Papi’s Popup and Yu’s No No

Let’s stop for a moment. Let’s stop for a moment a think about how stupid this thing we are arguing about really is. Friday night in Texas, Yu Darvish had a no-hitter going. In the seventh inning, David Ortiz popped up the ball to short right field. Routine as it gets. But the Rangers had on the shift, so fielders were in somewhat unfamiliar places. Rangers’ rookie second baseman Rougned Odor was sort of in the vicinity of where the ball was going to land. This seemed to confuse him and it also seemed to confuse right fielder Alex Rios who should have stepped in to catch it. Instead, the ball dropped between them.

Here is my best guess.

1. Ninety-nine percent of baseball fans believe that ball absolutely should have been caught.

And …

2. Ninety-nine percent of baseball fans believe it should not be ruled an error because neither guy touched the ball.

This is the stupidity of errors in 21st Century America.

You will sometimes hear baseball people mock the concept of Defense Independent Pitching Statistics or DIPS. The idea behind DIPS is a fairly new one — last decade or so — and it is meant to separate the contribution of the pitcher from the contribution of fielders. DIPS does this based on the theory that there are only three things that a pitcher has demonstrable control over (strikeouts, walks and home runs) and everything else is some foggy mix of luck and defense and whatever ambiguous ability a pitcher has of controlling how well a ball is hit.

Many of the critics I’ve heard of DIPS do not rip specific details but the whole idea. How can you say pitchers don’t have control on balls hit in play? Madness! Baseball has a rich history of giving way too much credit to pitchers. Heck, people used to say pitching is 90% of baseball. NINETY PERCENT. No other player can get credited with a victory. More to the point, no other player can have his stats boosted by a benevolent scorekeeper who sits up in a press box and says, “Oh, hey, don’t worry about it, that run wasn’t your fault.” It troubles many people that DIPS does not give pitchers credit for preventing hits on balls in play. It doesn’t seem to trouble as many that baseball has long given pitchers credit for amazing plays that fielders made behind them.

But then … the counting of errors and the calculation of ERA are just a prehistoric form of DIPS. Very early in the game’s history, when pitchers would actually pitch the ball like horseshoes and were only responsible for starting the action (kind of like slow-pitch softball pitchers today), defense was everything. To determine the best fielders, newspapers began to put “Errors of Fielding” into their early box scores. According to Alan Schwartz’s fascinating “The Numbers Game,” the father of baseball statistics Henry Chadwick — who basically framed the way baseball games would be quantified for more than a century — did not like the error concept and wanted instead to judge fielders by the number of successful plays they made. That was one of the few statistical battles Chadwick lost. Errors became the dominant way to judge fielding and, in a less visible way, judge pitchers.

As pitching developed into the most important part of run prevention, the error stayed in the game — the general motivation being the same as DIPS. They wanted to separate defense from pitching. Only these statisticians came at it from a different angle. They came at it assuming that pitchers have COMPLETE control of balls hit in play. They deserve 100% of the blame when the player gets a hit. But if they compel a batter to hit a ball right at a fielder and the fielder doesn’t do his job (turn it into an out) then, well, that’s the fielder’s fault and not the pitcher’s fault. The fielder would get an error. And the pitcher, through the dominant ERA statistic, would get the assumption the out was made. It’s like pitchers — alone among all athletes in sports — have been allowed to live in this alternate universe.

And this is how baseball has been scored ever since, to very little disagreement, even though it is a logical nightmare. Why were pitchers CREDITED when fielders made dazzling plays that should have been hits (even home runs) but NOT DEBITED when fielders missed plays that should have been outs? Why were people in press boxes making determinations about what should have happened? (This kind of scorekeeping does not happen in any other sport). Why were official scorers going through the craziest hoops to figure out what the pitcher DESERVES (“OK, so let’s see here, if that error hadn’t happened, there would have only been a runner on first, and he probably would not have scored on that double, so that’s not a run, and then the second error would have been the third out of the inning so all the runs that scored after that are unearned and …).

