only) time that this has ever happened.

    At any rate, Joe is right that my opinion on this completely pegged me as a Baseball Reference guy. I very rarely go to Fangraphs. Not that I have anything against them. I’m sure they do good work. I just feel like Baseball Reference does everything I need.

    In fact, when I saw the huge gap in bWAR, I didn’t even bother looking up the fWAR. That was lazy on my part, and I almost did, but I decided against it. The reason why is because I thought that while there might be a slight difference between the two, surely it wouldn’t be a gap of two full WAR, let alone four. I thought that 2000 would still be comfortably ahead. Whoops.

    As for which is the best season, I think there actually is a clear winner: it’s 2000. Since Baseball Reference and Fangraphs disagree so wildly on WAR, I think we may as well just throw that out as the deciding factor and just take a look at what the other numbers tell us:

    1999: 23-4 2.07 ERA 243 ERA+ 0.923 WHIP 13.2 K/9 8.36 K/BB 6.8 H/9 .4 HR/9
    2000: 18-6 1.74 ERA 291 ERA+ 0.737 WHIP 11.8 K/9 8.88 K/BB 5.3 H/9 .7 HR/9

    Ok, so 2000 has a clear advantage in ERA, ERA+ (the best in the history of baseball), WHIP (best ever) and hits per nine innings (fourth best ever). Walks are about even, though 2000 has the edge there too. 1999 has the edge in strikeouts and home runs.

    Digging a little deeper, 2000 Pedro threw 68 percent of his pitches for strikes, compared to 66 percent for 1999 Pedro. He also had a line-drive percentage (Apparently, 75 percent of line drives go for hits, so it’s an important stat to consider) of 18% in 2000 compared to 22% in 1999. Those two stats suggest to me that he had control over some* of the decrease in BABIP.

    *I did some rough calculations in which I estimated that Pedro allowed 13 more line-drive hits in 1999 than he did in 2000. I’m in the process of looking at his game logs to get an accurate total, but I’ll use 13 for now. He also allowed 19 more doubles/triples (really, just doubles, as he only gave up one triple total over the two seasons) in 1999 than he did in 2000.

    1999 Pedro allowed 32 more hits than 2000 Pedro. The difference in innings pitched is very small, but it looks like those extra 3.2 innings would have been responsible for about two hits. So that increases the margin to 34 for the same number of innings pitched.

    So 19 extra doubles/triples and 13 extra hits off line drives is 32 extra hits over the course of the season, which is exactly the margin between the two seasons. Of course, there is some overlap between the two, but it still explains a lot. I wouldn’t argue that defense has nothing to do with batting average on balls in play, but I’m skeptical that it made up all (or even most) of the difference in line drives and doubles between the two seasons. At any rate, once I finish charting his hits (hopefully, by tomorrow), I’ll post an update of what I discovered.

    I guess it comes down to this: Given the choice, I’d rather have a pitcher who is both dominant *and* lucky* than a pitcher who is dominant and unlucky. I don’t know how much control Pedro had over balls hit into play during those years, but I don’t care.

    All I know is that he allowed significantly fewer hits, base runners and runs in 2000 than he did in 1999, which is the name of the game. I’d gladly trade 1.4 strikeouts per nine and double the (still tiny) chances of allowing a home run in exchange for his lucky horseshoe. Also, the chances of getting a complete game or shutout were better in 2000 than 1999.

    *I think there’s a certain irony in Joe pointing out that Pedro was luckier in terms of ball hit into play, when it’s clear that he was actually much luckier in 1999 in terms of run support. He got 5.68 runs of support in 1999, compared to 4.51 in 2000, which explains why the Pedro actually had a much better winning percentage (.852) than he did in 2000 (.750).

    Then again, this opens up some other questions. Because of the increased run support, it seems that you had a better chance of winning the game in 1999 than you did in 2000. Of course, neither of those teams won the World Series. The 2000 team didn’t even make it to the playoffs. I think I’d rather have a pitcher who wasn’t quite as great playing on the ’27 or ’98 Yankees than a better pitcher playing on the 2010 Mariners.

    But all things being equal, if 2000 Pedro was pitching against 1999 Pedro, and my life was on the line, I’d take my chances with the 2000 edition.

    Reply
      1. John Gale

        I didn’t say that I cared about a pitcher’s W/L record when deciding which one I preferred (in fact, I took the 2000 edition, which had a significantly worse W/L record). My point was that there’s more than one way for a pitcher to be lucky.

        Reply
  26. DjangoZ

    1999 Pedro for me. I don’t think the pitcher has much control over BABIP, maybe a little, but not much, and possibly almost zero.

