Who Invented Runs Produced?

Referral to the glossary of statistical terms in the first edition (1989) of Total Baseball by John Thorn and Pete Palmer allows one to easily find not only the meaning and utility of numerous baseball statistics but also the persons credited with inventing them.1 For example:

Assist average. Assists divided by games played. Stat created by Philadelphia baseball writer Al Wright in 1875.

Average bases allowed. A pitcher’s total bases allowed, divided by his innings pitched—what might be termed opponents’ slugging average. Created by Alfred P. Berry in 1951.

Linear weights. A system created by Pete Palmer to measure all the events on a ball field in terms of runs.

On-base percentage. Created by Allan Roth and Branch Rickey in its current form [hits plus walks plus hit by pitch, divided by at-bats plus walks plus hit by pitch] in the early 1950s.When OBP was adopted as an official stat in 1984, the denominator was expanded to include sacrifice flies.

Runs created. Bill James’s formulation for run contribution from a variety of batting and base- running events. In its basic expression, the formula is [(hits + walks) x (total bases)] divided by (at- bats + walks).

Total average. Tom Boswell’s formulation for offensive contribution from a variety of batting and baserunning events. The concept of the numerator is bases gained; that of the denominator is outs made: [total bases + steals + walks + hit by pitches] divided by [at-bats − hits + caught steal- ing + grounded into double plays].

Runs produced. Runs batted in plus runs scored minus home runs.

Curiously, the inventor of the runs-produced (RP) statistic is not mentioned. Similarly, other sources of such information have provided only the definition or formula for runs produced—nothing at all about its creator. For example, the third edition (2009) of The Dickson Baseball Dictionary gives the following for runs produced: “An informal statistical measurement that equals runs scored plus runs batted in, minus home runs. Of unknown origin, the measure was evaluated by Bill James (Baseball Abstract, 1987).”2

Because of my interest in determining who has the major-league record for the longest consecutive- games run-produced (CGRP) streak,3, 4 I deemed it appropriate to find out who created the runs-produced statistic.5

So, who did invent runs produced?6

Here’s the fascinating chronology of my discovery. In a phone conversation with fellow SABR member Seymour Siwoff (Elias Sports Bureau), I mentioned my CGRP-streak research (and the need for accurate data for runs and RBIs alike on a game-by-game basis). Seymour told me that he recalled runs-produced stats first being presented in Sports Illustrated—a couple of years after its first year of publication, which was 1954. Similarly, in an email exchange with Pete Palmer, Pete thought that runs produced “was introduced by Sports Illustrated, maybe in the ’50s or ’60s.”

With that lead, I went through every “baseball season” issue of Sports Illustrated from 1955 through 1964, looking for anything on runs produced. Here’s what I came up with:

1955. Nothing at all on runs produced. In each weekly issue, SI included an information box (titled “Major League Baseball”) that gave the scores of the previous week’s games and the indi- vidual leaders in BA, RBI, HR, and pitching W–L.

1956. Each weekly issue of SI included “The X-Ray Box,” which (in addition to the usual stats) presented a chart for the top five “runs produced” leaders for each league.

In the first baseball-season issue (May 14, page 52), the column headings in the runs-produced chart were

• Player’s name (team and batting average)
• Runs Scored
• RBI
• Total Runs Produced

In the next issue (May 21, page 46), the column headings in the RP chart were

• Player’s name (team and batting average)
• Runs Scored
• Teammates Batted In
• Total Runs Produced

Note the difference for the third column heading in the first two issues—“RBI” (i.e., Runs Batted In) in the May 14 issue and “Teammates Batted In” in the May 21 issue.

It is pointed out that SI provided no explanation whatsoever of the change from “RBI” to “Teammates Batted In.”

For the remainder of the 1956 baseball season, the column heading “Teammates Batted In” was used.

In the end-of-the-season “X-Ray” (October 7, page 55), the distinction between “RBI” and “Teammates Batted In” is crystal clear (though not expressed by SI). For example, Mantle (the AL RBI leader) is listed in a chart of “month-by-month leaders” with a total of 130 RBI; in the Runs Produced chart, he is listed with 78 teammates batted in. Thus, “Teammates Batted In” is equal to RBI minus HR (Mantle having hit 52 home runs in his 1956 triple-crown season).

It is emphasized that in none of the baseball articles accompanying “The X-Ray Box” was any mention made or discussion given of runs produced (or of “teammates batted in”).

