RBI Men

2017-9

              Let me ask you a few questions about RBIs.. . .basic, pretty obvious questions, but questions to which I would speculate that you will not know the answers:

              1)  Could an average hitter drive in 100 runs in a season, if he had enough opportunities?

              2)  Do the best RBI men, over the course of a season, actually get the most chances to drive in runs?   If so, how many more chances to do they get?

              3)  Did Edwin Encarnacion have a big uptick in RBI last year because he had more RBI opportunities, or because he was more productive with the chances that he had?

              4)  Edwin Encarnacion drove in 127 runs last year and Caleb Joseph drove in none.   Obviously Encarnacion had more RBI opportunities than Joseph, but how many more?   Five times as many?  Ten times?  Three?   What is the real difference?

              Of the 127-RBI difference between them, how many do we attribute to Encarnacion’s performance, and how many to the difference between them in RBI Opportunities?

              5)  To what extent did Caleb Joseph make history last year (141 plate appearances without an RBI) because he just didn’t have a lot of RBI opportunities, and to what extent was it because he didn’t drive in runs when he had a chance? 

              6)  Or, to pick a more relevant comparison than Caleb Joseph, Jean Segura.  Segura had more at bats than Encarnacion did in 2016, more hits, and exactly the same number of total bases as Encarnacion (318 each)—but drove in only half as many runs.    To what extent did Encarnacion drive in more runs than Segura because he had more chances to drive in runs, and to what extent was it because he was a better RBI man?   Do you think it was 70% because Encarnacion was a better RBI man, or 30%?

              7)  The year that Miguel Cabrera won the Triple Crown, was he actually the best RBI man in the American League?

              8)  Who was the best RBI man in baseball last year, comparing his RBI to his RBI opportunities?

              9)  Who is the best RBI man in baseball, year in and year out?

              10)  Is being a good RBI man entirely a predictable function of one’s hitting skills, or is there something about it that you wouldn’t anticipate?

              11)  How many runs do teams gain by putting their best RBI men in the prime RBI positions? 

              12)  One team in 2016 had three of the four best RBI men in their league.   What team do you think it was?

              I’d ask you, if you’re in a nice mood and your basketball team didn’t just lose a big game, to actually re-read that list of questions, and make up answers, guess what the answers are.   If you do that, you’re realize that you are actually learning something meaningful about baseball by reading this article.

              OK, let me explain the method; this won’t take long.   A player’s RBI opportunity are the sum of his actual RBI, and his missed RBI opportunity.   But what is an RBI opportunity?

              If a runner is on third base with less than two out and the hitter does not bring him in, that is a full RBI opportunity (1.0) that is lost.

              If a runner is on second base, or if he is on third base with two out, that is a 70% RBI opportunity (.7) that is lost.

              If a runner is on first base and the batter does not score him, that is a 40% RBI opportunity (.4) that is lost.

              If there is no one on base and the batter makes an out, well, he could have hit a home run.  That we score as a 10% RBI opportunity (.1) that is lost. 

              HOWEVER, a batter is never charged with a lost RBI opportunity if he doesn’t make an out in the plate appearance.    If there is a runner on first, let’s say, and the batter hits a single; the runner goes to second or third, but doesn’t score.  The batter hasn’t driven in the runner from first, but the RBI opportunity still exists and in fact has been improved, rather than lost.  So if the batter doesn’t make an out, he is not charged with losing an RBI opportunity. 

              Recently I had a query in "Hey, Bill"—I wish I could remember who it was from, and I wish I had kept the query somewhere—to the effect that it would seem like RBI Opportunities should be a basic stat that people referred to all the time whenever RBI are referenced, so. . .why isn’t it?   When he asked that I thought. . .yeah, why isn’t it?   And why haven’t we (at BIS) done anything with the eight years of data that we have? 

              We at Baseball Info Solutions have actually been tracking this data since 2009; we can debate the method later on.  Frankly, we have done astonishingly little WITH the data; we have just been tracking it, putting it on a shelf, and forgetting about it.  

              OK, let’s address these 12 questions:

              1)  Could an average hitter drive in 100 runs in a season, if he had enough opportunities?

              In the eight years that we have been tracking the data, no hitter has come close to doing that.   It could in theory happen, I guess, but it would be an extreme outlier event.  

              An average hitter drives in 30.4% of his RBI opportunities, or 7 out of 23.    In the eight years we have been tracking the data, 146 hitters have driven in a hundred runs.   The WORST RBI man to drive in 100 runs in that time was Pedro Alvarez in 2013, at 34.1%, which is still a pretty good distance above average, although the average for a regular player is 33.4%.   Alvarez wasn’t much above average compared to regular players. 

              In order to drive in 100 runs with a below-average RBI percentage, a player would have to have 230 Missed RBI opportunities, as a theoretical minimum.    The most missed RBI opportunities of any player in the last eight years is 205.4.     

