By In Baseball

Runs Created and MVPs

Many years ago, around the time that Bruce Springsteen was writing and recording Born to Run (spoiler alert), Bill James invented a simple but beautiful little formula to help determine how runs are created in baseball. He intuited from all his baseball reading and research that a run-scoring formula would need three components:

A: Getting on base.

B: Advancing on the bases.

C: Opportunities.

Makes sense, right? To score a run, you have to get to the plate. You have to get on base. You have to work around the bases. So, he came up with an equation that I think is sort of baseball’s E=MC2. In its most basic form, it looks like this:

[(H+W) * (TB)] / (AB + W)

That’s it. That’s the whole formula.

[(Hits plus Walks) times (Total Bases) / (At-bats plus walks)]

There’s so much goodness in this equation. For one thing, as you can tell by the plainness of it, Bill was working with what little he had. It is sort of like early human artwork done with rocks. This was in a time long before Baseball Reference and Fangraphs. Bill in those days was copying statistics out of The Sporting News and off the back of baseballs cards. In those days, it wasn’t even easy to get hit-by-pitch numbers. Caught-stealing numbers were guarded like the crown jewels. To get total bases, you would have to do some backward math, figuring out how many singles a hitter had by subtracting doubles, triples and home runs from the hit total. You had to work with what you were given.

And that was OK because even if you had all these other statistics, you couldn’t do much with them. There were no computers then to do complex equations. Bill was doing everything by hand. He came up with a simple formula because, well, he needed a simple formula.

But here’s the thing: The formula worked. I suspect it worked beyond Bill’s wildest hopes. Many people — Bill included — have tinkered with runs created, bent it, twisted it, weighed it down, lifted it up, stretched it and stepped on it. Some of these are great statistics. But just this original formula, which seems to leave so much out, still gives an astonishing estimate of how many runs will be scored. It is, in a word, magical.

How magical? Well, if you go back to 1950, the basic runs created formula has estimated that 1,126,591 runs would be scored. And, over those 65 years, teams have actually scored just 3,695 more runs than that. Think about that for a moment. Bill will tell you: He’s no mathematician. He’s no scientist. He’s not statistician. But through instinct, he came up with a basic formula that is 99.5% accurate.

If you go year by year,  the basic runs created formula is a bit more volatile. It has been as much as 3% off, which is still not bad. In other years, though it has been so accurate as to seem mystical. In 1994, for instance, the formula estimated that 15,753 runs would be scored. How many runs were scored? Well: 15,752 — one run less. There have been many years like that.

Why do I bring this up now? Well, I think it’s time we celebrate the runs created formula again. In many ways, it has faded into the background as more complicated and more thorough formulas have taken its place. But there’s something arresting and magnificent about runs created — as I found out when trying to do a little experiment. 

The experiment had nothing at all to do with runs created, at least in the beginning. I was tinkering around and trying to answer a question: How much has the popularization of Wins Above Replacement (both Baseball Reference and the Fangraphs version) changed the way MVP voters vote? People argue all the time about WAR, its value, its flaws. Some call it voodoo. Some swear by it. But no matter where people stand individually, I have a strong suspicion that collectively, WAR — that sort of one-stop shopping statistic that attempts to quantify a players ENTIRE contribution, including offense, defense and base running — has changed the landscape of MVP voting.

What I found out, I must admit, is not especially interesting or surprising. By breaking down MVP voting over the last 50 years in several unscientific ways, I found that WAR has changed the voting in some ways but not in others. Brilliant, right? What I mean is, I don’t see WAR leaders becoming MVPs any more often than the did before the statistic gained favor. From 1975-1984, for example, 11 Fangraphs WAR leaders won the MVP award — and that was obviously decades before the statistic was even invented. Over the last 10 years, nine Fangraphs WAR leaders have won the MVP award. So I don’t think the WAR impact has been that direct.

But where I think WAR has entirely changed the landscape is in getting rid of quirky MVP winners. Since 2008, which is just about when WAR and similar complex statistics started to become mainstream, every single MVP has finished Top 5 in WAR. In fact, every single winner except Miguel Cabrera finished first or second in WAR. Cabrera finished either third or fourth the two years he won, so he wasn’t exactly an outlier either. 

