In the Atlanta chapter of PFP 2006, I ran similarity scores for Michael Vick's 2005 season and showed that the two most similar seasons since 1978 belonged to Steve Young (1986, his first year in the NFL) and Steve McNair (1997, his first year as a starter). Certainly Vick's 2005 numbers did not suggest that he would never develop as a passer. But the top 10 list of similar seasons also featured players like Kordell Stewart, Dave Brown, and Charlie Batch, players with similar skill sets who never developed (although Batch has turned into a nice backup).

So here we are, halfway through 2006. We've gone through a roller-coaster ride where the conventional wisdom on Vick changes roughly every two or three weeks:

• This new shotgun option is going to help Vick succeed, even as a run-first quarterback.
• No, never mind, he's never going to be any good.
• Wait, he's coming around -- two big games!
• No, never mind, he's never going to be any good.

Add the good games to the bad ones and Vick is suffering yet another terrible year as a passer. His completion percentage is 52.6 percent, which would be the worst of any of his full seasons. He's gaining just 6.3 net yards per pass, only slightly higher than the 6.2 from last year. He's on pace for a career high in touchdowns, but also in interceptions.

I pro-rated Vick's stats over the whole year and ran similarities again to see whether McNair and Young were still on the list. Assuming Vick is going to actually stay healthy for 16 games is always a bit of a stretch, but he's doing well so far, so we'll give him 16 games in the projections.

Comparing just a single season, the most similar player is Randall Cunningham in 1989, followed by Trent Dilfer in 1998 and Steve McNair in 1997. Yes, that's the same McNair season that was similar to Vick last year, but that makes sense because the fifth most similar season to Vick in 2006 is Vick in 2005. Also on the list are Kordell Stewart twice (1997 and 1998), Cunningham again (1990), Cliff Stoudt (1983), Shaun King (2000), and Timm Rosenbach (1990).

Extending the list to 13 gets really strange, because 11-13 includes Vick himself in 2004 and 2002 and Ron Jaworski in 1978. If I could name the quarterback that Vick was least likely to turn into, it would be Jaws -- can you see Vick debating Sal Palantonio Jr. and Mike Alstott on NFL Matchup in 2034? But Jaws in 1978 had passing numbers similar to what Vick is going this year -- he shows up below the top ten because the running numbers are so different. (Part of this is that 1978 was still a transitional year for the new passing rules, so the quarterback numbers from that season are all over the place.)

Anyway, what really matters about this list of similar players is how dissimilar they are. For those unfamiliar with similarity scores, they work on a scale where 1000 is the same player, and the lower the score, the less similar. The top ten for Vick's 2005 that I ran in the book ranged from 762 (David Carr in 2005) to 855 (Young in 1986). This top ten ranges from 660 to 745.

At this point, there is really nobody to compare Vick with, at least not in NFL history since 1978. There has never been a quarterback this inaccurate, this far into his career. Those few quarterbacks who come close were not runners, and of course, Vick runs with the ball more than any quarterback since Bobby Douglass.

Three-year similarity scores, which we usually use to help project careers, are also fairly useless here. The top player is Dave Brown 1994-1996, but the similarity score there is 685. Then comes Trent Dilfer, Randall Cunningham, Donovan McNabb, and Steve McNair. McNabb and McNair are similar to Vick if, by similar, you mean "numbers with similar ratios but dramatically superior in every passing category."

The one player who keeps showing up is Cunningham, but Cunningham's lower completion percentage makes some sense because he was throwing for more yards and more yards per attempt than Vick. You can live with a lower completion percentage if you throw more long passes.

The moral of the story, I guess, is that Vick is sui generis. There's never been anyone like him. Last year, there were historical parallels of players who developed, and this year, there aren't. Certainly that doesn't decide the fate of Vick's career, but it's a point in the argument that he's just never going to "get it." At what point is it just too late for him to develop into an actual NFL passer?