Mike Tomczak better than Jim McMahon?

Yes, at least at avoiding sacks. The latest issue of the NFL's GameDay magazine features an article by Steve Hirdt which offers a new system for measuring quarterback performance: rate of sacks per dropback.Hirdt's point is that in looking at attempts, completions, yards, touchdowns and interceptions - the factors used to compute NFL quarterbacks' passing ratings - an important element is often overlooked: the quarterback's ability to avoid being tackled behind the line of scrimmage.

For example, suppose one signal-caller completes five of his first 10 passes for 80 yards, but on his team's next possession is sacked twice, each time for a 10-yard loss. A second quarterback, meanwhile, also completes five of 10 passes for 80 yards, but on his team's next possession throws the ball out-of-bounds twice to avoid being sacked.

Under the NFL's system, the second quarterback would be rated lower because of a lower completion rate. But Hirdt suggests that the second quarterback would have helped his team more by avoiding a loss of yardage.

Hirdt offers the Chicago Bears' McMahon and Tomczak for comparison. Over the past two seasons, he notes, McMahon has been sacked about once every 15 times he drops back to pass. Tomczak's sack rate, on the other hand, is 1 in 26.

And the two top-rated quarterbacks in Hirdt's list are Dan Marino and Doug Williams, which seems to lend it some credibility.

But whoa.

The problem with this nifty little stat, like all other stats, is that it can only be considered as part of the whole picture.

Because after Marino and Williams, you start getting into some weird results. Like Tomczak third among NFL quarterbacks. Mark Malone seventh. Phil Simms 41st. And Randall Cunningham dead last.

Would Hirdt seriously start Tomczak over McMahon? Or over Simms or Cunningham?

McMahon's passing rating in 1987 was 87.4. Tomczak's was 62.0. That seems enough of a disparity in McMahon's favor to overcome a few more sacks.

Besides, Hirdt's system doesn't take into account the fact that some quarterbacks hold the ball longer looking for an open receiver, rather than flinging it out of bounds the first time they glimpse an opponent's on-rushing jersey. And those extra few seconds several times a game could mean several extra touchdowns over the course of a season. And a couple more wins.

Like most "new" stats, Hirdt's are flashy but only useful in context.