beyond ERA for free-agent pitchers (NY Times)

http://www.nytimes.com/2004/11/21/sports/baseball/21score.html?oref=login

Earned Run Average? Some Haven't Earned It and Are Only Average
By ALAN SCHWARZ

The recent major league general managers meetings in Key Biscayne, Fla., ended up being the equivalent of most Barry Bonds at-bats: enticing home run potential, but ultimately little to connect with. No big trades, no big signings, no specifics on them whatsoever. But there was plenty of subtle gamesmanship as team officials assessed this winter's crowded class of free-agent pitchers.

"There's not a No. 1 starter out there on the free-agent market, in my opinion," Yankees General Manager Brian Cashman said. "There's a lot of guys who their agents will tell me are No. 1's, but there aren't any out there."

As general managers opined that the available arms are all No. 2 and No. 3 types, with little to distinguish one from another, the splish-splash sound in the Ritz-Carlton lobby was not of the nearby Atlantic surf, but of the pitching market flooding. If every pitcher is essentially the same, then demand for each can go down, and the price to acquire them can go along with it.

Sure enough, the group of starters available this off-season is rather indiscriminate - most are around 30 years old and have generally posted decent earned run averages (3.60 to 4.00) and records (14-10, give or take a few decisions).

Brad Radke, Russ Ortiz and Odalis Perez are certainly not ace material, and Pedro Martínez, for whom the Yankees are no longer a daddy but perhaps a magnanimous uncle, doesn't rank as a No. 1 anymore because of his arm's ever-questionable endurance.

With all teams looking to add a starter or two this winter, they will value pitchers by their projected effectiveness - essentially, their future E.R.A.'s. Yet E.R.A. does a poorer job of presaging success than do several other statistics, and a general manager must consider this before lavishing a three- or four-year contract on a pitcher.

"Kind of like R.B.I.'s, it's subject to a lot of circumstantial factors," Nate Silver, a writer for Baseball Prospectus, said of E.R.A. Silver's research proves that a pitcher's strikeout and walk frequencies tell far more about his future E.R.A. Next come home run and hit rates, with won-lost records all but irrelevant.

That knowledge helps us organize this year's free-agent pitchers more discerningly.

Although Atlanta's Ortiz (36-16) leads a group of nine attractive in-their-prime pitchers who started at least 50 games the past two seasons, warning signs still exist. His rate of 6.3 strikeouts per nine innings is quite average, and his rate of 4.6 walks is alarmingly high. His 3.97 E.R.A. at pitcher-friendly Turner Field was already somewhat hefty; his new employer should not expect the effectiveness he has delivered Atlanta.

It is also easy to see why Mets General Manager Omar Minaya has taken his time in re-signing Kris Benson. Given his low strikeout and high hit rates - not to mention injury history - paying top dollar long term would be an immense risk.

The sleeper of the group is clearly the Cubs' Matt Clement, whose electric stuff has yet to translate into performance. Although he walks more than many prefer, his rate of 8.5 strikeouts and 7.6 hits allowed the last two years are by wide margins the best among his free-agent competition. It is these numbers that have some executives wondering if he will be the best investment of the group.

In many ways, Clement resembles Jason Schmidt, who entered the 2001 free-agent market as an enigmatic right-hander. He signed with the Giants and has blossomed into an ace.

"The guy I like is Clement," said one National League general manager, who withheld his identity and limited his remarks for fear of driving up the price. "We're buying the future, not the past."

Another hidden gem could be the Dodgers' Perez, whose paltry 19 victories the past two years could discourage some teams. Yet he received the N.L.'s lowest run support last season (3.3 runs a game), has slightly better under-the-radar numbers than Ortiz and, at 27, is three years younger.

Speaking of run support, Baseball Prospectus developed a method it calls Support Neutral Win-Loss, whereby a starting pitcher's outing receives not a full victory or a loss, as box scores have recorded for a century, but a slice of each depending on his full pitching line, the support he received, the park he was pitching in and more.

The aggregate numbers show how Derek Lowe's 31-19 record for the Red Sox tells more about the team's offense than about his pitching. He "deserved" to be 19-29. They show how a high victory total for Ortiz should be viewed with suspicion. They also illustrate the effectiveness of Minnesota's Radke and Florida's Carl Pavano, whose low strikeout rates and exceptional control have made them the market's safest bets.

But from Clement's whiffs to Ortiz's walks, each pitcher carries statistical pros and cons, leaving this market quite jumbled. With such little difference among the pitchers, the trick may be to wait for the best bargain; acquiring Perez for $5 million could very well be better than signing Radke for $8 million.

Each club will look at the statistics differently, the traditional organizations beguiled by won-lost records and E.R.A.'s, while the more progressive ones consult more esoteric percentages. Clubs won't identify their allegiances yet. Their money will talk soon enough.

Copyright 2004 The New York Times Company