Before 2008, you would get very odd MVP choices every now and again. Here are only a few (the WAR I use here is the average between fWAR and bWAR):

2006: Justin Morneau, 21st in WAR.

2006: Ryan Howard, 9th

2002: Miguel Tejada, 14th

1998: Juan Gonzalez, 18th

1996: Juan Gonzalez, 31st

1995: Mo Vaughn, 13th

1987: Andre Dawson, 19th

1979: Don Baylor, 26th

1979: Willie Stargell, 34th

1976: Thurman Munson, 10th

1974: Jeff Burroughs, 24th

1974: Steve Garvey, 19th

1970: Boog Powell, 10th

Well, I tend to think those days are over — at least until a new statistic takes hold. Many people may dislike WAR and many may revolt against it. But realistically, I just don’t see how anyone who isn’t at least near the top of the WAR chart can build enough of a consensus to win the MVP award these days. I just don’t think it’s possible now.

Here’s an example: Mark Teixeira. As you might know, Tex is having a superb comeback season for a Yankees team that surprisingly leads the American League East. He’s among the league leaders in homers and slugging, he should finish with 100-plus RBIs, he is exactly the sort of player who, in the past, could gain all sorts of MVP momentum in September (especially if he hits a few key homers) and take away the award from someone perhaps more worthy overall.

But I think WAR more or less eliminates this possibility. Tex is 15th in WAR. Sure, he will get some MVP votes. I could even see him getting a first place vote or two (and I can also see the Web sites mocking the voters for the choice). But, in the end, I just don’t see him coming close to winning, not with Mike Trout and Josh Donaldson so far ahead in WAR.

So, I think that’s changed. I don’t think WAR is necessarily is picking more winners, but I do think it is eliminating more contenders. I think it’s narrowing the field. I just don’t think quirky and emotional choices like Willie Stargell will be winning MVP awards now. You can either celebrate or bemoan this, but I think it’s the new reality.

Like I say, I didn’t find all that particularly interesting or surprising. But I did find one surprise. In trying to answer the question, I looked at a bunch of different baseball statistics — including Fangraphs and Baseball Reference WAR — to see how often the MVP led the league in those stats. I went back 50 years. Now, wome of these statistics admittedly were kind of meaningless. For instance, seven MVPs since 1965 led the league in doubles (the last being Dustin Pedroia). Other than being surprised that Miggy didn’t lead the league in doubles in either of his MVP years (he led the league the year before and the year after), that stat offered nothing.

But ten of the stats provided what I think is sort of an interesting view of the MVP voting.

Stolen base leader: 2 MVP’s out of 92 (2%)

— The only two players since 1965 to win lead the league in stolen bases and win the MVP award are Ichiro in 2001 and Ricky Henderson in 1990. I do think that’s sort of interesting. Voters rarely take stolen bases into account when voting for the MVP, which I suspect is why Rickey only won the one MVP (he had a strong case four or five other times), why Tim Raines never won one (which, I think, has hurt his Hall of Fame argument) and why Kenny Lofton was pretty severely underrated. The last player to win the MVP bases almost entirely on stolen bases was Maury Wills in 1962.

Hit leader: 10 out of 92 (11%)

— The only player in the last 10 years to win the MVP as the hit leader was Dustin Pedroia. The year Ichiro set the hit record, he finished seventh in the voting.

Batting champion: 18 out of 92 (20%)

— Well, I did expect the batting champion to win more MVP awards, especially back in the 1960s and 1970s when batting average was the biggest stat going. Even if you go back to the 1930s and 1940s, though, the batting champion just doesn’t get named MVP very often. Pete Rose probably won his MVP in 1973 because he was the batting champ. And Willie McGee’s high batting average might have won him the award in 1985. But the batting champion-MVP double is rarer than I thought.

Runs leader: 27 out of 92 (28%)

— Over the last 15 years, more MVPs have led the league in runs (9) than RBIs (4). That’s kind of an interesting shift. Of course, both runs and RBIs are team-driven statistics, and it’s kind of silly to use either one for an individual award. But people have been doing that forever, and it’s nice to see runs scored get its due. Once again, if people had looked more closely at runs scored, Tim Raines would have won an MVP award and, I think, he’d rightfully be in the Hall of Fame right now.

