By Aaron Schatz
Special to Page 2

Note: Go to the bottom of the table for a more complete explanation of how Aaron's QB rating system works.

Here's a typical e-mail I've received:

"I am a Giants fan, please tell me that Eli will be alright! Do you think that Coughlin even knew who was coming on the schedule when he decided to play Manning?"

Eli Manning
Manning played better in Week 15, but has had many reasons to hang his head so far.

Saturday's game probably quieted talk about Eli Manning as a bust for the time being, but there's no doubt his rookie year hasn't been what Giants fans were expecting. Even after a good game, his numbers look nothing like his brother's rookie year, or what people expect from the No. 1 overall pick. Here are Eli's stats, prorated to include two more starts:

81 for 186 (43.6 percent completion rate), 977 yards, 4 TD, 10 INT

Yep, that's bad. How bad? Eight quarterbacks since 1978 have had a completion rate below 50 percent and at least twice as many interceptions in the first season of their careers with at least 125 pass attempts. It isn't exactly the roll call of victory: Steve DeBerg (1978 49ers), Jeff Komlo (1979 Lions), Art Schlichter (1984 Colts), Bruce Mathison (1985 Bills), Mike Tomczak (1986 Bears), Jack Trudeau (1986 Colts), Browning Nagle (1992 Jets), and -- oh, no -- Ryan Leaf (1998 Chargers).

Those quarterbacks, however, don't necessarily make the best comparison for Manning because they put up terrible numbers in more games. Saturday's game is reason to believe Manning will improve with more playing time, and that his completion rate and touchdown-to-interception ratio would be better if he had played from Week 1.

A good way to compare Manning to other rookies who played only part of the year is with similarity scores. Similarity scores are a Bill James invention familiar to baseball fans, but I have a similar formula that I use to compare football seasons. (A couple of other football writers use variations on the idea as well. You can read more about the specifics in this article about Drew Brees. I took the prorated Manning numbers above and compared them to every quarterback since 1978, the year the modern passing rules went into effect and the season expanded to 16 games.

The new list of similar QBs looks a little better than the list above, but it isn't tremendously better. It is hard to find quarterbacks who had a completion rate as bad as Manning's and threw for as few yards per attempt (and only played a few games). Instead you get guys who were somewhat close in each area. The most similar quarterback is actually another No. 1 overall pick who has had a very long career -- but not kind of career that Giants fans are hoping to see out of Manning. The rest of the top 10 are a mix of guys who had reasonable but unspectacular careers, guys who really went nowhere, and guys who are famous for attacking Jim Rome. Here is the top 10, along with two players who would be in the top 10 except that the formula also includes rushing yardage:

Rookie QBs Most Similar to Eli Manning
Player Year Team G Comp Att Yards TD INT Pct Yd/Att
Manning proj. 2004 NYG 8 81 186 977 4 10 43.6 5.25
Vinny Testaverde 1987 TB 6 71 165 1085 5 6 43.0 6.55
Don Majkowski 1987 GB 7 55 127 875 5 3 43.3 6.89
Eric Zeier 1995 CLE 7 82 161 864 4 9 50.9 5.37
Billy Joe Tolliver 1989 SD 5 89 185 1097 5 8 48.1 5.93
Steve Walsh 19898 DAL 8 110 219 1371 5 9 50.2 6.26
Koy Detmer 1998 PHI 8 97 181 1011 5 5 53.6 5.59
Tim Hasselbeck 2003 WAS 7 95 177 1012 5 7 53.7 5.72
Cody Carlson 1988 HOU 6 52 112 775 4 6 46.4 6.92
Danny Wuerffel 1997 NO 7 42 91 518 4 8 46.2 5.69
Jim Everett 1986 Rams 6 73 147 1018 8 8 49.7 6.93
Michael Vick 2001 ATL 8 50 113 785 2 3 44.2 6.95
Quincy Carter 2001 DAL 9 90 176 1072 5 7 51.1 6.09


Putting Michael Vick aside for the moment, this is not an encouraging list for Giants fans. The two first-round picks, Testaverde and Everett, had long and successful careers, but only three combined Pro Bowl appearances. The other guys never really put it together, except for Don Majkowski for a couple years, and a number of them did have high expectations. Remember that Steve Walsh was a supplemental draftee who cost Dallas a first-rounder, Billy Joe Tolliver and Quincy Carter were second-rounders, and a couple more of these guys were third-rounders.