1957. Each weekly issue of SI was organized essentially just like those in 1956—“The X-Ray Box” included a runs-produced chart with the same column headings:

• Player’s name (team and batting average)
• Runs Scored
• Teammates Batted In
• Total Runs Produced

1958. Exactly the same as in 1957 (and 1956).

1959. The “X-Ray Box” was replaced by “Baseball’s Week,” which included text by Les Woodcock as well as some performance charts, including “Runs Produced,” which was exactly the same as those employed in 1958 (and 1957 and 1956)—with one significant midseason addition. Beginning with the July 13 issue (page 10), and continuing for the rest of the baseball season, the column heading “Team- mates Batted In” was asterisked, the asterisk directing the reader to the explanation “Derived by subtracting HRs from RBIs.”

1960. Same as in 1959—“Baseball’s Week,” which included text by different authors as well as some performance charts, including “Runs Produced, which for the column heading “Teammates Batted In” had an asterisk indicating the explanation “De- rived by subtracting HRs from RBIs.”

1961. Same as in 1960.

1962. Identical to 1961.

1963. Similar to 1962. However, the performance charts provided only runs-produced information— no columns for “Runs Scored” and “Teammates Batted In.”

1964. No performance charts; text only—no mention of runs produced.

In none of the baseball articles published in Sports Illustrated from 1955 through 1964 was any mention made of the creator of runs produced; likewise for the period 1965–2008.7, 8

So, I wrote the following summary and emailed it to Seymour Siwoff, Pete Palmer, and John Thorn:

The batting performance statistic, “Runs Produced” (which is defined as Runs Scored plus Runs Batted In minus Home Runs) first appeared in Sports Illustrated in 1956 (May 14 issue, page 52). A “Runs Produced” chart was included in nearly every issue of SI during the baseball season from 1956 through 1963. The specific person(s) deserving credit for creating the “Runs Produced” statistic has/have not yet been identified.

Thorn wrote back the following: “This sticks in my memory—that the inventor of the SI Runs Produced formula was none other than Bob Creamer.”

That Bob Creamer could have been the creator of the runs-produced statistic seemed unlikely to me for the following reason. In the Sports Illustrated issue with the very first presentation of the runs-produced statistic (May 14, 1956), “The X-Ray Box” accompanied the article “End of Round One” with the by-line “Baseball by Robert Creamer.” (“Round One” referred to the fact that each club was supposed to have played every other team in its league at least once.)

While Creamer provided his assessments on the round-one performance of each team in each league, he gave only brief mention to individual performances (through May 6), and then only of a few players— Mickey Mantle (who was ahead of Ruth’s 1927-season homer pace), Whitey Ford (who won his first four decisions), Bill Wight (who lost his first four starts), and Cardinals pitchers Tom Poholsky, Jackie Collum, and Ellis Kinder (who combined to pitch a “rare three-man shutout”). Creamer made absolutely no mention of runs produced.

So, I asked Thorn for contact information for Bob Creamer. Thorn responded that, while he didn’t have contact info for Creamer, another SABR member might—Marty Appel. (Appel had been public-relations director for the New York Yankees during the middle 1970s and is the author of the book, Now Pitching for the Yankees: Spinning the News for Mickey, Reggie, and George [Kingston, N.Y.: Total / Sports Illustrated, 2001].)

I sent an email summarizing the situation to Appel, asking for Creamer’s contact information. Appel replied: “Happy to provide it; very interesting story. I’ll be surprised if Bob was the creator, in that I don’t see him as a ‘stat guy,’ but you never know!”

Next, I sent an email to Creamer, including some of the salient points from above. I concluded my missive with the following:

So, I wanted to contact you to find out if you are the creator of the Runs Produced stat. I would greatly appreciate it if you would please let me know if you did indeed originate Runs Produced and your recollections of SI including RP in their weekly coverage of baseball during the 1956– 1963 period.

I was hoping that, even if Creamer was not the inventor of runs produced, he would recall who was.

*****

Pay dirt! Two days later, Bob Creamer sent the following email to me:

My computer has been down—it’s still not working right—or I’d have answered your email before this.

I’ll be honest and admit that I was delighted to get your message.

Yes, in 1956 in working up a weekly stat report for Sports Illustrated I suggested the Runs Produced idea. Les Woodcock, another original member of the SI staff, worked closely with me and helped refine it. At first I thought adding runs scored and runs batted in was enough, but that gave an over-preponderance to home-run hitters, who got two RPs for a home run, the one they batted in and the one they scored. To level the playing field, so to speak, and to give more weight to less powerful hitters who nonetheless seemed to get around the bases and score a lot, we arbitrarily decided to deduct home run totals.