Home run leader: 27 out of 92 (28%)

— In the last 50 years, 14 players have won the MVP award without leading the league in any major statistical category. It’s a compelling list that includes:

Roberto Clemente, 1966
Boog Powell, 1970
Steve Garvey, 1974
Thurman Munson, 1976
Willie Stargell, 1979
Kirk Gibson, 1988
Frank Thomas, 1993
Barry Larkin, 1996
Ken Caminitii, 1996
Juan Gonzalez, 1996
Ivan Rodriguez, 1999
Jeff Kent, 2000
Miguel Tejada, 2002
Justin Morneau, 2006

In addition, there have been seven MVPs who led the league in just one statistic.

Orlando Cepeda, 1967 (RBIs)
Jeff Burroughs, 1974 (RBIs)
Dale Murphy, 1982 (RBIs)
Robin Yount, 1989 (Runs Created)
Mo Vaughn, 1995 (RBIs)
Chipper Jones 1999 (Runs Created)
Andrew McCutchen, 2013 (fWAR)

I thought for sure that someone in the last 50 years won the MVP award based on hitting a lot of home runs. But it really isn’t so. Mark McGwire did not win the MVP award the year he hit 70. Cecil Fielder did not win the MVP award the year he broke the 50-homer barrier for the first time in a decade. Barry Bonds did win the award his 73-homer year, but he did so many other ridiculous things that year that I hardly think the homers won it it for him. I think the last time a player won the MVP award predominantly because of home runs was Roger Maris in 1961.

RBI leader: 34 out of 92 (37%)

— And now we come to the most controversial of MVP statistics. Between 1965 and 1998, more than half the MVP winners were also RBI leaders. This included some of the most bizarre choices in voting history — Mo Vaughn, Sammy Sosa, George Bell, Don Baylor, Jeff Burroughs and so on.

No can argue that RBIs cast a powerful spell over baseball fans. If baseball statistics have the power of language, as Bill James famously wrote, then RBIs shout. But, it’s silly. In 1998, when Juan Gonzalez led the league in RBIs, he came to the plate with more runners on than anybody in the league. Combine that with the fact that he played in a good hitters park and balls were flying everywhere that year and it would have taken a pretty amazing effort for Gonzalez to NOT have led the league in RBIs. Still he won an MVP award for it. In 1979, Don Baylor came to the plate with 40 more runners on base than any other player in baseball. And he was a good hitter. Of course he led the league in RBIs. It shouldn’t have made him MVP. But it did.

Anyway, as mentioned above, the RBI train has left the station. You can’t win the MVP award now based entirely on RBIs. Sure, there are still some who bemoan the RBIs’ fading light and there are some who will still give the statistic a lot of weight come MVP voting time. But there just aren’t too many RBI worshipers left.

Baseball Reference WAR leader: 35 out of 92 (38%)

Fangraphs WAR leader: 40 out of 92 (44%)

OK, now we get to the two WARs. I have gone back and forth about their being two distinct versions of WAR. For a while there, I hoped that they would come together a bit more — it seemed to me that it didn’t help the credibility of either version of the statistic to have such divergent results.

Now, I think differently — I think the two statistics should break apart even more. And I think they should have different names. I understand, as Tom Tango says, that they are just two different methods for the same framework (and technically they have different names, fWAR and bWAR). But I don’t completely buy it. I think they take two different approaches to valuing players, especially pitchers (bWAR looks at run prevention, fWAR look at strikeouts, walks and homers). They value defense somewhat differently. I now believe they should break apart, become two completely different statistics. One can be called WAR. The other can be called PEACE (Price Effective Above Common Earthling).

Hey, I think I’m going to do that. Baseball Reference WAR will still be called WAR. Fangraphs will be known as PEACE.

As you can see, the PEACE leader has won the award more often than the WAR leader. I’m not sure why — I think it comes down to defensive values. I’ll ask Tango his thoughts.