As Bill James wrote long ago, an error is a “moral judgment, really, in the peculiar quasi-morality of the locker room.”

I really think the crazy, illogical error concept has lasted all these years because we as baseball fans are desperate to credit pitchers rather than crediting entire teams. We like that pitcher-hitter matchup; like bloodless boxing. We want to credit pitchers for victories, for no-hitters, for perfect games even though they don’t do these things alone. We have spent more than a century thinking of defenders as Pips to the pitchers’ Gladys Knight. We have spent more than a century thinking of fielders as automatons who should ALWAYS make plays that look routine. If they happen to make a dazzling play now and again that keeps runs from scoring, OK, that’s nice. We’ll give you a gold glove at the end of the year, like the gold watch after working for 25 years. Nice work. Now, go support your pitcher.

The Darvish-Papi play shows you just how ridiculous this has become. Defenders as a group have never made fewer errors. Last year, teams made 2,747 errors in almost 5,000 games; that error-per-game percentage (56%) is the lowest in baseball history. Compare that with the 10,000 errors made in 3,000 games back in 1890, when the error was being formed as a concept.

Why are errors so far down? I think it comes down to a couple of things. One, fielding has advanced. Gloves are way better, defensive positioning is way better, field conditions are way better and so on. But two, we still give errors based on some antiquated system that barely made sense 100 years ago. Here are grownups arguing FURIOUSLY whether the pop-up that dropped between Odor and Rios should be called an error? Do we realize how stupid that sounds? We know Rios should have caught it. We KNOW Rios should have caught it. We KNOW KNOW KNOW Rios should have caught it.

But should it be called an error? Hmm. We never called it an error before. Hmm.

This is just plain dumb and it really comes down to the basic fact that Yu Darvish was going for a no-hitter. That’s the key — we see it as an individual achievement. It wasn’t the Rangers going for a no-hitter. No. It was Yu Darvish going for a no-hitter. There has never been a pitcher in baseball history who threw a no-hitter by himself, but if you look up the list of no-hitters you find only pitcher’s names.

Our insistence on trying to give too much credit to pitchers has blinded us to how daft all of this has become. A no-hitter should be what it sounds like … it should mean no batter reached base after hitting the ball. It should be credited to a team, with the pitcher playing the starring role. These are obvious things. But we don’t see them, in the same way we don’t see how absurd it is to argue about whether that Rios-Odor drop is officially an “error” or simply a “play that should have been made but wasn’t an error by the silly 19th Century standard we still use.”

We don’t see these things because we have been conditioned not to see them. We grew up with the error and so it makes sense to us, even if it doesn’t make sense at all. I think the error is an outdated concept. I know what we consider an error is an outdated concept. The goofy little ground ball David Ortiz hit to break up Darvish’s no hitter in the ninth was no more deserving of a hit, off the bat, than the routine fly ball that Ortiz hit that was called an error. One bled through. Another plopped untouched. The fact we are still arguing about stuff like this tells you just how powerful even the most ridiculous sports statistics can be.

42 thoughts on “Papi’s Popup and Yu’s No No

  1. mrdardy

    Joe and others – how do you feel about the idea of the ‘team’ error? So, instead of Rios being credited with the error, the error is just assigned to the Rangers. That might help clear up the fog about those dopey non-errors just because no one touched the ball. If it’s pretty clear that someone SHOULD have touched the ball if they were doing their job correctly, then an error is assigned to the team.

    Reply
    1. McKingford

      Except that the scoring rules already dictate that a player need not touch the ball for an error to be attributed. Fools like Harold Reynolds who argue that there should not have been an error scored on the Rios play simply want to ignore the plain text of the rules. Otherwise, nobody but nobody thinks there is any ambiguity in whose fault that play was.

      Reply
      1. Artie

        You know, fools like Harold Reynolds and every official scorer in the Major Leagues. Fools! Fools, the whole lot of them!

