Skip to content

45% hitting, 25% fielding, 25% pitching, and 100% not telling us how they did it

A University of Delaware press release reports:

This month, the Journal of Quantitative Analysis in Sports will feature the article “An Estimate of How Hitting, Pitching, Fielding, and Base-stealing Impact Team Winning Percentages in Baseball.” In it, University of Delaware Prof. Charles Pavitt of the Department of Communication defines the perfect “formula” for Major League Baseball (MLB) teams to use to build the ultimate winning team.

Pavitt found hitting accounts for more than 45 percent of teams’ winning records, fielding for 25 percent and pitching for 25 percent. And that the impact of stolen bases is greatly overestimated.

He crunched hitting, pitching, fielding and base-stealing records for every MLB team over a 48-year period from 1951-1998 with a method no other researcher has used in this area. In statistical parlance, he used a conceptual decomposition of offense and defense into its component parts and then analyzed recombinations of the parts in intuitively meaningful ways.

The good news is that the numbers add up to less than 100% (I assume the remaining 5% can be attributed to baserunning, strategy, and teamwork—those are the only other variable factors I can think of that could influence winning). The bad news is that the press release does not link to the article or to any technical report.

So I have no idea whether to take Pavitt’s claim seriously at all. I think this sort of press release is just silly: the claim is empty without the accompanying analysis.

I don’t have high hopes, though, given that the author appears to be analyzing “team winning percentages” rather than runs scored and runs allowed. As Bill James has pointed out, runs scored and allowed are more directly related to offense and defense, and you’re pretty much just throwing away information by looking at winning percentages. It’s hard to know more, though, given that we have no link to the article and I can’t find anything with that title on the web.


  1. Joe Levy says:

    I wonder if anything is attributed to plain luck.

  2. John says:

    I should hope it adds up to less than 100 even with those other factors accounted for… unless you think there’s no random chance in baseball.

  3. @Joe and John: In most models, there’d be an estimate of the probability of winning a game between two teams under some conditions that’s somewhere between 0 and 1, with non-boundary estimates allowing for some unexplained randomness in the outcome. Often, there’s a “noise” parameter in the model itself and sometimes models can be reformulated in terms of noise parameters (see the Gelman and Hill regression book for different formulations of logistic regression for examples).

  4. Eric Tassone says:

    But don’t all players on all teams give 110%? Anyway, here’s a link to Bill James himself saying baseball is 42% hitting, 8% baserunning, 37% pitching, and 13 percent fielding: Worth a skim!

  5. That formula is obviously wrong. We found out a long time ago that 90 percent of this game is half mental.

  6. Alex Cook says:

    JQAS really need to learn from the other top journals and introduce a press embargo until the paper appears on-line.

  7. Michael Humphreys says:

    The statistical model of defense (Defensive Regression Analysis, or DRA) used and fully disclosed in my book Wizardr: Baseball’s All-Time Greatest Fielders revealed, indicates that the ratio of the standard deviation in runs saved/lost attributable to defensive events under the control of pitchers to that of fielders is about 65:35. Since the standard deviation in runs scored by teams is about the same as the standard deviation in runs allowed, that has led me to conclude that baseball is about 50% offense, 33% pitching, and 17% fielding, which is close to how Bill James allocated things in his book Wins Shares (2002).

    Now since good fielders are generally weaker hitters, there is probably a negative correlation between team offense and team fielding, which could cause the net combined impact of a team’s non-pitchers to fall relative to pitchers. Tom Tango has found that teams spend something north of 40% of their payroll on pitchers, and concluded they are generally correct in doing so.