Fraction of balls put in play is at an all-time low

Here’s a plot showing the percentage of balls in play each year. Specifically, this refers to the fraction of plate appearances that result in a ball being handled by the defense.

BIP

The formula uses at-bats in the numerator, subtracting out home runs and strikeouts, and adding sacrifice hits and sacrifice flies. That total is then divided by plate appearances, which of course includes walks and hit-by-pitch.

As you can see, 2013 is on pace to have the lowest percentage of balls put in play in MLB history. And even though 2013 is far from over and could change, 2012 itself set the all-time record, at just 68.7%.

The two biggest factors are, of course, home runs and strikeouts. Even in this year of very low offense, home runs are still quite high. I wrote about that in my USA Today Sports Weekly piece this week. Strikeouts continue to go higher and higher, and drive the percentage of balls in play lower and lower.

Some comments and implications about the above graph:

  • Think about 2013, at 68%, vs baseball in the 1940’s, around 80%. With teams averaging around 38 plate appearances per game, that’s a difference of nearly 5 balls in play per game. Think about that–5 batted balls fewer per game, every game! That’s astounding.
  • The fewer balls in play also means that defense matters less. In the current game, defense is the best it’s ever been. Equipment and fields are of uniformly excellent quality and players have better range than ever. However, the lack of balls being put into play means that the defense has fewer chances, so overall, there is less variability on defense from team to team.
  • We think of 1968 as the year of the pitcher, but 74.2% of balls were still put in play that year, meaning the defense played a much bigger role.
  • Take a look at fielding metrics over the years, here. Putouts have remained the same over the years, because they are basically all outs. But total chances has dropped gradually over the years, from about 41 in the 1920s to 39 in the 1940s to about 37.5 in recent years. This number has fallen a bit because errors have continually dropped, but more because assists have dropped. And why have assists dropped? Because a higher fraction of putouts are to the catcher, i.e. strikeouts, when there can be no assist.

The game is changing, for sure…

 

38 thoughts on “Fraction of balls put in play is at an all-time low

  1. 1
    Pseu says:

    The improved defense makes the “swing for the fences” strategy even more attractive, as it shrinks the “cost” of a strikeout.

    • 2
      Timmy Pea says:

      I agree, most players don’t shorten the swing on 2 strikes anymore. Great stuff!

    • 4
      no statistician but says:

      Pseu:

      I hope you’re being humorous. The differential in fielding 2012/1960 is less that 1 error per three games. I don’t know how to figure the % of errors that result in runs, but is has to be considerably less than the number of errors, probably less than half.

      Plus, how much of that “improved” defense results from the all or nothing approach? It’s impossible to say, of course—stats guys, correct me if I’m wrong—but I’d guess there’s a small correlation.

      • 5
        Andy says:

        Actually, in my recent conversation with Reggie Jefferson (referenced in the USA Today piece), he said exactly what Pseu said–that players today know the likelihood of a ground ball turning into an out is higher than ever, and therefore guys want to hit fly balls, and if you hit a fly ball, you want it to go over the fence.

        • 10
          no statistician but says:

          Andy:

          I don’t doubt you or Reggie Jefferson, but is my figure wrong? Less than 1 error in three games difference? since 1960? I’d say rather that players today “know” what they think they know that supports the point of view they already have decided to take. It’s nothing new. People do it all the time in all walks of life.

          Separating conception from reality is hard for rocket scientists, much less ballplayers.

      • 9
        Andy says:

        See #3 below–the issue isn’t so much errors as it is shifts and range. Players make fewer errors, but they get to so many more balls now than they used to.

  2. 3
    brp says:

    I think there’s another, bigger factor in defense in the past 10 years or so – shifting. Especially the dramatic shifts with 3 guys to the left/right of 2nd base, or a guy playing short right field for lefty power hitters, etc. Maybe I was too young to remember, but I don’t recall teams shifting/shading/etc., nearly as much until somewhat recently.

    The defensive skill levels of today’s players doesn’t seem that much more impressive than the guys I grew up watching (Vizquel, Ozzie, Alomar in the infield, Griffey or Andrew Jones in the OF, etc.), and the caliber of the equipment and fields also does not seem have changed much recently. I guess there are fewer turf fields so fewer goofy bounces in that regard.

    • 7
      Andy says:

      Most definitely true, brp. Great point.

    • 12
      Doug says:

      Fewer turf fields also contributes to fewer ground balls getting through the infield.

      Defensive efficiency on ground balls can be gauged by these stats:
      2006: .299 BABIP, 1.15 GB/FB ratio
      2013: .290 BABIP, 1.26 GB/FB ratio

      So, despite the hitters’ best efforts to elevate the ball, they seem to be losing that battle with the pitchers.