        Reply
    2. NevadaMark

      There was a long article in Baseball Digest in the 70′s regarding the very concept you mention.

      Reply
  2. Ty Sellers

    What does it say about me that my major takeaway from the article was Joe’s miscalculation of how many games were played last season (*2430)? And thus, his miscalculation of errors/game (*1.13/gm)? .565 errors/gm per TEAM would be correct but as we all know two teams compete in each game.

    Anyway, the idea that starting pitchers, and starting quarterbacks, are individually credited with wins or losses is one of the most maddening aspects of sports analysis and has bothered me to know end for some time, particularly in football. In baseball, at least theoretically, a pitcher can win a game with only one other teammate, the catcher. He can strike out 27 batters and hit one home run and he has won the game. Will this ever happen? Absolutely not, but at least it is possible. A quarterback has absolutely no bearing on the performance of his team’s defense while they are on the field nor could he possibly move the ball against eleven defenders were it only he and the center (taking from the pitcher/catcher combination above and staying with the modern rules of the game which require a center/quarterback exchange). If I had one sports wish that could be granted this would be it: outlaw attaching a win/loss record to both pitchers and quarterbacks. They are TEAM GAMES!

    That reminds me of the great Caddyshack quote:
    Judge Smails: “Ty, what did you shoot today?”
    Ty: “Oh, judge, I don’t keep score.”
    Judge: “Then how do you measure yourself against other golfers?”
    Ty: “By height.”

    Ok, I’m calm now. As you were saying…

    Reply
    1. Jay

      What does it say about me that my major take away from your comment is you don’t know that the real phrase is “bothered me to no end”.

      Reply
      1. Artie

        I think what he’s saying it this phenomenon has made him so angry he is prepared to kill himself, to ‘know’ the ‘end’ of life. Ty, don’t do it! You have so much to live for, so many more pedantic comments to make on the Internet!

        Reply
        1. NevadaMark

          Hey Ty, how about this one to tickle your brain? Theoretically, it is possible for a pitcher to have a nine inning complete game shutout and only throw 9 pitches. The catcher in this case would not be involved at all.

          Think about it.

          Reply
          1. Guest

            Perhaps it is theoretically possible for a pitcher to have a nine inning complete game shutout without throwing a single pitch.

            The pitcher must throw a pitch within twelve seconds if there are no base runners. Rule 8.04 The penalty is the umpire calling a ball. Id.

            So the pitcher could do this four times. Then balk twice. Then pick off all three runners. Repeat for nine innings. I’m just not sure if the umpire would eventually eject the pitcher for unnecessarily delaying the game.

            If a pitcher unnecessarily delays the game by throwing to other players without the intent of trying to make an out, the penalty, after a warning, is ejection. Rule 8.02(c). If a pitcher simply unnecessarily delays the game with runners on base, the penalty is a balk. Comment to Rule 8.05(h). The intent of Rule 8.04 is to avoid unnecessary delay. Rule 8.04. The rule also states ” Obvious delay by the pitcher should instantly be penalized by the umpire.” I think there is a colorable argument that “be penalized” is referring to being penalized in the general sense applicable to avoiding unnecessary delay, i.e., ejecting the pitcher after a warning, rather than penalized in the specific sense of Rule 8.04, calling a ball. After all, how could an umpire “obviously” know that a pitcher was going to take more than 12 seconds to throw the ball? And it’s a weird use of “instantly” because, after the 12 second mark, why would the umpire wait around a bunch more seconds before calling ball?

            On the other hand, if a pitcher is intentionally waiting more than twelve seconds in order to walk the batter, then balks two more batters, and then is able to pick off all three batters, and is able to complete this sequence nine times, and thus pitches a complete game shutout without throwing a pitch, the pitcher has obviously come up with some sort of video game-like loophole and is following a deliberate, efficient strategy. As such, can the pitcher be said to be causing “unnecessary delays”? Such a game would almost certainly be much shorter than a standard nine inning game.