OPS leader: 41 out of 92 (45%)

Plenty of people will tell you that OPS is a junk stat, that it makes absolutely no sense to add together on-base percentage (which uses plate appearances as a denominator) and slugging percentage (which uses at-bats as a denominator). But for whatever reason, it has gained the power of language. I think it’s in part because OPS numbers very loosely match up with the grading scale at schools:

.900 and above: A

.800 to .899: B

.700-.799: C

.600 to .699: D

Below .600: F

OK, it’s not exactly like that. As we all know, some years teams score more runs than others, and league-wide OPS goes up and down. Around 2000, the league average OPS was .780 or so. Now it is much closer to .700. So the corresponding grades change. These days, an .850 or better OPS is probably an A, and on down.

This is why many people, myself included, find adjusted OPS+ to be so much more valuable than the raw numbers, because it adjusts the number based on ballpark, run scoring environment and so on. But you might be interested to know that the raw OPS leader has won the award seven more times since 1965 than the adjusted OPS+ leader.

And all this leads — big finish — to the whole point of this. The statistic that best predicts the MVP winner is … yep, Bill James’ little invention.

Runs created leader: 44 out of 92 (48%)

There is a lot that is amazing about this, but the main point is this: Nobody looks at runs created when voting for MVP. Well, maybe SOME people do, but I suspect its very few. That means that this stat has done a pretty amazing job picking the MVP award even though nobody refers to it. People just instinctively think along the lines of runs created.

As mentioned, the thing I love about runs created is that it is simple. There are more accurate ways to measure runs, but this is one you can do on a napkin. And it is such an elegant formula. Remember above, I wrote that Mark Teixeira will not win the MVP award this year? Well, runs created explains why in just a few quick calculations.

Who are the two top MVP candidates right now? Mike Trout and Josh Donaldson, probably.

Trout has 132 hits, 64 walks, 260 total bases in 444 at-bats.

Donaldson has 145 hits, 51 walks, 281 totals bases in 480 at-bats.

With a piece of scrap paper, I can quickly figure from that small bit of information that Trout has created 101 runs, and Donaldson has created 103. That’s really close. So now, you start thinking about their defense, their base running, their other contributions. It’s a great race.

And what of Teixeira?

He has 100 hits, 59 walks, 215 total bases in 389 at-bats.

Quick calculation: He has created 76 runs. And, while that’s good, it puts him 25-plus runs behind the other guys. How is he making up that difference? Speed? No. Defense? Not at first base. Leadership? Well, you’d have to give him a whole lot of leadership points.

Runs created really is a little bit of brilliance.

Print Friendly, PDF & Email

49 Responses to Runs Created and MVPs

  1. Doug says:

    Superb essay. I “knew” Runs Created was the perfect stat, but could no articulate why. Thanks for explaining — and documenting — how RC is such a great way of measuring offense.

  2. It is amazing how good this stat is. It even works in high school where we’ve used it to determine our MVP every year I have coached. Even at a less than professsional level, it is over 95% accurate.

  3. Jake says:

    Excuse me while I obsess over this stat and calculate it for the NL contenders. Great essay.

  4. Kendell says:

    I would suggest that Andre Dawson won his MVP award in 1987 based primarily on his HRs.

    • mjm says:

      RBI is what did it for the Hawk that year. That and “Leadership”, on a last place team.

      • Jay says:

        Y’all gotta remember that Dawson was blackballed by MLB in free agency in the winter of 1986. He signed with the Cubs either eight before or during spring training 87. This story that the writers could grab a hold of in addition to his chase of that magical *50* number was what won him the MVP. That, and that there were no run-away clear choices.

    • I was going to type the very same thing. I suspect that Joe might argue that Dawson’s RBI numbers had something to do with it, too.

  5. albrooke says:

    I think that OPS developed as a surrogate for the linear weights assessment of _The Hidden Game of Baseball_, which was originally published in the 1980s. Pete Palmer’s idea was to use linear regressions to figure the actual run value of every single offensive event (thus, I recall a single was — on average — worth .41 runs, but a walk was worth maybe .37 runs, for obvious reasons). A home run had a value of perhaps 1.4 runs (since men might be on base). The resulting linear weights (1.4*HR) + . . . + (.41*S) + (.37*BB) + . . . = a batter’s value in runs. Divide the number of runs by the number of extra runs that results in an extra win (about 10, most years) and you had the player’s offensive value in wins. (This was not WAR, though.)