      That 9 point drop in BABIP is big – matches the high and low BABIPs over the past 20 seasons (since 1993). In the 20 years before that (1973-92) there was also only a 9 point range in BABIP (.274-.283), excepting the freak year of 1987 (.285). Which makes the 9 point jump from 1992 (.281) to 1993 (.290) look really suspicious (PEDs ?), more especially since that .290 would become the floor BABIP for the following 20 years.

      • 13
        bstar says:

        Doug, I just don’t see how PEDs could be the major cause of a 9-point increase in BAbip from ’92 to ’93.

        I’m not saying steroids weren’t a big factor in the increase in offense from ’93 on (of course they were) but I fail to see how the steroid effect could be there in full force one year (’93) and completely absent the previous year (’92). That’s just not logical.

        Did a couple hundred major leaguers get together over a hot stove in the winter of ’92 and decide to start injecting themselves with steroids en masse? (sorry for the snark, but it seems unavoidable right here.)

        We all know why offense went up in ’77 and ’87, don’t we? So why can’t the same be true for ’93? It’s not a conspiracy theory, or some wild postulate invented by a couple of writers.

        In fact, it’s exactly what the hitters and pitchers were saying, if one recalls, in the summer of ’93: the ball was indeed livelier.

        • 14
          Richard Chester says:

          I remember Ed saying more than once that the ball was juiced up a bit for the 1993 season.

          • 37
            mosc says:

            I don’t buy the ball thing. Coors field, expansion itself, these were all huge factors. The next biggest factor was the general weight lifting/GNC craze of the early 90s. Prior to that, being bulked up was somewhat synonymous with being slow. Pitchers, for example, trained primarily by distance running and not weight lifting. There was a huge shift all across atheltics at all levels during those years to much more heavily favor weight training and it came to a peak in the mid 90s.

        • 15
          Ed says:

          Bstar – But isn’t there an even more likely explanation…namely expansion? Doesn’t offense tend to increase when there are new teams due to the lack of availability of quality pitching. And of course one of the new teams in ’93 was Colorado.

          • 17
            Andy says:

            I think you’re both likely to be right. I think steroids were part of it, but bstar’s right that they wouldn’t cause such a dramatic league-wide jump. Difference in the ball as well as expansion are more likely to cause an abrupt change. The other thing, though is psychology. As we see today, players have a different approach. In 1993, as some players bulked up dramatically, I think most players started trying to jack homers, even before they’d started juicing.

          • 18
            bstar says:

            Good point, Ed. I had considered expansion but didn’t think it would be that big of a factor. I’m at least slightly wrong. Here’s the years of expansion with the corresponding increase in BAbip:

            1961 +.002
            1962 +.002
            1969 +.007
            1977 +.006
            1993 +.009
            1998 -.001

            On average, that’s about a 4 point increase in BAbip for expansion years, excluding all other factors. So yes you are right, Ed. I should have attributed part of it to that.

            But then how do you explain 1987? Just saying it was a fluke doesn’t work for me.

          • 19
            bstar says:

            Andy, I’ve been looking for someone to agree with me about the different ball post-’92 since I came to this site, so your at-least-partial acceptance of the idea fills that need. Thanks.

          • 20
            no statistician but says:

            bstar @ 18:

            The ball was juiced in 1987. Have you forgotten? Wade Boggs hit 24 HR; he only made double digits one other time(11). Andre Dawson hit 49,32 his highest otherwise; Keith Moreland 27/16; George Bell 47/31; Alan Trammell 28/21; Juan Samuel 28/19; Ozzie Virgil 27/19; etc. Numerous players had their highest or second highest HR total that year; several younger players like Mark McGwire had more HRs that year than they would produce until the late 1990s; several other players had their franchise high in HRs, and only did better later when they moved to friendlier confines.

            I, personally, would rather see the ball juiced than the players.

          • 27
            Ed says:

            Bstar – You also need to account for the “Mile High Effect”. Colorado had by far the highest BABIP (.333); that alone accounts for about 2 of the 9 point increase.

            Beyond that, if the ball were juiced, wouldn’t you expect to see fairly uniform BABIP increases across all stadiums? Instead we see some ballparks with large increases but others with decreases. For example, BABIP decreased from .303 to .290 in Cleveland whereas it increased from .268 to .289 in Oakland.

            Same thing in ’87 btw. For example, BABIP in the Metrodome decreased from .305 to .283 whereas it increased from .286 to .325 (holy crap!) in Milwaukee County Stadium.

            BTW Doug, where are you getting your BABIP numbers? I’m looking on the league batting splits and finding different numbers. I’m still showing a 9 point increase from ’92 to ’93 but from .285 to .294, rather than the .281 to .290 that you cited.

          • 31
            bstar says:

            nsb @20: That was my point!! The ball was juiced!