          2. Guest

            Fuck. Balks don’t advance the batter, just the runner. None of my theories work. Back to the drawing board. Sorry, Mom.

          3. Ty Sellers

            You’ve got me stumped Mark. Batting out of order is the only thing that comes to mind but I’m not quite sure the penalty for that infraction.

          4. MCD

            Bet guess:

            First batter triples. With second batter at plate, runner attempts to steal home, batter interferes, umpire calls batter out , rules no pitch and sends runner back to third. The third and fourth batters of inning are also called out in same fashion. Repeat all of this for nine innings.

          5. NevadaMark

            MCD nailed it. I got that from an old Ripley’s Believe it or Not paperback in the sixties. It was a special sports edition and it had some weird stuff in it.

    2. Xao

      That depends on how you calculate games: games played or team-games played. Much like work hours versus man-hours, both are useful and valid measurements.

      Reply
  3. cf

    Mitch Williams and Harold Reynolds argument last night about the scoring was priceless. Totally agree with you (and Harold) … it’s not an error and never has been. Only considered doing that because it was a perfecto. I’m a Yanks fan and I was so happy Big Papi ended the discussion. Do not challenge the “Baseball Gods”.

    Reply
    1. JaLaBar

      I am fairly certain Joe doesn’t agree with you, and does agree with the official scorer. If a ball is hit, and 99.9% of the people believe it should be caught, and it isn’t caught, the fact that an error was committed is obvious. Why do we need one of the players to muff the ball to call it an error. If they stand there and watch it drop… conversely, why should the batter be credited with a hit just because the fielder neglected to even attempt to catch a 99.9% catchable ball?

      Reply
  4. Spencer

    @cf

    Did you read the post? I don’t think Joe agrees with you really, and he certainly doesn’t agree with Harold Reynolds.

    Reply
    1. bellweather22

      Joe lost interest long ago. He’ll probably regroup and get them posted at some point, but it’s likely to drag on for months. This much like the BR HOF that died a slow painful death by neglect last year.

      Reply
  5. McKingford

    Oh man, does this ever hit home with me. Errors have become so incredibly arbitrary as to be completely useless. I’m grateful for this play, because it will forever be my go-to argument in the pointlessness of errors:

    http://mlb.mlb.com/assets/images/0/4/8/73655048/cuts/upton_sun_1_cbrcfv4a_641tjx41.gif

    If that is not an error, then NOTHING is an error. I was watching that game, and the Braves announcers were initially bemoaning the play because it was going to end some ridiculously long streak of “errorless” play by the Braves, and were falling over themselves excusing Upton because he obviously lost it in the sun. How is that an excuse?! I mean, you can excuse basically EVERY error in some form or another: doesn’t a defender basically “lose” the ball momentarily before misplaying any “error”? And if the difference between “error” and “hit” is whether the fault of the play should be attributed to the defender or pitcher, are you telling me that inducing a shallow pop up to centrefield is the *pitcher’s* fault?!

    Of course the other dirty secret is that what is or isn’t an error is so completely dependent on whether it’s a home or away game, given the bias of the scorer. I’m still a little bitter because my dad and I flew down to Baltimore the first year Camden was open and saw Charles Nagy 1 hit the O’s. That single “hit” was a sharply hit ball right at the shortstop, who played the ball off his chest and threw half a step late to first. Does anyone really think if that game’s played in Cleveland Nagy doesn’t get his no hitter (ie. that play is ruled an “error”)?

    So, yes, do away completely with errors. The interesting thing about learning about DIPS is that I finally realized that all those heartbreakingly close no-hitters or perfect games (I still remember George Kell’s call when Milt Wilcox’s 8.2 inning perfecto was broken up – “OH NO!”; then there’s Dave Stieb with consecutive starts going 8.2 no hit innings; and Galarraga’s iconic perfecto, together with Yu last night) share one thing in common: they don’t end with a strikeout. The one sure way the pitcher could have brought the game to a successful close was by striking out the last batter and they failed to do so. Once that happens, you take your chances with the defence. See this play at the beginning of the Dave Stieb clip – the last play that breaks up the no-no takes an epically comical bounce over the second baseman’s head; it’s really something out of a video game.