    Linear weights regressions are (or at least were) the gold standard, but they had to be refigured all the time. Palmer noticed that LW was closely correlated to the simple sum of SLG and OBP, so voila! he had something that could be easily calculated from the pages of The Sporting News, and was a good guess for the “real” number.

    OPS was thus not a statistically justifiable number, it just happened to correlate to the statistically justifiable Linear Weights calculation.

    Or that’s what I remember from reading the book in 1990.

  6. Gary W says:

    Good stuff as always Joe. One thing you didn’t mention about those quirky MVP winners is their collective inability to field the ball. I think WAR (and PEACE) has done a great job of illustrating the importance of defense to voters, most of whom obviously took it for granted if all those one-dimensional boppers could take the trophy home.

  7. David Gardner says:

    Joe, not only was Bill’s development of Runs Created absolutely brilliant, but he also wrote a very compelling introduction to it in one of the abstracts. He pointed out that…since “runs” are the coin of the realm…a batter’s job is to do the things that help his team score runs, An example he used is that…when a team has a game facing a great pitcher…the manager doesn’t think “I hope we can go out and hit .250 against this guy today.” Instead, he thinks “Gosh, i hope we can score a couple of runs and give our pitcher a chance.” And he also used the example of three straight batters in a row hitting singles. The first batter gets credit for a “run scored”, the third batter gets credit for a “run batted in”, while the second batter doesn’t get credit for any of the “payoff” stats, despite the fact that his single helped his team…by moving the first runner into scoring position, and also putting himself on base…as much as either of the other two batters. And Runs Created gives players credit for things like that.

  8. I believe Branch Rickey developed a formula which was quite similar, though not identical, to runs created. He did this while with the Dodgers and had Alan Roth to crunch the numbers for him, but the formula was his and as I recall it definitely did take into account walks and total bases as key elements.

  9. invitro says:

    I bet if you just take walks out of the RC formula, the result will predict MVP’s better than RC does.

  10. invitro says:

    And the result of no players out of the top 5 in WAR winning MVP since 2008 is most definitely an interesting result, and an important one. Giving awards to people that deserve them, rather than to people that don’t deserve them, is real and true progress.

    • Brett Alan says:

      Agreed, and I would add that what Joe describes–that WAR isn’t determining the winners, but it’s narrowing down the contenders–is EXACTLY what it should be doing. If you believe that the guy with the highest WAR (either version) is ALWAYS the best player in the league, I think you’re giving that stat too much credit. But, with rare exceptions, if you think that the best player in the league is someone who is way behind the WAR leader, you’re almost certainly wrong. The stat isn’t perfect, but it’s not off by much. IMO.

  11. jim says:

    Ok, if I was the stat king of the world, I would highly value a stat I will term “bases created”. You get a base created for a walk, single, stolen base, or however many bases you advance by hitting the ball including if there is an error. You get 2 for a double, 3 for a triple, and 4 for a HR. You also get bases created as a result of the advancement around the bases of runners on in front of the hitter (those runners’ stolen bases do not count). So you hit a HR with a man on 1st, this gets you 7 bases created……or if there are runners on 1st and 2nd and you ground out to right side and both advance, you get 2 bases created…..This rewards all positive offensive outcomes like high batting average, walks, speed, rbis, and putting the ball in play while punishing Ks. Thoughts????

    • Jon Alan Schmidt says:

      I have thought about something like this for many years, but I would count both “bases gained” and “bases lost,” assigning the latter by some appropriate procedure to the players who make outs in an inning when others reach base but fail to score. Then we would end up with a total for each player that can simply be divided by four to get a version of runs created–and when aggregated, it will always add up to exactly the number of runs that are actually scored. The same methodology could also be used on the other side, for pitching and defense.

    • tangotiger says:

      You’ve basically described a simple version of RE24.

      You’ll enjoy this:

  12. Phaedrus says:

    Why are there only 92 MVP winners from 1965-2014? Seems like there should be 100, or 98 if no winners were declared in the strike year.

