In that pivotal sixth game of the 2002 World Series, the Angels trailed the Giants 5–3 coming to bat in the bottom of the eighth. With Tim Worrell relieving, Darrin Erstad homered and Tim Salmon singled. Garrett Anderson moved Chone Figgins (running for Salmon) to third with a single, taking second on Barry Bonds's bobble in left. Dusty Baker then called on his closer, Robb Nen, who promptly allowed a base-clearing double by Troy Glaus before getting out of the inning. The other Troy (Percival) tossed a perfect ninth and, stunningly, Anaheim won 6–5, forcing a seventh game. The pitching line: Worrell—three runs (two earned), the loss, and a game ERA of 54.00 (he got the last out in the seventh); Nen—0.00 ERA, two of two inherited runners scored, plus a blown save.

But why can't ERA reflect Nen's responsibility for allowing the runners to score as well as Worrell's for putting them on? Proportional Run Allocation (PRA) does exactly that. By allocating .25 of a run per base advanced, each pitcher's liability is identified. Worrell allowed baserunners to reach third (.75) and second (.50), so in addition to the homer, he's tagged with 2.25 runs (1.75 earned) and the loss; Nen let runners score from third (.25) and second (.50), or 0.75 runs, 0.25 earned. Factoring in PRA, the game ERAs become Worrell, 47.25, and Nen, 0.25—numbers more representative of each man's outing.

PRA works in all situations, affecting not only ERAs but, at times, won-lost records. Imagine Figgins only reaching second on the Anderson single and both runners still scoring on Glaus's double. Worrell would be charged with the runners who reached second (.50) and first (.25), a total (with the Erstad blast) of 1.75. Nen would be hit with 1.25 for the portion of the runs scoring from second (.50) and first (.75) and the loss, since the theoretical score attributable to Worrell would be Giants 5, Angels 4.75. Their game ERA? Worrell, still 47.25, but Nen, 11.25.

PRA impacts ERA to varying degrees, inversely proportional to the number of innings worked. Starters, with many innings and no inherited runners, experience slight decreases. Middle relievers and specialists, who put men on base and allow inherited runners to score, move up or down, depending on their overall effectiveness. Closers undergo moderate to significant increases. The ten or so in each league who head the statistical listings slip into the pack with ERAs more indicative of their skills.

Prior to 1969, all pitchers were rated by ERA and won-lost only. To denote aspects of relieving not evident in those numbers, the save category was added, eventually to be supplemented with its illegitimate descendants, blown saves and inherited runners scored. But those peripheral facts, much like the infamous asterisk assigned to Roger Maris's sixty-one home runs, function merely as explanations, modifiers. By utilizing PRA, these additional factors can be incorporated into pitching's two most quantitative statistics, thus eliminating the need for additional subcategories of measurement. Now all pitchers can again be compared equally by using just ERA and won-lost totals.

It's time to fold PRA into baseball's statistical mix. The fans, the scorers, the computers, and maybe even the pitchers can handle it.

—EFQ

• In the Batter's Box • Bring Us Home • On the Newsstand • Sample an Issue •
Submit a story • Tell a Friend • Advertise with us • Our First at bat • Privacy Statement

© 1999 - 2006 Elysian Fields Quarterly Web Master Dahlke Designs