    Reply
  27. EnzoHernandez11

    As others have mentioned, if it’s one game for my immortal soul, I’m still going with Gibson in 1968. With my soul on the line, I need a guy who can go the distance. I ain’t handing my soul over to Tim Wakefield or Derek Lowe in the eighth.

    Of course, with my luck, the devil will have signed Mickey Lolich and Jim Northrup. Maybe, I’d be safer going down to Georgia and learning to play the fiddle. :)

    Reply
  28. KHAZAD

    I’m for 2000, and it is not really close. John Gale makes a good point about the extra line drives and extra base hits in 1999. He also induced more infield flies in 2000. The anomaly is between the two seasons is really the low percentage of fly balls that turned into home runs in 1999 despite giving up more line drives, more outfield flies, and more overall extra base hits that year.

    Pedro also gave up more balls in play when behind in the count in 1999, and got hitters to put balls in play when he was ahead in the count more often in 2000. He also did the small things better in 2000, having only one wild pitch, 1 sacrifice fly and 10 stolen bases, as opposed to 6, 6, and 21 in 1999.

    Fangraphs looks at the K’s and home runs, but Pedro quite simply was better in 2000 at EVERYTHING else. He also allowed 191 balls in play after getting to two strikes in 1999 as opposed to 171 in 2000, so the ratio of closing the deal instead of allowing contact was nearly identical, he was just better at dispatching the hitter on his pitch earlier in the count, leading to less strikeout opportunities in 2000.

    Reply
    1. John Gale

      All excellent points. I’m not really sure why I was wasting my time trying to figure out how many line drive hits Pedro gave up each year, when BR already has that information (I was about halfway through 1999 when I realized this, and boy did I feel stupid). My estimate was actually off, as Pedro only allowed six more line drive hits in 1999 than he did in 2000, ironically because of an unlucky .809 batting average on line drives in 2000 compared to .757 in 1999.

      The more I think about it, the more I’m amazed at how Fangraphs calculates pitcher WAR. It is so illogical and ridiculous that I don’t know how anyone takes it seriously. It only considers strikeouts, walks and home runs, and it doesn’t consider runs. So Fangraphs appears to be saying that pitchers are not responsible for allowing a run to score unless it was a homer This isn’t like trying to include W/L record in the calculation. It’s *runs.*

      And am I really to believe that a pitcher is 100 percent responsible for a home run, but he bears no responsibility for a booming double off the wall or a rocket into the gap or down the line? Fangraphs seriously expects me to accept that with a straight face? Perhaps I’m missing something, but that’s the way Joe’s explanation makes it sound.

      I’m not sure what the comparable oversight would be for position players. Perhaps not considering hits that aren’t home runs. After all, that’s dependent on defense too, right? What’s more important to a pitcher’s success than not allowing runs? They didn’t convince me that 1999 Pedro was better than 2000 Pedro. All they did was discredit themselves in my eyes.

      Reply
      1. KHAZAD

        Actually, the comparable thing might be to rate hitters WITHOUT counting HRs, Ks, and BB’s as the root of the fangraphs theory is that these are under pitcher control, and everything else is the hitter or defense.

        Even if you believe the hype, the math is fuzzy as it is results based, not really based on any actual value of the events. The multipliers that are used are simply to make it come close to fitting, with a constant that is changed a bit each year to make the make the results look more (leaguewide) like ERA.

        With each new defense independent pitching stat, they kind of debunk prior theories. The first new one basically said “We are wrong about home runs, those are luck as well” and normalized the home run value (in relation to flyballs allowed). Later ones said that pitchers do somewhat control contact, and adjusted (mostly again using results based math) for ground ball, flyball, popup and line drive percentage. (But doubles and triples, despite the fact that some pitchers fairly consistently have higher or lower amounts, even than pitchers with similar K and BB and GB/FB rates, are still allegedly completely luck of course!)

        Fangraphs continues to stick with FIP, (though they also include some newer ones on their site, though not in their WAR, for later years) and for me that means I can ignore their WAR for pitchers.

        Someday the root though that pitchers have no control over the type of contact they give up will be looked at kind of the way leeching is looked at by doctors now.

        Reply
        1. John Gale

          I see what you mean by that. Of course, if Fangraphs tried to convince us that Babe Ruth’s home runs or Ted Williams’ walks or Adam Dunn’s strikeouts were a function of luck, no one would take that seriously. So it’s interesting to see that they put so much faith in the flip side of that premise.

          Reply
  29. Pak

    I also would want to be certain that I am counting on just one man to pitch until the game is complete, so I would choose Gibson, Koufax or Walter Johnson before Pedro, probably in that order. Pedro was not expected to finish games at the same rate that his predecessors were, I understand that, but I think this is the flaw in WAR for pitchers. How much did these pitchers lose off their WAR as they tired, but were still their manager’s best option?

    Reply
  30. Bothrops Asper

    And how much WAR did they gain by going 100 IP more than modern players just because of usage patterns of the time? Don Sutton, who never finished higher than 5th in IP had more replacement points per season in his prime than Maddux, who led his league 5x or so. That really isn’t fair either. Makes one want to only use WAA for pitchers.

    Reply
  31. Todd

    I’m just a regular reader that never submits entry, but I wanted to add to this column comments. I have been fortunate to attend to several historic Red Sox games in my time (including both the Grady Little “I’ll leave Pedro in” fiasco finalized by Aaron Boone and the following year’s “Break the Curse” game in NYC to go to the 2004 World Series), but the one game I have attended in person that exemplifies “performance” – about 8 rows from the field – was 6/4/1999. Pedro Martinez struck out 16 against the Atlanta Braves, going up against Tom Glavine. Ryan Klesko hit one of the nine homers that year….squeaking around Pesky’s Pole in right field (which one could argue for fun and raise debate if that home run should be included with the other 8 given up that year). Crowds chanted Pedro’s name religiously whenever there was a 2 strike count as if it was a playoff game. He was on top of his game and it was obvious then as it is today. That memory epitomizes baseball for me. Thanks Joe for such a great article (one again).

    Reply
  32. Carl

    Hi Joe,

    I was 9 years old and lived in NYC, so to me Ron Guidry’s 25-3 w a 1.74 (w the DH) was the gold standard of dominant pitching seasons. My father, who grew up a Dodger fan would always claim Koufax ’65, period (much like some of the posters you referenced). I also remember many articles in Baseball Digest touting Gibson’s 1968. Still in NY in 1985, I could understand those who argued for Gooden in 1985.

    This article sent me to BR to do some research and some thinking. The top post 1900 BR WAR for pitchers showed the following list:
    Rank Pitcher WAR Year
    1 Walter Johnson+ (25) 14.6 1913
    2 Walter Johnson+ (24) 13.5 1912
    3 Cy Young+ (34) 12.6 1901
    4 Dwight Gooden (20) 12.2 1985
    5 Pete Alexander+ (33) 12.1 1920
    6 Steve Carlton+ (27) 12.0 1972
    7 Roger Clemens (34) 11.9 1997
    8 Walter Johnson+ (26) 11.9 1914
    9 Pedro Martinez (28) 11.7 2000
    10 Wilbur Wood (29) 11.7 1971
    11 Joe McGinnity+ (32) 11.6 1903
    12 Ed Walsh+ (31) 11.4 1912
    13 Eddie Cicotte (33) 11.4 1917
    14 Red Faber+ (32) 11.3 1921
    15 Bob Gibson+ (32) 11.2 1968
    16 Hal Newhouser+ (24) 11.2 1945
    17 Walter Johnson+ (22) 11.2 1910
    18 Walter Johnson+ (27) 11.2 1915

    Man, that Water Johnson could pitch. By this simple metric, Walter Johnson would get the nod as to the most dominant and valuable pitching seasons. For those who prefer post-integration, Dwight Gooden’s 1985 leads the list. Gibson’s 1968 is up there, as is Carlton’s 1972 season. Both Guidry’s and Koufax’s best season don’t make the cut.

    However, WAR is a counting stat. Like any counting stat, whether RBIs, HRs or WAR, the number of opportunities needs to be considered. For example, Pedro’s 2000 with 11.7 WAR was in 217 innings while Joe McGinnity’s 11.6 WAR in 1903 came in 434 innings. Same value over exactly twice as many opportunities. To adjust, I divided each season’s WAR by the innings pitched, and then for presentation, multiplied by 100. The most dominant pitching seasons list then becomes:
    Rank Pitcher WAR Year IP WAR/IP * 100
    1 Pedro Martinez (28) 11.7 2000 217 5.392
    2 Roger Clemens (34) 11.9 1997 264 4.508
    3 Dwight Gooden (20) 12.2 1985 276.2 4.417
    4 Walter Johnson+ (25) 14.6 1913 346 4.220
    5 Bob Gibson+ (32) 11.2 1968 304.2 3.682
    6 Walter Johnson+ (24) 13.5 1912 369 3.659
    7 Hal Newhouser+ (24) 11.2 1945 313.1 3.577
    8 Wilbur Wood (29) 11.7 1971 334 3.503
    9 Steve Carlton+ (27) 12.0 1972 346.1 3.467
    10 Red Faber+ (32) 11.3 1921 330.2 3.422
    11 Cy Young+ (34) 12.6 1901 371.1 3.395
    12 Pete Alexander+ (33) 12.1 1920 363.1 3.332
    13 Walter Johnson+ (27) 11.2 1915 336.2 3.331
    14 Eddie Cicotte (33) 11.4 1917 346.2 3.293
    15 Walter Johnson+ (26) 11.9 1914 371.2 3.206
    16 Walter Johnson+ (22) 11.2 1910 370 3.027
    17 Ed Walsh+ (31) 11.4 1912 393 2.901
    18 Joe McGinnity+ (32) 11.6 1903 434 2.673

    As shown, Pedro’s 2000 season was, by BR-WAR, the most dominant pitching season of all time and by a fair margin.
    Seeing Pedro’s name atop the list made me, as a NYer, feel sort of like the first time I walked into the MoMA and realize all the great art I’ve overlooked because I had been raised looking at the Old Master at the Met. Thank you for writing this article and opening my eye’s to Pedro’s 2000 pitching season.

    Reply
  33. mwarneridx

    The biggest surprise about this post for me was that a discussion of Pedro’s brilliance, and superior pitching as a whole, Omar Olivares shows up…

    Reply
  34. JB

    Without getting into all the stats, one thing really stuck out. Five complete games. So that is the guy you want to win one game? A guy who MIGHT go 8? Okay.

    Reply
  35. bl

    For the fun of it. Here’s a link to a blog post with vidoes of Pedro’s greatest moments with the Red Sox. It has the entire Game 5 of the 1999 ALDS against Cleveland. It was said above that his relief appearance was on one day’s rest. That isn’t accurate. He actually hadn’t pitched since Game 1 which he left early with a back injury. He was still injured and couldn’t get to 90 mph with his fastball and still he managed to shutdown the offensive juggernaut that was the Cleveland Indians. Absolutely amazing performance. I’m pretty sure it’s this one game that caused Joe to pick 1999 Pedro to pitch for his soul.

    http://www.overthemonster.com/2014/1/31/5350324/pedro-martinez-greatest-red-sox-games

    Reply
  36. Karyn

    Everyone whinging about complete games and number of starts and innings pitched: Come on. If Koufax and Gibson and Carlton pitched in the same conditions that Pedro did, they’d very likely have similar results.

    Reply
    1. Stephen

      Well, I can’t speak for everyone, but if you look at my posts above you’ll see that I’m not saying that Martinez was worse than those other guys, or that time periods don’t affect how pitchers are used.

      I don’t know who had the best single season. Maybe it was Martinez in ’99. Maybe Martinez in 2000. Maybe somebody else. As Joe explains so well, the metrics we use don’t always agree. There is room for argument.

      But we have to use the information that we have. And the question that kicked this off is, Who would you want to pitch a single game with your soul on the line. And I suppose that’s maybe nothing more than a nice literary device for framing the question that Joe is asking, not to be taken seriously, but he DID ask it, and some of us are answering it; and from that perspective the fact that Pedro often left the game with two innings to go does matter. It’s a part of his record; maybe he would have been able to complete more games if he’d been asked to, but the reality is that he didn’t.

      [And while we're playing what-if: What if Koufax had played in a time when medical knowledge and training methods were forty years more advanced? What might he have achieved if his elbow hadn't been killing him on every freaking pitch?]

      Reply
  37. Mr Furious

    Absolutely. Both years are Pedro at his peak, but those two performances in particular are among the most memorable in his career, and not to be overlooked in the same incredible 1999 season is Pedro coming out of the bullpen in the decisive ALDS Game 5 tied 8-8 in the 4th and pitching six shutout innings, striking out 8 Indians to carry the Sox on to the ALCS.

    1999 has the clear “domination intangibles” edge over 2000.

    Reply
  38. Pingback: Ultimate Sports Talk – Red Reposter: Real Fans go to Game 2!

  39. KB

    My answer: It depends on your bullpen. Pedro was not a 9 inning guy. In a game for my soul, I’m kinda hoping for a guy I can count on for nine innings, unless I have Goose Gossage and Mariano Rivera in my bullpen. If I don’t have either of them, then I am thinking it’s a tossup between the 1912 and 1913 Walter Johnson, leaning toward the 1912 version. 2 HR allowed in 369 innings, 303 Ks, 34 CG out of 37 starts.

    Reply

Comment: