By In Stuff

Talking Pitching and Defense

OK, let’s begin by going over some well-known baseball history. When the game was first played professionally, not long after the Civil War*, pitching was simply a trigger to get the play started. 

*Or as my daughter’s seventh-grade history teacher calls it: “The War of Northern Aggression.”

Pitchers stood just 50 feet from home plate and were required to pitch the ball underhand like horseshoes (which is why they are still called pitchers). In those early days, the rules were often adjusted to keep pitchers from doing anything cunning or illusory to trick hitters. It took NINE called balls for a walk then — pitchers were just supposed to get the action started.

In those days, you would probably say that run prevention was split this way:

Pitcher 5-10%
Defense: 90-95%.

Over time, of course, pitchers took a more aggressive role, and what baseball people found was that the one-on-one battle between pitcher and hitter was quite exciting. Rules changed rapidly. Pitchers were given the freedom to throw the ball from above the waist, and balls necessary for a walk were reduced almost annually. The pitcher’s mound was moved back. Catchers were allowed to wear padded gloves. Numerous other rules were changed and added in to deal with this major shift in the game. 

Ever since then, there has been an effort to determine just how much of run prevention should be credited to the the pitcher, and how much of credited to the eight other defenders on the field.

This will eventually lead us to Steve Kluber** and Felix Hernandez, but before that we have to begin with the earliest and so far most popular effort to separate pitcher and defense: Earned Run Average. 

The idea of “earned runs” and “unearned runs” goes back to just about the beginning of baseball record keeping. As Alan Schwartz explains in “The Numbers Game,” earned runs began as an OFFENSIVE statistic — Henry Chadwick, the inventor of the boxscore, wanted hitters to only be credited for the hits they made and not for the blunders of fielders. Pitchers were still throwing underhand then and so were not much of a consideration.

For most of the early years, the only pitching statistics that mattered were wins and losses — a theme that has kept some momentum for more than a century — but about 100 years ago, earned run average was introduced to much acclaim. Here was a statistic that, using errors and a bit of logic and ingenuity, would tell you exactly what runs had been the pitcher’s doing and what runs had scored simply because of defensive incompetence.

You know the simple formula:

ERA = (Earned runs x 9) / innings pitched.

Of course, the formula is not THAT simple because “earned runs” is a complicated deal. We don’t think much about it now because earned runs have been around for so long, but the earned runs concept takes up a full four pages in the baseball’s officials rules and includes a full paragraph on how to properly score earned runs after catcher’s interference. There’s also this gem:

“When a fielding error occurs, the pitcher shall be given the benefit of the doubt in determining to which bases any runners would have advanced has the fielding of the defensive team been errorless.”

All of this earned run stuff is based on some guy’s decision up in a press box whether or not a fielder should have caught the ball. Then everybody has to imagine what WOULD have happened if the error had not occurred. There really isn’t a concept like it anywhere else in sports.

Well, as you know, ERA for many people settled the question of what responsibility a pitcher bears for preventing runs. If a fielding made a bad play, hey, that wasn’t the pitcher’s fault. You can’t blame him because the ball went through the third baseman’s legs or a centerfielder threw the ball over the catcher’s head. ERA became the run prevention statistic of choice, and remains so to this day.

There are, of course, some problems with ERA. For one thing, it manages to clear pitchers of responsibility when fielders mess up but it doesn’t counterbalance that by giving the fielders credit when they save the pitcher with great plays. If an outfielder leaps at the wall and saves a home run, the pitcher doesn’t have an imaginary run added to his ERA, even though by ERA logic he should. He gave up that run. The fielder took it away.

But even more to the point, ERA itself is make-believe. It pretends that those unearned runs scored somehow didn’t count. But they did count. The runs did actually score. We’re just not recording them on the pitcher’s record. We’re chalking them up to some vague notion of lousy defense.

I was thinking about this when looking hard at two pitchers with very similar records: Curt Schilling and Kevin Brown.

By some statistics, Schilling’s career and Brown’s career are freakishly similar. They pitched almost the exact same number of innings:

Schilling: 3,261 innings
Brown: 3,256 innings

They had very similar won-loss records.