              The most runs driven in by a below-average RBI producer in the last eight years are 85, by Robinson Cano in 2009.   Cano has been a good RBI producer most of his career, but in 2009 he hit .376 with the bases empty but just .207 with runners in scoring position, his OPS dropping more than 400 points when there were runners in scoring position.     Cano (2009) was also the player who had 205.4 missed RBI opportunities.   

              Brief digression.   About ten years ago I did some research into the history of the phrase "in scoring position".     I learned, to my surprise, that the expression "in scoring position" was originally used in every sport you can possibly think of, and in most sports more often than in baseball.   I assumed it was a baseball expression which had bled a little bit into the other sports.   This is not true at all.   Though it was used in baseball as early as 1905, it was not often used in baseball through the 1930s.   I found examples of the expression "in scoring position" being used in boxing, basketball, football, tennis, bowling, golf, polo, distance running, many others.   And not infrequently; it was used A LOT in football and bowling, in particular.   For some reason the expression almost entirely died out in other sports in the 1940s, becoming an expression exclusive to baseball.  

              2)  Do the best RBI men, over the course of a season, actually get the most chances to drive in runs?   If so, how many more chances to do they get?

              They do, yes.    Good RBI men frequently get 250 to 300 RBI opportunities in a season, occasionally over 300.     Bottom of the order and top of the order hitters may be more in the 160 to 225 range.    A little of the difference is accounted for by home runs counting as RBI, thus as RBI opportunities, but at least half of it is not.

              That paragraph didn’t QUITE get where I was trying to go.   Teams do, in fact, give more RBI opportunities to players who are better RBI men.   The traditional form of the batting lineup does work, in that sense.  

              3)  Did Edwin Encarnacion have a big uptick in RBI last year because he had more RBI opportunities, or because he was more productive with the chances that he had?

              Entirely because he had more chances.   He had a huge surge in RBI opportunities in 2016, far more than he had ever had before.  He actually had his LEAST productive RBI season in the last five years (in 2016), despite having a career high in RBI.  

              4)  Edwin Encarnacion drove in 127 runs last year and Caleb Joseph drove in none.   Obviously Encarnacion had more RBI opportunities than Joseph, but how many more?   Five times as many?  Ten times?  Three?   What is the real difference?

              Of the 127-RBI difference between them, how many do we attribute to Encarnacion’s performance, and how many to the difference between them in RBI Opportunities?

              Encarnacion had about seven and a half times as many RBI opportunities as Joseph did.   Encarnacion had 321, Joseph had 43.  

              Of the difference between them of 127 RBI, two-thirds is accounted for by RBI opportunities, and one third by the hitter’s productivity.     Given the number of chances that each hitter had to drive in a run, an average hitter would have driven in 97 runs with Encarnacion’s opportunities, and 13 with Joseph’s.      That’s a difference of 84 runs.   Encarnacion was 30 runs better than an average hitter; Joseph was 13 runs worse.    So 84 runs of the 127-run difference are accounted for by opportunities, and 43 are accounted for by productivity.  

              5)  To what extent did Caleb Joseph make history last year (141 plate appearances without an RBI) because he just didn’t have a lot of RBI opportunities, and to what extent was it because he didn’t drive in runs when he had a chance? 

              One-third because he didn’t have a lot of chances; two-thirds because he didn’t produce.   An average American League hitter last year had .47 RBI opportunities per plate appearance.   With 141 plate appearances, Joseph should have had 66 RBI Opportunities.    In fact, he had only 43—about a third less than random expectation.  

              The 23 Opportunities he was missing should produce 7 RBI—remember, 7 out of 23.   The 43 chances he DID have should have produced 13 RBI.    So he’s short by 7 RBI because of circumstances, and by 13 because he didn’t produce. 

              The 43 RBI Opportunities with no RBI is the most in our data, obviously.   Second on that list is Hiroki Kuroda, 2011, at 32.3.    Kuroda had an unusual season in that he had a LOT of RBI opportunities, for a pitcher, and never came through, but he is, after all a pitcher.   

              Joseph is first on that list, and then the next 36 players are all pitchers.   The most RBI opportunities without an RBI, after Joseph, is 19.6, but Koyie Hill in 2013.  

              6)  Or, to pick a more relevant comparison (for Encarnacion) than Caleb Joseph, Jean Segura.  Segura had more at bats than Encarnacion did in 2016, more hits, and exactly the same number of total bases as Encarnacion (318 each)—but drove in only half as many runs.    To what extent did Encarnacion drive in more runs than Segura because he had more chances to drive in runs, and to what extent was it because he was a better RBI man?   Do you think it was 70% because Encarnacion was a better RBI man, or 30%?