The Runs Produced stat was sometimes dismissed by mathematical purists and I confess I was surprised and pleased when Total Baseball included it among its many measures of batting performance. I’d always felt that despite its mathematical flaw it was a good honest way of evaluating an offensive player’s worth. I had that belief reinforced in the 1950s by my great friend Seymour Siwoff of Elias, who said something to the effect that while it may not be mathematically valid, “It works!” (Seymour, who was a tremendous help to us at Sports Illustrated in those early days of the magazine, often spoke with exclamation marks in his voice.)

The Runs Produced stat had its origins a decade earlier, in 1946, just after World War II, when a bunch of us returning from military service to southern Westchester County formed a softball team and joined a Sunday league (Sunday because lots of people still had to work Saturdays in those days and evening games were difficult for guys who commuted to jobs in New York City). I was the manager for some reason and because I was smart enough to keep myself on the bench most of the time (I wasn’t much of a player) I was able to keep a meticulous scorebook of all our games. Because of my fondness for stats (I love Marty Appel but he pegged me wrong on that one) I kept working up lists of team leaders in various categories.

I had an On Base Percentage that included not just hits and bases on balls but getting on base because of errors. We had a little right-fielder who batted about .220 but could bunt beautifully and was fast as a rabbit going down the line to first, with the result that pitchers, catchers and infielders hurrying to throw him out made error after error. Jay had a very high OBP and a remarkably high number of runs scored. We also had a rotund third-baseman who could hit and drive in runs but who didn’t get around the bases to home plate that often. Jay would be high on the list of runs scored, and Fred would be among the leaders in RBIs. I got the idea of adding runs scored and runs batted in to see who overall were the best run producers on the team.

That Runs Produced figure worked well in softball because we didn’t play on a fenced field and home runs were hard to come by. But when Les Woodcock and I applied the Runs Produced idea to major league baseball it became distorted by the great number of homers, which led us to the idea of deducting them from the overall total. And there we were.

How I do run on. Sorry for the length, but it was a pleasure.

So: Mission accomplished!

Bob Creamer (with refinement input from Les Woodcock) is the inventor of runs produced.

 

Acknowledgments

It is a pleasure to thank all the persons who contributed to this chronology—Seymour Siwoff, Pete Palmer, Gary Stone (who helped me search some of the issues of Sports Illustrated), John Thorn, Marty Appel, and, especially, Bob Creamer.

 

Notes

1. John Thorn and Pete Palmer, , Total Baseball, 1st ed. (New York: Warner Books, 1989), 2286–93.
2. Paul Dickson, The Dickson Baseball Dictionary,” 3d (New York: Norton, 2009), 729.
3. Herm Krabbenhoft, “Lance Berkman Joined Select Group of Run Producers in 2008,” Baseball Digest, May 2009, 40–43.
4. Herm Krabbenhoft, “Who Has the Major-League Record for the Longest Consecutive-Games Run-Produced (CGRP) Streak?” The Baseball Research Journal 38, no. 1 (Summer 2009): 125–134.
5. The runs-produced (RP) statistic should not be confused with the statis- tic estimated runs produced (ERP) devised by Paul As reported on pages in The Bill James Baseball Abstract (1985), Johnson’s ERP “is a method for estimating run production which is more accurate than even Bill James’ runs created formula” (276–81). The ERP formula is: ERP = 0.16 x {2 x [TB + BB + HB] + H + SB – [0.605 x (AB + CS + GIDP – H)]}.

6. The Baseball Almanac (www.baseball-almanac.com) states the following in its section “Career Leaders for Runs Produced”: “Runs produced is a SABERmetric statistic that describes a hitter’s overall effectiveness by measuring his ability to produce runs for (his) team either by scoring them himself or driving them in at the plate. Runs produced was created by baseball great Bill James during the 1970’s and the way it is calculated is adding runs to runs batted in [and] then subtracting home runs.” Likewise, in The Hidden Game of Baseball: A Revolutionary Approach to Baseball and Its Statistics (Garden City, N.Y.: Doubleday, 1984), John Thorn and Pete Palmer (with David Reuther) wrote the following: “Bill James, at about the same time [i.e., that Steve Mann introduced his run-productivity average in an unpublished 1977 manuscript] came up with a similar formula, since shunned, with values based on runs plus RBIs minus home runs” (64). However, in The Bill James Baseball Abstract (1984), James wrote that “there is another road toward the same truth [ascertaining a player’s contributions to offense, i.e., his runs created] that I would like to say something about. That is the statistic ‘Runs Produced’” (17–19). James concluded his discussion with the following statement: “Ah, well, I didn’t build the road” (i.e., invent runs produced). Then, three years later, James in The Bill James Baseball Abstract (1987) wrote: “Runs produced were invented by Spiro Agnew, an attempt to measure the same thing [as total average—i.e., to sum up the total effectiveness of an offensive player]. The ‘formula,’ of course, is runs + RBI – home runs (Spiro never was too complex)” (25).