Schilling: 216-146
Brown: 211-144

They have ERAs that are close, and their ERA+ —which compares them against league average and considers park factors — is identical:

Schilling: 3.46 ERA, 127 ERA+
Brown: 3.28 ERA, 127 ERA+.

And yet most people — including me — think Schilling was a noticeably better pitcher (For instance: I annually vote Schilling for the Hall of Fame, and did not vote Kevin Brown). Why? Well, there are some clear reasons. Brown was mediocre in the postseason, Schilling was legendary. Schilling had more good seasons. Schilling has the greatest strikeout-to-walk ration in baseball history.*

*I’m not counting Tommy Bond, who pitched in the 1870s and 1880s.

But there is something else that is right there, in front of our faces, but has become hidden in the record because of ERA. Kevin Brown gave up 107 more unearned runs than Schilling. That’s a big difference. Brown’s ERA was .18 better than Schilling, but his RA — Runs Against — is not as good.

Brown: 3.75 runs per game.
Schilling: 3.63 runs per game.

Now, you consider the ballparks they played in, the leagues, and so on — look, Kevin Brown was a superb pitcher. But, in reality, he wasn’t as good as Schilling at preventing runs. His ERA is, at least in part, an illusion.

This is when some people will say: No! Not an illusion! You can’t blame Kevin Brown for those unearned runs.

I want to talk about this word “blame” in a minute.

Point is: This is the baseball argument that goes on and on.

In the last few years, there has been a new kind of effort to separate pitching and fielding and it goes by the not-quite-natural-sounding acronym: FIP. Fielding Independent Pitching. FIP is based on Voros McCracken’s theory that pitchers have much more control over strikeouts, walks and home runs allowed than they do over balls hit in play. The statistic FIP (and its more belligerent cousin xFIP) is significantly more complicated than ERA:

FIP == ((13 x HR) + (3 x (Walks but Hit By Pitch)) – (2 x strikeouts)) / innings pitched plus constant.

The constant at the end there is just meant to make FIP look like ERA so that it’s easier for people to digest. The main point is, the only three things that make up a pitcher’s FIP are homers, walks (and hit by pitch) and strikeouts.

FIP, like ERA, has believers and skeptics at all levels. Some think pitchers have almost no control over balls hit in play, that the fate of those balls in play is entirely based on defense and luck. Some think pitchers have some but limited control over balls in play. Some think pitchers have the majority of control and that the whole FIP thing is bunk.

Which leads us, finally, to Kluber and Hernandez. A few people have written in to say that they are disappointed in my Kluber choice for Cy Young, and they will throw a flurry of statistics that prove their point (mostly ERA and WHIP) and that it’s not fair to “blame” Hernandez because he happened to have a better defense and home pitching ballpark.

But I think “blame” is the wrong word. I don’t blame Hernandez for any of it. I’m just one of the voters trying to determine who was the better pitcher.

If you have two sprinters, and one runs a 10.0 100-meter dash and the other runs a 10.2, then you know that the first performance is better.

But what if the first one ran with a tailwind, and the the second ran into the wind? What if the first ran on an Olympic track and the second ran on gravel? What if the first ran wearing running shoes and the second ran wearing boots?

We’re just trying to determine which performance is better. That’s all. Nobody blames the first runner for running with the wind behind him, nobody says that was in his control. But it’s reality. He had the faster time. But he might not have had the better performance.

Hernandez’s advantages over Kluber come down to two things: He gave up fewer hits and gave up fewer runs. He did these things even though Kluber struck out more batters and gave up fewer home runs. How? Essentially, it comes down to this:

Batting average on balls in play:

Felix Hernandez: .260
Corey Kluber: .318

So, now you ask — how much of that major difference was Hernandez’s pitching and how much of it was Hernandez’s circumstances? We are back to that question: How much of this was due to the pitcher and how much was due to the conditions and fielders?

Those who think balls in play are almost exclusively the domain of the pitcher’s work will say that Felix Hernandez had the better season.

Those who think balls in play are almost exclusively the domain of defense and luck will say that Kluber had the better season.

And those of us somewhere in the middle will go back and forth. I sure did. They both deserved the award. I picked Kluber by a nose.

One final thought to wrap this up: We again had an ERA illusion in this race. Hernandez’s ERA was 2.14, while Kluber’s was 2.44. That’s a big difference.