        • 15
          Ed says:

          Bstar – But isn’t there an even more likely explanation…namely expansion? Doesn’t offense tend to increase when there are new teams due to the lack of availability of quality pitching. And of course one of the new teams in ’93 was Colorado.

          • 29
            Andy says:

            Responding actually to your comment at #27. It’s an interesting point, that BABiP varies a lot league-wide. I presume this is down to the actual strength of the pitching. The Twins put up a surprisingly good (lucky?) year in 1987, and I’d guess that a large part of this was pitchers yielding fewer line drives and thus a lower BABiP. Sometimes I think we take the notion that a pitcher has no control over the fate of a batted ball too seriously, or forget that if batters make more contact, they will get more hits. So I don’t really see a problem with large team-to-team fluctuations in the context of the league-wide average trend. The real question would be how the changes in BABiP in expansion years compare to the magnitude of changes in other years.

          • 30
            Ed says:

            Oh sure Andy. Obviously I wasn’t doing any sort of in-depth investigation. I just like to be the contrarian so that multiple views are discussed.

  3. 6
    Kenny says:

    The USA today piece is extraordinarily clear and certainly worth reading through to the end. Nice work Andy…

  4. 21
    John Autin says:

    Good piece, Andy. Especially timely on a day that saw the 2nd and 3rd pitcher games in this week with 1 baserunner over 9 innings.

    And it seems to me that the strikeout trend is finally getting some mainstream attention, so I’m hoping the conversation keeps growing. Occasional dominant games are exciting, but this brand of baseball in general is not so much fun to watch.

    BTW, it’s interesting that so far, the rise in K rate is concentrated among starting pitchers. The reliever K% is virtually the same as last year’s, but the SP K% is up from 18.7% to 19.3%.

    On a tangent, I’m puzzled by your mention of this year’s “very low offense.” Through Thursday, runs per game were not quite 2% below the 2012 full-season rate, and not quite 1% lower than 2011. Given the generally horrid weather so far, I expect this year’s scoring rate to be higher than last year’s.

    P.S. The current HR rate, per game or per PA, would fit right in with most seasons of the PED era.

    • 26
      Andy says:

      JA, what I really meant by low offense is low offensive production. If you read the USA Today piece, I reason that the amount of production, such as advancing of base runners, is remarkably low this year thanks to the increasingly feast-or-famine approach at the plate.

    • 32
      bstar says:

      Good point about reliever Ks not going up this year. Here’s the top 5 K/9 relievers from 2012 and how they’re faring this year in that stat, ignoring the 2013 SSS for now:

      Player name: 2012 SO per 9/2013 SO per 9, % change

      Craig Kimbrel: 16.7 / 13.2, 21% decrease
      Aroldis Chapman: 15.3 / 12.9, 16% decrease
      Antonio Bastardo: 14.0 / 7.9, 44% decrease
      Jason Grilli: 13.8 / 14.1, 2% increase
      Kenley Jansen: 13.7 / 11.5, 16% decrease

      So a lot of the big guns out of the ‘pen just aren’t as dominant as they were in 2012. This isn’t that unusual, given that so many relievers can have one or two great years and never be heard from again (hello, Fernando Rodney).

      But Kimbrel, Kenley Jansen, and Chapman are 1-2-3 all-time for SO/9 based on their # of IP, so it’s a bit surprising to see all three not destroying hitters right now.

  5. 22
    e pluribus munu says:

    Great post, Andy. And your USA Today article went from one sharply reasoned point to another. I hope your prediction at the close is correct – I’ve begun to picture baseball devolving into a game of K’s, BB’s, and HR’s, with nothing else but the occasional excitement of a WP.

  6. 24
    Mike L says:

    Nice piece, Andy. I wanted to also suggest two possible pitching trends. The first is the increase in the number of pitchers on the roster, for the primary purpose of creating platoon advantages. The second is the marked increase in the number of pitchers who just throw extremely hard.

  7. 33
    Doug says:

    Ed and Bstar,

    The BABIP data I related are from FanGraphs.

  8. 34
    MikeD says:

    As someone who came of age baseball wise in the 1970s, it’s not surprising to see the blip upwards during that decade. The style of hitting was probably a reaction to the 1960s, the Year of the Pitcher and the rise in turf parks where teams wanted to put the ball in play more. Hitting coaches such as Charlie Lau made their name during that time.

    • 35
      Andy says:

      Definitely true. In the 1970s there was a concerted effort to reduce strikeouts. By the 1980s I think people (even in the absence of WPA data) began to intuitively understand that this led to a lot of weak ground balls, double plays, etc, and should be limited to those with really good bat control.

  9. 36
    mosc says:

    Can we break down this drop off in BIP by position? I would suspect particularly the drop off is at first base and third base chances.

  10. 38

    It’s nearly impossible to find well-informed people for this topic, but you sound
    like you know what you’re talking about! Thanks

Leave a Reply

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