Column 2 3 4 5 or 6

Sorry, Andre, but "The Numbers" Don't Lie

By Kenneth Broder
Los Angeles Herald Examiner
April 3, 1988

In this year's Sporting News Baseball Yearbook, sports writer Dave van Dyck calls Andre Dawson's 1987 MVP season a "monstrous" year.

Former teammate Tim Raines, in describing why the Chicago Cubs outfielder was able to so impressively dominate the Most Valuable Player balloting in the National League, said, "The numbers that he put up this year were able to just overwhelm the reporters who vote."

Now, Tim Raines has an excuse for spouting such nonsense; he's a good friend of Andre. "Friends of Distinction," the yearbook calls them. They go back half a dozen years together -- back even before Andre blew his knees out and had to move from centerfield to right and ease up on the Montreal artificial turf play. But what's behind Dave van Dyck's blathering?

Apparently it's "The Numbers." You can't deny them. Forty-nine homers and 137 RBI. Andre led the league in both categories, two legs of the fabled Triple Crown. Van Dyck was impressed. Raines was impressed. Most of the baseball world was impressed.

But I wonder if the seven National League hitters who actually had better years with the bat than Andre were impressed. These were guys who couldn't match Andre's homer and RBI totals, but dwarfed his puny totals of 32 walks, 24 doubles, 11 stolen bases and .287 average.

But how do you prove who's better?

Take Tony Gwynn, for instance. Dawson hit more homers, Gwynn walked more. Dawson drove in more runs. Gwynn hit more doubles. Dawson had 16 game-winning RBI. Gwynn hit for a higher average. You get the picture. Too many numbers to compare. Which stats count more? Who's to say?

Well, the Society for American Baseball Research (SABR), for one -- 5,000 or so sports fanatics who seem to take a particularly twisted delight in trashing nearly every baseball folk myth held dearly by us.

The list of heresies is lengthy, although it varies depending on the heretic you talk to. They dismiss runs scored and RBI as virtually meaningless, practically ignore a player's fielding percentage, worship the base on balls, downplay the stolen base, ridicule the hit-and-run and look askance at -- heaven help us -- the batting average.

The most famous of these so-called SABRmetricians is Bill James, author of the "Baseball Abstract," high priest to a growing flock of statistics freaks and inventor of the Runs Created formula. It is upon that mathematical rock that James has built his unholy church.

James, and others like him, take a whole range of statistics like homers, doubles, walks, etc., and distill them down to one number. A picture painted with a single stroke. A story told with a single word.

This is how he does it:

Runs Created = (Hits + BB - CS) x (Total bases + .7 x SB) / (At Bats + BB + CS)

The result is a single number that tells you how many runs a player contributed to his team. Not how many he actually drove in, or how many he scored. Those statistics say as much about whom else is hitting in the lineup as it does about the man at the plate. Runs Created deals only with how many runs the hitter theoretically contributed to either getting on base or advancing a runner already there. Scoring runs is, after all, what the game of baseball is about.

This is, admittedly, pretty abstract stuff. If we don't see a runner score, how do we know he has? How can we tell the formula is accurately describing what James says it's describing? How do we know that a certain combination of hits, walks, homers, etc. will produce a certain number of runs?

You obviously can't tell looking at individuals, but you can get a pretty good idea if the formula works by applying it to team and league totals where you know just how many runs were scored. Theoretically, if it works on teams and leagues it should work on players, with a bit less accuracy because of a smaller sample of statistics.

At least, Bill James thinks so. I'm not a mathematician, but I do have a calculator and this is what it shows:

by RCRunsAccuracy

American League10,98811,11298.9%
National League8,7978,77199.7%

Looks pretty good. I can live with a 90 percent-plus accuracy. Anyway, it confirms a hunch I've always had about whining, gimpy Andre Dawson and his puny walk totals. He's an overrated ballplayer. Here's a vote for the actual MVP -- Tony Gwynn.
These were probably the real Top Ten hitters in the American and National leagues, or at least the most productive over an entire season. Next week we'll look at why Andre Dawson is even worse than these statistics indicate.

American LeagueNational League

1.Wade Boggs148.8Tony Gwynn143.5
2.Alan Trammell137.2Dale Murphy136.3
3.George Bell129.5Tim Raines130.6
4.Mark McGwire126.6Daryl Strawberry123.4
5.Dwight Evans126.1Eric Davis121.6
6.Danny Tartabull123.2Pedro Guerrero120.4
7.Paul Molitor120.7Jack Clark114.3
8.Kirby Puckett119.5Andre Dawson113.8
9.Don Mattingly119.0Mike Schmidt111.5
10.Kevin Seitzer119.0Juan Samuel108.0