But again, the RA is closer. Hernandez gave up 12 unearned runs this year; Kluber gave up eight.

Hernandez: 2.59
Kluber: 2.75

When just about everyone — including those pushing for Hernandez — seem in agreement that King Felix had markedly better defense behind him, it doesn’t seem right that Felix gets bonus ERA points.*

**The Steve Kluber reference is an inside joke … I was going to leave it without mention, but it occurs to me that some might not get it. In my Cy Young column, I called Corey Kluber Steve twice for reasons that even now I cannot quite get clear in my mind. And so every time I write about Corey Kluber, I intend to call him Steve at least once.

52 Responses to Talking Pitching and Defense

  1. Brian says:

    Nicely balanced. Hard to argue your final decision.
    Just to comment – watching my son in Little League, it does seem good pitchers seem to induce better fielding. By that I mean, watching my son, he seems to induce more ‘bloopers’ than other pitchers, making for easy outs – something I never appreciated before. He would hold a team scoreless, then his reliever would get rocked. Before witnessing that, I assumed K’s were all that mattered.

    anyways, love your columns

    • DjangoZ says:

      I think the key word is “seems”. It makes sense to us and we seem to see it with pitchers…and yet if you analyze pitchers there is little evidence of this effect.

      • Brian says:

        I started to add to my note about how in MLB the effect may be less pronounced – but changed my mind. While I used the word ‘seems’, for my son pitching in Little League, it was definitely true. When he was ‘on’ he induced bloopers, and other pitchers did not. I believe that it was because his fast ball had a natural movement that made it hard to hit square. In his final year of Little League (12 yrs) he did not give up a single home run. He was the number one pitcher on his team, and had actual results (wins and saves) to show.

        • fhomess says:

          The whole DIPS theory (defense independent pitching statistics) has never been proven to apply to 12 year olds. There is little doubt in my mind that major league hitters could hit .500 off me on balls in play (no, I’m not 12). There is a difference in terms of pitcher skill in their ability to limit hits on balls in play, but by the time pitchers make the major leagues, that difference becomes much harder to convincingly identify. The talent is much closer together.

          Your son probably IS quite a bit better at inducing weak contact than other pitchers in his league.

    • Spencer says:


      I can’t tell but are you drawing a comparison between little league and professional baseball? It’s safe to say Voros McCracken’s theories don’t apply to 12 year olds.

      • Brian says:

        Nah. Just pointing out when I first started to appreciate when a pitcher can affect outcomes even if he does not strike out the batter.
        It is clearly more pronounced in Little League.

        I think Joe hit the right balance.

  2. DJ MC says:

    Or as my daughter’s seventh-grade history teacher calls it: “The War of Northern Aggression.”

    To be fair, no matter where one’s sympathies lie (on a 150-year-old conflict), this is a much more fun way of discussing the war. It manages to convey both the spirit of the Southern side as well as their complete delusion in just four words.

    • Goirish says:

      Also fun – it calls to mind the best X-Files episode ever (featuring the inbred Peacock family), where the mother/matriarch referred to the conflict as the War of Northern Aggression. First time I had heard that expression and it has stayed with me since then.

      • RickyB says:

        If by ‘best’ you mean ‘grossest,’ then you are correct. As a huge X-Files fan, Home is in the top 20, but there are other episodes that are superior: War of the Coprophages, Bad Blood, Jose Chung’s From Outer Space, Clyde Bruckman’s Final Repose and Humbug among the stand-alones, and there are a number of mythology episodes I would rank higher.

        Also, according to my wife (who hails from Louisiana), the South consists of Louisiana, Arkansas, Mississippi, Alabama and maybe South Carolina and Georgia (but not Atlanta). Everything else is the North to her.

    • I teach history in New York and I tell my students that many Southerners — USED TO call it the “War of Northern Aggression.” I honestly would not expect any present-day teacher to use that phrase.

      Then again, maybe the South will rise again?

      • stevemillburg says:

        I live in Alabama. I hear “War of Northern Aggression” a lot. Sometimes it’s at least mostly a joke. Sometimes it’s not AT ALL a joke.

  3. Edward says:

    Corey Kluber, not Steve Kluber.