    Reply
    1. McKingford

      Sorry, I didn’t make clear, but that gif of Upton dropping that pop up was scored a hit, not an error. A HIT! Not only a hit, but a double! Which is why I say if that’s not an error, then nothing is an error.

      Reply
      1. NevadaMark

        I was very interested in your comment, Mr. McKingford, because I had recently viewed some highlights on the 1966 World Series, which of course included the famous Willie Davis 3 errors in an inning game.

        The first flyball (and it was an easy flyball) Davis did not touch; it landed right behind his knee. The second (equally easy) he dropped.

        Both were ruled errors. Willie also heaved a throw into the 3rd base dugout but that is not really germane.

        Thought you might want to check it out yourself.

        Reply
  6. Phaedrus

    Serious question here. Let’s say the Royals are playing in Boston. Boston’s leadoff hitter hits a ball high off the Monster for a double. The next batter hits one over the Monster for a home run. Why is the pitcher assumed to have control over the home run, but not the double?

    Reply
    1. Ian R.

      As I somewhat understand it, that’s the sort of issue with DIPS that tends to get worked out in the aggregate. Obviously a deep fly ball off the Monster shouldn’t be treated the same as a little blooper to left, even though they’re both balls in play. But over the course of a season, home runs end up being a decent proxy for the number of really hard-hit balls a pitcher gives up overall – there aren’t many pitchers who give up a ton of deep fly balls without giving up their share of HR.

      Reply
  7. Ian R.

    How much of the decline in errors can be attributed to the soaring strikeout rate? Fewer balls in play, fewer errors. It would be more interesting to look at errors per ball in play over that same span.

    Reply
  8. Old Soul

    I like a lot of the new stats, such as OPS+ and WAR. But I like a lot of the old ones, too.
    For me, the highlight of the off season was watching “Clubhouse Confidential” when Brian Kenny was talking to stats revolution Godfather, Bill James, and James said he liked pitcher’s wins. James explained that although wins often give a distorted picture for one season they work reasonably well over a career.
    I wish I had recorded it because the look on Kenny’s face was priceless.
    And the exchange between Kenny and James illustrates what disturbs me about these debates.
    It’s not enough to have new stats that you feel add to insight about the game. It’s not enough to feel a little smug. It’s not enough that Bert Blyleven gets into the Hall of Fame.
    No, the old stats must be crushed.
    Hey, don’t like errors, don’t pay attention to them. That’s cool. I still will.
    Is it a perfect system? No.
    So?
    What’s the prefect stat?
    And the old stats have something going for them. I mean if it’s all DIP or FIP and line drive rates, is there any drama at the end of a Rangers’ blowout win over the Red Sox in May?
    Is there anything about that game for us to discuss?

    Reply
    1. MisterMJ

      I love the MLB commercials where they present Kenny as some quant whiz kid … there’s one showing him writing graduate-level math equations on a friggin’ pane of glass (a la Beautiful Mind). Not bad for a college BFA to anoint himself the Media Guardian of Sabermetrics.

      Reply
  9. Dennis Toll

    Love the post and the statistical concepts. But I still have a problem with DIPS (and I love advanced metrics, generally). You say that DIPS is trying to measure things over which a pitcher has “demonstrable” control. In that list is strikeouts. But does a pitcher really have “control” over strikeouts? What about that fat, hanging curve ball that floats over the plate for a called strike three when the hitter plain blows it and doesn’t swing. And even when a batter swings, sometimes he just happens to get enough wood on a great pitch to steal a hit, or sometimes he just plain misses a bad pitch over the plate. Doesn’t the hitter share in the responsibility for striking out? There are some hitters who have a lot of strikeouts, some who don’t, so isn’t the “ability” to strikeout as much a part of the hitter’s skills (or lack of) as of the pitcher’s? I think in this case, DIPS is just a little bit dippy in estimating the pitcher’s control over the strikeout. It’s admirable to want to remove the defense’s contribution from measuring pitching skill, but by the same logic, shouldn’t you also want to remove the hitter’s contribution?