    Are you not including years in which a pitcher won MVP?

    • Doug says:

      There were MVPs in the strike year:Jeff Bagwell in the NL and Frank Thomas in the AL. Pitchers were MVPs in both leagues in 1968, in the AL in 1971, 1981, 1986, 1992, 2011 and in the NL in 2014. That would leave 92.

  13. Tom Geraghty says:

    It may make absolutely no sense to add together on-base percentage (OBP) and slugging percentage (SLG), but it does make sense to multiply them together, since essentially:

    Runs Created = OBP x SLG x AB

  14. bpdelia says:

    Joe to be fair war has 3 pretty glaring holes imo. One it returns consistently puzzling results for 1b defense and does not take scooping into account.
    2) it seems to undervalue top relievers.
    3) it absolutely fails to capture catcher defense in any meaningful way. And that HUGE when one considers that catcher defense is so important that teams will not trade offense for defense there at all.

    Doesn’t include pitch framing or game calling.

    Which is why berra and Munson and bench etc all have lower war than one would assume. The defensive component leaves out a bunch of value.

    So basically that was all a way of saying that by WAR Munson’s value would have been severely undervalued relative to guys at other positions where the defensive component is at least trying to capture their value. With catchers WAR has basically thrown up it’s hands.

  15. McKingford says:

    Got to give a shoutout to one of the strangest MVPs – and I say this as a Tigers fan: Willie Hernandez.

    In 1984, he was 4th on the team in bWAR, and was the 4th *pitcher* on the Tigers in terms of fWAR. But still the league MVP!

  16. Kubi says:

    “In 1994….there have been many years like that”

    Thankfully, there has been only one year like 1994…

  17. DJ MC says:

    Is it possible to make this work from a pitching perspective, where you could judge a pitcher based on fewest RC of the batters he faced?

  18. […] Story: On runs created and MVP awards. The Monkey Says: It correlates best to predicting the MVP, but I wonder if it is really something […]

  19. Jay Stevens says:

    One, the comparisons are a little unfair. You’re comparing how well isolated stats — HR, hits, SB, eg — against a combination of those stats — “runs scored” — as a measure of offensive performance. Well, yes, of course the latter is better! I’d bet if you just took total bases as a stat, you’d see similar results.

    Second, you’re asking the wrong question. Runs scored doesn’t better measure greatness — which seems to be the point of bringing WAR into the fray — it better accurately measures what criteria baseball writers use to make MVP votes. Writers — somewhat unsurprisingly — don’t take defense, park factors, or positional adjustments into account, which is why Miggy won two MVPs over Trout, although the latter was the more valuable player to his team.

  20. PhilM says:

    While I like the simplicity and correlation, I’m not sure this is much more than confirmation bias. If I look at Cy Young winners, the Wins leader matches 67% of the time, far better than bWAR (53%) or ERA+ (41%), for example. This doesn’t mean that Wins are an indicator of the best pitchers: it means that the voters vote for them. Yes, it’s “better” lately: but still 8 of the 14 Cy Young winners since 2008 (57%) were also the Wins leader.

  21. Cuban X Senators says:

    Love RC and have used it to draft many a sim league.

    One may only be able to appreciate it if one lived in the Sunday Sports section and a calculator era of stat crunching, but my favorite Bill James quick-figure stat was always: add R and RBI & compare that to H; the more the ratio tilts toward R + RBI the more that player is helping you in (then) subtle ways; if it tips toward H, that player is less valuable that is/was apparent.

  22. nickolai says:

    This post is another instant classic from Joe P. Thanks Joe

  23. JB in GP says:

    Interesting, but I wonder about the continual bashing of Juan Gonzalez. OPS of 1.011 in 96 yet an WAR of only 3.2? I watched almost every game that year and the man was a beast. I know this population hates RBI, but he had 144 of them and played in 134 games. Another hated thing – clutch – comes into play. With runners on, you knew he’d do the job. Plus look at how Juan Gone almost beat the eventual champion Yankees by himself in the playoffs. The new obsession with WAR dismisses a career like Juan’s – that is a shame.

    • Phil says:

      Gonzalez’s clutch stats from ’96 are kind of a mixed bag. With RISP, pretty close to his seasonal line (i.e., very good): .321/.424/.625. With runners on, some drop-off: .306/.378/.585. With two outs and RISP, and late and close (about 75 plate appearances for both), very poor. “High leverage,” not bad, not great: .282/.328/.518.

      You can look at what was behind all those RBI a bunch of different ways. But I don’t think there’s strong evidence that he was a great clutch hitter.

  24. I wonder how many wOBA, or wOBA x PA, leaders have won compared to RC leaders?

  25. Richard says:

    You know, there’s a good way to figure out how good a “stat” like OPS or RC really is. If you can create a stat like this for an individual player, you can use the same formula with the total numbers for a given team. Then you can use basic statistical methods (as built-in to any spreadsheet program worthy of the name) to determine which stat correlates best with the team’s winning percentage.

  26. Interesting analysis, Joe, and a fitting ode to Bill James’ Runs Created.

    The “goodness” and “simplicity” of Runs Created is still captured in the MVP voting. In fact, using the post-WAR era that you mentioned (2008-present), *every* position player MVP since 2008 (12 in all) also finished in the Top 5 in RC. Moreover, 7 of the 12 (58%) led their respective league in RC. In that same period, only 5 of 12 (42%) were bWAR leaders.

    Some other statistics of note regarding the 12 MVPs from 2008-2014:

    * Of the 5 bWAR-leading MVPs, three also led their respective league in RC. (The other two finished 3rd and 4th.)

    * Of the 4 RC-leading MVPs who were not bWAR leaders, none finished lower than 4th in bWAR.

    * The two bWAR leaders who did not finish in the RC Top 5 did not fare well in MVP voting: Carlos Gomez (13th in RC) finished 9th in the 2013 NL MVP race, and Ben Zobrist (8th) finished 8th in 2009.

    Clearly, there have been no total surprises from a WAR or RC perspective since 2008. In addition, a high-ranking WAR has not guaranteed a finish near the top of the MVP race.

    That leads to one glaring omission in your analysis: the importance of team success in MVP voting.

    Since 2008:

    * 11 of 12 (92%) MVPs played for a team that qualified for the post-season. (Pujols in 2008 is the lone exception. He led the NL in bWAR and RC; STL won 86 games, but finished in 4th.)

    * All 7 of the bWAR leaders who did not win the MVP played for a team that failed to qualify for the post-season.

    Voters like winners.

    Going back to your article, citing the importance of playing for a winning team would have shed more light on those “odd” MVP choices; providing only WAR rankings ignored relevant context. The list below shows each of those 13 MVPs, his team win total and post-season fate, followed by the bWAR leader, his team win total and post-season fate:

    YEAR LEAGUE: MVP (Team Wins, Post-Season?), bWAR Leader (Team Wins, Post-Season?)

    1970 AL: Boog Powell (108, Y) , Yastrzemski (87, N)
    1974 AL: Jeff Burroughs (84, N) , Carew (82, N)
    1974 NL: Steve Garvey (102, Y) , Schmidt (80, N)
    1976 AL: Thurman Munson (97, Y) , Nettles (97, Y)
    1979 AL: Don Baylor (88, Y) , Lynn (91, N)
    1979 NL: Willie Stargell (98, Y) , Winfield (68, N)
    1987 NL: Andre Dawson (76, N) , Gwynn (65, N)
    1995 AL: Mo Vaughn (86, Y) , Valentin (86, Y)
    1996 AL: Juan Gonzalez (90, Y) , Griffey (85, N)
    1998 AL: Juan Gonzalez (88, Y) , A. Rodriguez (76, N)
    2002 AL: Miguel Tejada (103, Y) , A. Rodriguez (72, N)
    2006 AL: Justin Morneau (96, Y) , Sizemore (78, N)
    2006 NL: Ryan Howard (85, N) , Pujols (83, Y)

    Again, winning seemed to matter most to the voters:

    * 10 of 13 (77%) MVPs played for a post-season qualifier.

    * All three MVPs who played for a non-qualifier were on teams that won more games than the bWAR leader.

    Some additional notes on RC and bWAR for that group:

    * 11 of the 13 “odd” MVPs lost the battle of WAR and RC. In fact, 9 of the 13 bWAR leaders finished 1st or 2nd in RC (5 finished 1st and 4 finished 2nd).

    * None of the “odd” MVPs finished in the Top 5 in WAR, and only four finished Top 5 in RC.

    The subtext to the RC equation is that there is a correlation between creating runs and improving your team’s chances to win. Regardless, some voters prefer to lounge on the lido deck of the SS Intangibles.

    The selection of Munson (teammate Nettles finished 16th in the 1976 AL MVP race) provides an example of voters putting more stock in leadership. In this case, the non-MVP teammate led the league in WAR and also finished higher in RC.

    Stargell in 79 (over teammate Dave Parker, who significantly bested Pops’ WAR and RC, but finished 10th in the MVP voting) was another example of leadership trumping stats; that MVP trophy was a de facto Lifetime Achievement Award.

    In this post-WAR era, easily accessible data and powerful new metrics (in addition to well-thought and clearly presented analysis of you and others…and the not-so-subtle arm-twisting of the Twitterverse, blogosphere, etc.) have enhanced the selection process. Still, voters prefer winners.

    With that in mind, the extended wild card format may be the most important development in the MVP landscape. Allowing additional teams into the post-season has expanded the list of “winners”. Consequently, the list of possible MVP candidates in the eyes of voters has grown.

    And the brilliance of runs created continues.


  27. Ortega says:

    I’m tired of people saying how much they hate the RBI statistic and I think it’s still a perfectly good stat to judge players by. Generally speaking, to create a run, you have to have multiple things happen, obviously. You need someone to get on base, which is why AVG is important. Stealing bases help get a player into scoring position, hence SBs are important. (Doubles and triples both get merit for essentially doing the same things as above.) And then what is needed to top it all off and bring a run in? An RBI! Wow!

    So the fundamental stats that have ruled baseball for a long time – AVG, 2Bs, 3Bs, SBs, RBIs, and of course HRs are still 100% relevant.

    Stop devaluing RBIs, Joe and Joe’s sycophants and stop trying to reinvent the wheel.

    • Marc Schneider says:

      But the point is, RBIs are heavily related to where the guy hits in the order and how good the team is. If you have enough runners on base, almost any decent hitter will have a lot of RBIs. But then there are great hitters that did not have an inordinate number of RBI. Mickey Mantle had over 100 RBI, I think only four times. So what does RBI actually tell you other than the guy came up with a lot of runners on base? In effect, the hitter has little control over RBIs. He does have over doubles, triples, and home runs.

      Not only that, but you don’t need to get a hit to get an RBI and the way you get an RBI can effect the team negatively. For example, suppose the team is down 3 runs in the 7th inning with runners on second and third, one out. Hitter hits a weak ground ball to second and the run scores. Hitter gets an RBI but he has, in effect, likely killed the rally. It’s a cheap RBI but does little for the team.

  28. Steve Gardner says:

    Great article Joe. Just one question: Why didn’t you calculate total bases by taking Hits + Doubles + 2* Triples + 3* Homers? I remember doing it that way when I had to calculate my dice baseball team’s stats.

  29. […] a recent article, Joe Posnanski made an interesting observation about WAR statistics and the MVP contest. He noted […]

  30. […] that estimates how much a player has contributed to their team’s run total. As Joe Posnanski writes, it is perhaps the best way to measure offensive success. Well, by RC, Posey has certainly been […]

  31. MikeN says:

    Mo Vaughn didn’t win because of his RBI. It was because he was a nice guy and people didn’t want to vote for Albert Belle.

  32. Carl says:

    Real question here. Does anyone remember Bill James, in one of the old Baseball Abstracts, defining Runs Created as R+RBI-HR?

  33. […] Posnanski recently argued that despite its detractors, WAR has “changed the landscape” to the extent that the […]

  34. […] überhaupt und es ist ein Jammer, dass sie nicht halb so bekannt ist wie z. B. RBI. Hier noch eine Leseempfehlung, die die Aussagekraft von RC […]

Leave a Reply

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