  4. Faye Schlift says:

    And then there’s the pitcher making an error and any subsequent runs off that error don’t count as earned. Where’s the logic?

    • Jake Bucsko says:

      YES. Late in the season, Felix Hernandez made a fielding error that led to 3 “unearned” runs. In what way are those runs unearned? The whole thing is illogical. Imagine the uproar in fantasy football if Colin Kaepernick throws a TD pass through Richard Sherman’s hands and into Michael Crabtree’s.

      Oh, well, Sherman “should have” intercepted the pass. Yeah, well, he didn’t. I’m not interested in what should have happened.

      • RickyB says:

        Because the defense is in control of the ball in baseball, unlike other sports, comparing what should have happened to other sports is illogical. I agree ERA has its flaws, especially considering it does not take into consideration that there are different types of unearned runs. For instance, man on third, two outs, ball goes through the shortstop’s legs. That run was a direct result of the the error. Then there are the unearned runs that score well after the error, like when the leadoff batter reaches on an error, but later scores when a pitcher allows a home run with two outs. The latter scenario accounts for two unearned runs, when the pitcher knew the situation he was in yet failed in his job. Much like not having a reliever get away with an unearned run if he came in the game with a man on base via an error and two outs. He gives up a home run, the run is unearned for the team, but earned for the pitcher. Because he knew the situation and does not get the benefit of the error.

        Again ERA is flawed, but I can’t help but love its nuances.

    • RickyB says:

      The logic comes from determining what is caused specifically by the pitcher’s pitching, not by any defense, including his own. ERA isn’t supposed to calculate what the pitcher is responsible for, rather what his pitching is responsible for.

    • Karyn says:

      The logic is that ERA tries to measure how good a pitcher the guy is, not how good a player he is overall.

      How successful ERA is at this goal is open to debate, but that’s the logic for pitcher errors.

  5. MikeN says:

    There is a war between the sabr folks and traditional guys, to the point where they trash Clint Eastwood’s movie. They should have realized that Clint was the advanced stat guy, attacking batting average while he looked at(heard) the advanced stats.

    • Spencer says:


      Are you joking? That movie was an attack on sabermetrics, and it was absolutely terrible to boot. Of course they trash it, it deserves to be trashed.

  6. Ok what gets me is the first paragraph and the note that your daughters teacher calls it “The War of Northern Aggression” I went to school in North Carolina (in Person County) too in 70’s and that is what my teachers called it as well. It’s hard to believe that this is still going on but I am not surprised.

  7. BobDD says:

    No matter how you evaluate players, these two are hard to rank. I think while trying to convince King Felix fans the other guy was superior that calling him ‘Steve’ in that context is genius. Allows just enough escape time.

  8. How much variance in BABIP is there among the pitchers of one team?

    If one pitcher on a team has a relatively high BABIP, whereas another pitcher on the same team has a relatively low BABIP, then the variance between them can be attributed to the pitcher, because the other fielders tend to be the same players.

    If Hernandez would have a low BABIP, but Iwakuma, Young and Elias would have a high BABIP, then you could argue that Hernandez might be a pitcher who elicits a lot of easy grounders, etc.

    • According to Baseball Reference, the BABIP values of the pitchers who started at least 5 games for the Mariners:

      Hernandez: .260
      Iwakuma: .287
      Young: .240
      Elias: .297
      Ramirez: .308
      Paxton: .274
      Maurer: .311
      Walker: .287

      Team total: .277. League average: .299.

      While it seems that Seattle has a good defensive, it also seems that Young and Hernandez are good at eliciting easy grounders, etc., whereas Ramirez and Maurer had more difficulties doing so.

      • Bobby Mueller says:

        The thing is, Felix had a much better BABIP allowed this year than the team average, but that doesn’t happen every year for Felix, so I think it’s hard to say that he has an ability to elicit easy grounders/pop-ups/etc. Using data from FanGraphs, Felix has had a better BABIP than the team BABIP just 3 times in 10 seasons:

        2005: .253, .287–Felix was better
        2006: .312, .297
        2007: .333, .315
        2008: .314, .307
        2009: .278, .272
        2010: .263, .279–Felix was better
        2011: .307, .285
        2012: .308, .282
        2013: .313, .303
        2014: .258, .275–Felix was better

        For all 10 seasons, Felix’ BABIP is .295. The Mariners’ BABIP over those same 10 years is .290, so over 2060 innings, Felix has allowed a higher BABIP than his team’s BABIP.