    Reply
    1. Bono

      As noted above, the idea is that in the aggregate these issues will level out. Sure, some K’s are “lucky” and some hitters might skew games strikeout numbers. But over a season and especially a career a pitcher sees a great variety of hitters in so many situations that we can attest to his ability or inability (or anywhere in between) to generate K’s. I haven’t done it, but I know that the use of HR, K and BB in DIPS stats isn’t arbitrary but based on correlation analyses across seasons, and those three stats tend to be the most stable for individual pitchers. At the end of the day, you can’t really argue that Darvish gets a ton of K’s simply because he is lucky or only faces hitters who strike out a lot.

      Certainly there are many elements involved in any play in a baseball game, and no one player is every really fully responsible (fielding position, weather, opposing hitter/pitcher, pregame preparation, ballpark, etc.) but DIPS aims to clear out many of the things that were traditionally credited to pitchers that we have little evidence of their influencing, mainly hits. Hits, it turns out, is a stat with very little year to year correlation and so removing it from our pitcher evaluations gives us more reliable, more predictive stats, while also helping us advance out understanding of the batter/pitcher interaction, even if that remains, in many ways, mysterious.

      Reply
  10. Zach

    Despite some raggedness in implementation, I think the error is a useful concept. I mean, try writing a description of Yu Darvish’s game last night that doesn’t include the word “error” or something closely equivalent:

    “Yu Darvish dominated the Boston lineup, allowing no hits until David Ortiz hit a towering popup directly to the second baseman. Although the fielder was clearly in position to make the play with ordinary effort, the ball nonetheless fell safely due to a gross misplay.”

    Calling that play a hit is just obfuscating things. It wasn’t a hit. It was an error.

    Reply
  11. Nom

    Actually, the pitcher doesn’t need to throw a single pitch.

    Delay of game = ball, correct?

    “walk” three batters, pick them off…

    Reply
  12. Pat

    “There has never been a pitcher in baseball history who threw a no-hitter by himself, but if you look up the list of no-hitters you find only pitcher’s names.”

    There’s a great story in the recent Rube Waddell biography about a story close to this. It happened in the minors, but apparently one day Waddell pitched a perfect game with 20 or 21 strikeouts. That’s not the remarkable thing. Of the remaining six or seven outs, Waddell recorded a putout or assist on all but one—I forget how many he threw to first and how many he covered first one, and heck, he easily might have fielded a grounder to the right side and stepped on the bag or applied the tag himself, if the runner was slow enough. The one remaining out, if memory serves, was a flyout to the outfield.

    Reply
  13. Brent

    In response to Fenway Wall Balls above, of course those should be treated like Home runs and probably if someone thinks about it, there is some slight variation in pitchers with Fenway as their home park (and hitters with their BABIP) because of this. Honestly, those balls hit high off the wall in Fenway (and any other parks that have such a wall), should be treated just like HRs when calculating DIPS or BABIP. I assume it isn’t done because it would be too hard to keep track, although going forward it really shouldn’t be. Obviously much more difficult going back in time and trying to calculate them. I believe that BABIP at Fenway and, for an example of a past ballpark where it would be relevant, the Baker Bowl, is clearly arbitrarily elevated by Wall Balls.

    Reply
  14. Moeball

    When Mookie Wilson hit the dribbler to Bill Buckner, Buckner never touched it and it was called an error, so official scorers do occasionally call errors on balls the fielders never touched, but it is rare…what really cracks me up are the ones where the fielder DID touch the ball but they still call it a hit…the official scorer here in San Diego is completely incapable of calling any play an error unless the fielder kicked the ball at least 3 times…

    Reply

Comment: