Darrell Evans was a Better Hitter than Wade Boggs

The title of this post seems pretty ridiculous, doesn't it? How in the world could Evans even be close to Boggs offensively? Just look at their triple slash lines:


Boggs: .328/.415/.443
Evans: .248/.361/.431

The numbers aren't even close, and they came in virtually the same number of plate appearances (10740 vs. 10737). In terms of OPS, it's Boggs .858 to .792. Even accounting for the fact that Boggs played at Fenway through 1992, and in the Steroid Era from then on, he still leads Evans in OPS+ 131 to 119.

These numbers, however, are quite misleading. BA on its own is misleading because it treats all hits the same and discounts walks and HBPs. OBP does account for walks and HBPs, but still doesn't weight for hits, SLG does weight for hits, but doesn't factor walks and HBPs. Even OPS and OPS+ are actually inaccurate statistics, as they tend to favor players with higher batting averages. Since both OBP and SLG account for hits, OPS and OPS+ actually double count a player's hits. In essence, each of Boggs's 3010 hits are counted twice, as are Evans's 2223.

If we broke down the components that go into BA, OBP and SLG (and thus OPS), here's what we get:

Boggs: 3010 H, 1412 BB, 23 HBP, 4064 TB, 10740 PA, 9180 AB, 29 SH
Evans: 2223 H, 1605 BB, 35 HBP, 3866 TB, 10737 PA, 8973 AB, 34 SH

Let's create a new formula that doesn't double count the hits. Since TB is the weighted version of hits, we can use that and throw out the hits totals. In addition, AB and PA are mostly redundant, so well just use PA and subtract off SH to get our denominator. Therefore:

(TB + BB + HBP) / (PA-SH)

This is in essence OBP, substituting TB for H. This seems pretty similar to wOBA, except wOBA weights for run values. Here, we are weighting for bases. Using this formula for Boggs and Evans yields this result:

Boggs: .513
Evans: .514

Evans now rates as a tick better. These numbers, however, are not context adjusted. Let's see what  a league average player would have done in the same ballparks:

Boggs's context: .268/.336/.413
Evans's context: .263/.330/.392

Boggs obviously played in a higher run scoring context. Using these numbers we can see how much above average both he and Evans were. To start, lets calculate our "improved OBP" for the contexts. The above formula can easily be retrofitted for rate stats as such:

(OBP*(PA-SH)-BA*AB+SLG*AB)/(PA-SH)

In this case we'll use Boggs's and Evans's numbers for PA, AB, and SH.

Boggs's context: .460
Evans's context: .438

We can then compare the result with their actual and place on a 100-is-average scale.

Boggs: 112
Evans: 117

In essence, Evans was 17% better than league average, while Boggs was only 12% better.