    • sbmcmanus says:

      This is a good thought in theory, but the analysis depends on the volume of data for each pitcher, because there will be some statistical noise masking the “real” BABIP by pitcher unless you have a very large number of balls in play to go on. There almost certainly isn’t enough data by pitcher in a year to make conclusions – we know from experience that BABIP bounces around for the same pitcher from season to season, so the single season numbers for different pitchers can’t be compared to each other with much certainty.

      • That is absolutely true. Especially, for pitchers, such as Maurer and Walker, there are simply not enough observations to assume that the results are reliable (i.e., stable). Furthermore, it does not take into account the role of the pitcher as a fielder or the role of the stadiums in which the games are played.

        • I had a closer look at Hernandez’ BABIP numbers. While last year’s BABIP value was lower than the team total, his values are more often than not higher than the team total, suggesting that Hernandez might not be as good in eliciting ground balls after all. The years in which his values was lower than his team total are highlighted with an asterisk.

          2014: .260 / .277*
          2013: .316 / .305
          2012: .309 / .284
          2011: .308 / .287
          2010: .265 / .281*
          2009: .280 / .274
          2008: .316 / .309
          2007: .337 / .319
          2006: .313 / .299

          • Dave says:

            FYI on BABIP, for the years up until ’13–before he was hurt–Matt Cain consistently, year after year, had a low BABIP, lower than the rest of the Giants, lower than the league. Baseball Prospectus writers several times I believe called it a “skill” he had developed.

  9. Fin Alyn says:

    How many more errors were made behind Brown than Schilling? Merely citing how many more unearned runs he gave up is meaningless by itself. Did his teams make 100 more errors behind him? What was the amount the other pitchers on the same staff gave up? Brown played on some terrible defensive teams in Texas, so not only was he playing with guys who made errors, but guys who couldn’t get to the ball either.

  10. otistaylor89 says:

    I assume that the fact that Brown gave up more runs was due to the fact that once an error causing an unearned run to happen, he couldn’t get out of the inning, snowballing into more unearned runs? Or did his teams make that many more errors, causing more unearned runs?
    For Brown, it seems to be a problem during his time in Texas and his 3rd baseman was Dean Palmer. For Schilling, during his time in Philly, he he gave up very few runs above unearned runs and his 3rd baseman was Scott Rolen.

    • Anon says:

      Right. Because the difference between Schilling and Brown comes down to the 3.5 years in their 20 year careers that Rolen was Schilling’s 3B and Palmer was Brown’s 3B. No offense but this might be the stupidest thing I’ve ever seen posted on this board.

      • otistaylor89 says:

        You obviously didn’t look at those years for both players (earned runs, runs) very closely.
        Why is it a stretch to say that one of the best, if not the best 3rd baseman of the era helps playing behind you and guy who lead the league in errors multiple times at 3rd hurts playing behind you?
        3rd basemen commit a ton of errors and a lot of those errors can be big disaster type errors. Did you know that Rolen committed 186 errors in 17,479 innings at 3rd base in his career, which is 94th all time? Dean Palmer committed 185 errors (95th all time) in only 9,852 innings(?!?!).
        BTW, Scott Rolen never played another position besides 3rd during his time in MLB – and that includes DH.

        • Anon says:

          Oh, I looked at the years trust me. Again, it’s only about 3.5 years for each player over 20 year careers for both Schilling and Brown. That alone makes it an idiotic statement, but let’s look at the numbers .

          Brown/Palmer – Palmer only played half seasons in 91 and 94 (strike year but still, only 93 games). He wasn’t a good fielder with 6, 22, 29 and 22 errors in 91-94.

          Schilling/Rolen – Rolen only played 37 games in 1996 and Schilling missed some time in 1999 and was traded halfway through 2000. Call it 3.5 seasons. Rolen was generally good but his errors go 24, 14, 14, 10 from 97-2000. SO he had 1 season that really was no different from Palmer’s in terms of errors which are the important metric for considering unearned runs. In total the difference is maybe 10-15 errors over 3+ years.