              Of the 63 RBI difference between them, 40 were because Encarnacion had more chances to drive in runs, and the rest were because Encarnacion hit 42 homers to Segura’s 20, and a Home Run is also an RBI.   Setting aside the larger number of RBI Opportunities for Encarnacion and the fact that he drove in himself 22 more times with homers, they were otherwise the same as RBI men. 

              7)  The year that Miguel Cabrera won the Triple Crown, was he actually the best RBI man in the American League?

              He was not, no.    Cabrera drove in 11 more runs than Josh Hamilton of Texas, but had almost 50 more RBI opportunities, or 18% more opportunities.  

              8)  Who was the best RBI man in baseball last year, comparing his RBI to his RBI opportunities?

              Daniel Murphy.   Murphy had 104 RBI with 224 Opportunities, or 46.4%.     These are the league leaders in each league since 2009:

American League

2009     Jason Bay

2010     Jose Bautista

2011     Jose Bautista

2012     Josh Hamilton

2013     Chris Davis

2014     Mike Trout

2015     Josh Donaldson

2016     Mookie Betts

National League

2009     Albert Pujols

2010     Joey Votto

2011     Ryan Braun

2012     Garrett Jones

2013     Freddie Freeman

2014     Devin Mesoraco

2015     Nolan Arenado

2016     Daniel Murphy

              At the end of the article I’ll run a chart that lists these side by side with the total RBI leaders. 

              9)  Who is the best RBI man in baseball, year in and year out?

              You probably all got this one right.   It’s Miguel Cabrera. . ..well, Cabrera or Trout.   Trout if you consider a player eligible at 1,000 opportunities, Cabrera if you use 1,200.   Cabrera has never been the best RBI man in the American League in any one season since we’ve been tracking this, but he is always close:

              He didn’t lead in 2013 because he happened to run up against a guy (Chris Davis) having the best RBI season in the last eight years, at .483.     But over the eight-year run, Cabrera and Trout are the best RBI man in baseball:

              The WORST RBI men to get 500 RBI opportunities over those eight years are these guys:

              I’ll try to post these grades for all hitters tomorrow. . .not sure if I can get that done, but I’ll try.  

              10)  Is being a good RBI man entirely a predictable function of one’s hitting skills, or is there something about it that you wouldn’t anticipate?

              Well, you now know the answer to that question, because of the chart I just gave you, with all of the great hitters at the top.   It’s largely a predictable function of the player’s production.  

              In a single season, things happen.   The best RBI man in the National League in 2014 was Devin Mesoraco.   Mesoraco is a career .237 hitter with only 16 career homers, other than the 25 he blasted in 2014, in just 384 at bats.    That year he hit .241 with the bases empty, his normal average, but hit .315 with men on base, and homered 65% more often when there were men on base.   With the bases loaded he went 7-for-9 and hit three homers.   That year, he was the best RBI man in the National League. 

              That is actually the second half of a postman-always-rings-twice double image.   Mesoraco was the Reds catcher.    In 2010  the Reds backup catcher was Ryan Hanigan, who had a very similar year.    Hanigan—a career .250 hitter—hit .300 in 203 at bats, but, more particularly, hit .376 with runners in scoring position.   He hit four of his five home runs with men on base, although more than 60% of his at bats were with the bases empty, and he went 5-for-9 with the bases loaded.   In 2010, although he didn’t have enough RBI opportunities to be considered the league leader, he had a higher RBI production rate than the league leader.  

              But both Hanigan and Mesoraco came back to earth the next season.   There’s really no such thing as an ability to hit .376 with runners in scoring position, if you’re a .250 hitter otherwise.     If it happens, it is just something that happens.  

              I know I have talked about this before, but I have always been fascinated by what I call Floyd Robinson seasons.   Floyd Robinson in 1962 hit .312 with 11 homers, but 109 RBI.   He hit 45 doubles, 27 of them with men on base.   Robinson, batting behind Joe Cunningham, who had a .410 on base percentage, had 384 plate appearances with runners on base, an enormous number, and hit .347 with men on base.  

              I was twelve years old at the time; I figured this was Robinson’s real skill level, and he would do that every year. I didn’t know how many plate appearances he had had with men on base, or that he had hit .347 with men on base, or that there really is no such thing as an ability to hit with runners in scoring position.    I just figured this was who he was; this is what he would do.    Of course he wasn’t able to sustain any of the elements of magic that had made that a magical season.   The next season he batted just 301 times with men on base—still a high number, but not 384—hit just .287 with runners on base, and hit just 7 doubles with men on base, down from 27 the previous year.  

              In 1970 Wes Parker had a very similar season, hitting .319 with 10 homers, 112 RBI.     He also hit 47 doubles.    He had come to the plate 349 times with men on base, and 231 times with runners in scoring position.    By now I was a young adult, a little more skeptical, but still intrigued to see if Parker could repeat his Floyd Robinson season.