7. In a subsequent (13 November 2008) search of the SI Vault on the Sports Illustrated website (www.Vault.sportsillustrated.cnn.com) for the term “runs produced” for the period 1954–present, I found that runs-produced charts were included in some post-1964 issues: 1965, once each month [(April 19 [the 1964 season RP rankings], May 3, June 7, July 12, August 16, September 13, and October 11); 1966, once every other month (April 25, June 20, and August 15); 1968 (August 19); 1976 (June 21, October 25, and November 29); 1982 (July 5); and 1999 (June 21). In none of the articles in which runs-produced statistics were presented (with or with- out RP charts) was any mention made of the creator of the RP statistic. In the “Scoreboard” (a collection of snippets on a variety of current top- ics, edited by Robert W. Creamer [e.g., Philadelphia Flyers Bobby Clarke’s thoughts on the NHL’s decision to crack down on fighting and related violence by introducing more stringent penalties]) in Sports Illustrated (21 June 1976) was the following statement about runs produced: “A baseball statistic called Runs Produced, which first appeared in Sports Illustrated 20 years ago, is based on the premise that runs are what count most in baseball. The figure is arrived at by adding the runs a player scores to the runs he bats in and then subtracting from that amount the number of home runs he hits. Players at or near the top in Runs Produced invariably are the ones who win ball games, those who get on base and score, those who drive other base runners in. For example, last year’s Runs Produced leaders were Joe Morgan of Cincinnati in the National League and Fred Lynn of Boston in the American. Not by coincidence, each was voted Most Valuable Player in his league, even though neither finished first in any of the so-called Triple Crown categories— batting average, home runs, runs batted in. If you’re wondering why the Reds are moving away from the pack, or why Texas and Kansas City are running one–two, here are this season’s top Run Producers in each league through games of last Friday.” The accompanying chart provided the following information (Player, Team, Runs Produced): National League—Griffey (CIN, 85), Morgan (CIN, 79), Perez (CIN, 75), Rose (CIN, 72), Schmidt (PHI, 72); American League—Mayberry (KC, 65), Otis (KC, 65), Burroughs (TEX, 63), Chambliss (NY, 62), Hargrove (TEX, 62), Hisle (MIN, 62). Later in the “Scoreboard” in SI (25 October 1976), Creamer reiterated the position that “while hitters who win batting titles and home-run championships get the publicity, the most valuable players tend to be the ones who are at or near the top in runs produced.” An accompanying chart provided the top ten in each league in runs produced—in the American League, Thurman Munson of the Yankees finished second in runs produced with 167 (Rod Carew of the Twins finished first with 178); in the National League, Joe Morgan of the Reds was first with 197 (with teammate Pete Rose second with 183). Creamer’s prognostication turned out to be on the money, as Munson and Morgan each later claimed the Most Valuable Player Award in his league. These two commentaries are apparently the only editorial texts on runs produced provided in Sports Illustrated. However, in a later issue of SI (29 November 1976), in the “19th Hole” (where readers expressed their thoughts about SI ’s treatment of a given topic), two people wrote to criticize runs produced. Archie Motley (Chicago) claimed that, in order to have a meaningful statistic, home runs should not be subtracted. And, Phil Tortora (Milford, Connecticut) opined that the “runs-produced theory does not take into consideration the player’s team”—i.e., a player on a good-hitting club will likely produce more runs than if he were on a poor- hitting team.

8. Similarly, an analogous online search of The Sporting News at paperofrecord.com for the period 1954–2003 showed that, while runs-produced statistics have appeared numerous times over the years since 1962 (particularly in the columns of Edgar Munzel, Peter Gammons, and Moss Klein), no indication of the inventor of runs produced was ever provided.