          If you really think that alone explains the unearned runs difference between the two, that is absolutely the dumbest thing ever posted here.

          • otistaylor89 says:

            You failed to mention that Palmer, in that strike year playing only 91 games, made 22 errors and that Brown gave up 18 unearned runs . 18 unearned runs was the 2nd highest in his career next in 1992’s 19 – also with Dean Palmer as his 3rd baseman.
            I never said that Palmer/ Rolen was the sole reason for the difference between Brown and Schillings unearned run totals (Brown being more of a ground ball pitcher and Schilling a flyball pitcher, especially later in his career, being the #1 reason), but having Dean Palmer as your 3rd basemen for any amount of time isn’t going to help you much.

  11. Ed says:

    I can’t believe a teacher in Charlotte calls it the War of Northern Aggression.

    I grew up in Greensboro, which is probably not as liberal as Charlotte… and I never once heard anyone call it the War of Northern Aggression.

  12. Cliff Blau says:

    Run prevention is 100% defense. That’s the definition of defense (“the team in the field”, per the rule book.) It consists of pitching and fielding, not pitching and defense.

    Also, Henry Chadwick was not the inventor of the box score, which came from cricket and was first used for a baseball game 11 years before Chadwick started covering the game.

  13. Jim Townsend says:

    It’s an easy decision for me. Schilling is a jerk so Brown was better.

  14. John says:

    As a die-hard Maddux fan asking another die-hard Maddux- how would you reconcile his success if balls hit into play are merely up to luck and defense? I know what you will say, and I agree. Just interesting to hear you make such an argument after writing a long column some time ago about how Maddux made the Yankees “just miss”.

    • Anon says:

      The concept of pitchers having “no control” has never really been that pitchers have ZERO control over batted balls. Rather the observation is that pitchers have much less control than previously thought. Certain pitchers do have some control over batted ball contact -notably knuckleball pitchers.

  15. Other Brian says:

    Knuckleball pitchers also produce more passed balls, which lead to more unearned runs. I read an article many years ago (can’t find it now) arguing that we should use run average instead of earned run average when evaluating careers in that there was enough data to overcome fluctuations in a pitcher’s team’s defense over the course of his career, and it presented that the majority of pitchers with the greatest difference between RA and ERA were knucklers, given PB’s.

  16. Erik L. says:

    Here’s a decided disadvantage that King Felix has that nobody really talks about. Most of his Cy Young competition gets to pitch against the Seattle Mariners. (See Kluber, 2014: 9 IP, 3 hits, 0 runs.)

  17. J Hench says:

    The other issue with ERA is allocating runs between two pitchers in the same inning. 100 years ago, relievers were called upon with much less frequency than they are now. With increased reliance on bullpens comes an increase in the number of runs charged to a pitcher’s record that scored when the pitcher was no longer in the game. This is just as much an artificial construct as the distinction between “earned” and “unearned” runs.

    To be sure, when Jake Peavey loads the bases before being replaced by Yusmeiro Petit, then when those runs score, it seems more fair to hold Peavey accountable than Petit. But perhaps just as often you the pitcher who gets through 7 1/3, allows a scratch single, the manager pulls him for a reliever, and the reliever allows a 2-run shot, with the first run charged to the starter – from ERA’s point of view, the starter might as well have allowed a solo HR instead of a single.

  18. Mark A says:

    It seems to me that since baseball is already trying to account for the impact of defenders with defensive runs saved type statistics, that perhaps there should be a pitching stat that tries to normalize (rather than ignore) pitching performance for defence.

    ERA doesn’t just have a problem with not counting great defense. If the third out would have been made, then the pitcher walks 2 and gives up a home run (all totally within his control), those runs are all considered unearned. And that is pretty silly.

    Also, counting runs allowed gets into the same problems as evaluating hitters by RBI. You can have a bad pitching outcome without giving up a run. You can sometimes even have a good one and still give up a run.

    So it seems to me that you need to evaluate pitching much more similarly to the advanced measures in hitting. Count by the number of batters they faced. Evaluate the outcomes of those plate appearances. Then take off defensive runs saved. I assume that metric accounts for errors with the amount of extra runs one would normally expect such an error to give up.

    Still some grey areas (how to factor in baserunning for example, and whether to count intentional walks)

Leave a Reply

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