              Periodically a player has a year like that.    Keith Hernandez in 1979.   48 doubles, 11 homers, 105 RBI.    He had hit .385 with men on base.   Hernandez was a great player, but he never drove in 100 runs again.    RBI are an essentially predictable outcome of your other numbers, over time—varying, of course, with whether you hit leadoff or cleanup, if you are Rickey Henderson, and with whether you hit sixth or eighth, if you are Leonys Martin. 

              11)  How many runs do teams gain by putting their best RBI men in the prime RBI positions? 

              About ten to fifteen runs in theory, although in practice that is probably an overstatement.   A typical team has about 2200 RBI opportunities in a season, which are arranged high to low something like this:

              And a typical team has RBI producers of varying ability, something like this:

              If you maximize that so that the player with the most RBI ability gets the most RBI opportunities, the team will get 710 RBI.   Well, 711:

              If you randomize it—a random batting order—you’ll usually get somewhere between 695 and 700 RBI.    You can go as low as 684, if you make the batting order as illogical as possible.   You have a gain of about a dozen RBI by putting the best RBI producers in the prime RBI slots, or 27 RBI from the theoretical best to the theoretical worst. 

              Realistically, it is less than that, for two reasons.   One is that, at random, Miguel Cabrera might hit ninth.  In the real world, that’s not going to happen.   MOST of the distance from 684 to 711 is covered by just doing really obvious things like putting Miguel Cabrera somewhere in the middle of the order.  

              The other is that there are second-order effects which would tend to slightly reduce the measured difference.    There is a runner on second, one out.    There is an RBI potential of .7.    If the batter hits a single, the next hitter up has an RBI potential of .4.   If the batter strikes out, the next hitter up has an RBI potential of .7.    If one hitter produces the RBI, that reduces the RBI potential for the next hitter, if we ignore the "table-setting" half of the equation, which we are ignoring in this analysis.  

              12)  One team had three of the four best RBI men in their league last year.   What team do you think it was?

              It was the Red Sox.   These are the top ten RBI producers in the American League last year:

              The Red Sox had a terrific offense in part because they had three of the best cleanup hitters in the league.   One of those is retired now, and. . .you don’t want to EXPECT Mookie to perform at that level every year.    Whoever led the league last year, usually doesn’t this year.   Mookie is certainly not Floyd Robinson or Devin Mesoraco, but his season had quite a bit of magic to it, like Fred Lynn’s season in ’75.   He hit .355 with Runners in Scoring Position.   You can’t expect him to do that every year.  

              OK, before I go, I should explain a little bit about why the system is the way it is.     I know very well that a runner in scoring position does not have a 70% chance of scoring.   That is NOT what we are measuring here, and that is not a relevant fact.  

              It would be easier to design THIS measurement if we were measuring teammates batted in—TBI.   That would be easier.   We have to work with the definition of RBI that we have inherited. 

              I tried to set it up so that an average hitter has about the same level of measured success in each situation, more or less, and consistent with the goals of simplicity and conceptual clarity.    Let’s say a player makes 400 outs in a season, and hits 18 homers.   If those all come with the bases empty, his RBI percentage is .310—18 RBI in 58 opportunities.   Each out (with the bases empty) is .10 RBI lost.  

              Let us say he hits .280 with a runner in scoring position. . . .let’s say 50 for 180, which is actually .278, and let us say that 40 of the 50 runners score.    He’s got 40 RBI, and he is charged 91 lost opportunities (180 minus 50 is 130, times .70 is 91).   So he is 40 for 131, which is .305.  

              Let us say he bats with a runner on first only 150 times, hits .280, which is 42 for 150.   Let’s say that includes 12 doubles, 2 triples, 5 homers.   He is going to have about 17 RBI there.   If he is charged with .40 RBI opportunity when he makes an out, that’s 43.2. . . he is 17 for 60.2, or about .280.  

              I was trying to set it up so that, whether he hit doubles, singles, homers, whatever, his RBI production ratio would be about the same one way as another.     What matters is whether he produces or whether he doesn’t—if he is an average hitter.

              If he is a home run hitter—well, of course that makes him a better RBI man when the bases are empty.    That’s inherent in the definition of RBI that we have inherited from our ancestors; three point penalty to the writer for using "inherent" and "inherited" in the same phrase.     I think it is a reasonable premise, and a reasonable approach to measuring RBI opportunities.   You can make of it what you will. 

              Thanks for reading.   Tomorrow I’ll try to post the career ratings for everybody.    Here’s the list I promised you earlier. 

              There have been ten Floyd Robinson seasons in the majors since the end of World War II.    What you will notice about those guys is that they mostly lost about 25 RBI the next season.    Hernandez drove in 99 runs in his followup season, but then, he was actually really good.