Obviously, a look at similar rookies doesn't mean Vinny Testaverde represents Manning's limit at this point. All this struggle could be the fault of the offensive line, or the receivers, instead of Manning. This method doesn't include any adjustment for strength of schedule, and Manning has faced a gauntlet of this year's top pass defenses. And he's probably going to get better over his last two games. But even when I looked at similar rookies assuming that Manning's final two games would be like his game against Pittsburgh instead of the ones that came before it, the list of similar quarterbacks is pretty much the same, with a couple of unexciting additions like Craig Whelihan and Chad Hutchinson.

Actually, Manning's good game on Saturday got rid of the most successful of the players with similar rookie seasons. When I did this based on Manning's first five games instead of six, the most similar rookie was eventual Super Bowl MVP Doug Williams, who had a horrific 37.6 percent completion percentage his first season. After a few years in Tampa, Williams had to switch leagues and then switch to a new NFL team before he won a Super Bowl.

What about Eli's big brother? In Peyton's rookie year he threw more interceptions than touchdowns and completed only 56.7 percent of his passes for 6.5 yards per attempt, both numbers far lower than any of his other seasons. But Peyton's rookie season was so unique that there is only one similar rookie since 1978 (Jim Kelly).

This week's rankings:

  • Get overall season rankings at

    Quarterback Skinny DPAR
    1. Chad Pennington
    18/24, 253 yards
    3 TDs, 0 INTd
    Since it began the season in spectacular fashion, quickly collapsed, and has now redefined pathetic, I hereby nickname Seattle's D the "XFL defense." 17.0
    2. Kerry Collins
    21/37, 371 yards
    5 TDs, 1 INT
    Despite injuries and free-agent defections, Tennessee's D was playing reasonably well against the pass until it imploded three weeks ago. 14.1
    3. Daunte Culpepper
    25/35, 404 yards
    3 TDs, 1 INT
    I think Randy's hammy feels better now. 8-for-10 on third down with 234 yards and 3 TDs. 12.5
    4. Peyton Manning
    20/33, 249 yards
    1 TD, 0 INTs
    Second straight week where lack of TD passes does not indicate poor play. 12.0
    5. Eli Manning
    16/23, 182 yards
    2 TDs, 1 INT
    Flailed his arms around before the snap, changing plays and motioning guys over to pick up blitzers. Yep, he's a Manning. 11.1
    6. Josh McCown
    22/34, 287 yards
    2 TDs, 0 INTs
    At this point, is anyone not rooting for Arizona to win NFC West at 7-9? 10.5
    7. Trent Green
    16/19, 224 yards
    3 TDs, 0 INTs
    Rookie Sammie Parker, who had never caught an NFL pass, caught three for 84 yards and a TD. 9.3
    8. Billy Volek
    40/60, 492 yards
    4 TDs, 1 INT
    This game counts for 17.5 PAR, the highest of the week, but Oakland's secondary has been so bad that adjusting for opponent cuts the value by more than half. 8.1
    9. Ben Roethlisberger
    18/28, 316 yards
    1 TD, 2 INTs
    Very good game despite the INTs. Am I the only one who thinks the first one was partially Lee Mays' fault for being too passive? 7.8
    10. Patrick Ramsey
    18/27, 214 yards
    1 TD, 0 INTs
    If a quarterback has a good game on a Saturday, and nobody watches because both teams are terrible, does it make a sound? 6.5
    11. Matt Hasselbeck
    22/30, 201 yards
    2 TDs, 1 INT
    Don't blame him for terrible defense or Alexander fumble. Best field position to start drive was own 31-yard line. 6.0
    12. Aaron Brooks
    14/21, 169 yards
    2 TDs, 0 INTs
    Did absolutely nothing for much of the game while getting battered by the TB pass rush, then found a way to throw two perfect TD passes to win the game in the final minutes. 5.9
    13. David Carr
    13/28, 220 yards
    1 TD, 0 INTs
    Anyone else out there see name "J.Gaffney" in box score and think of jerk captain from "Homicide: Life on the Street"? 5.6
    14. Jake Delhomme
    24/35, 340 yards
    2 TDs, 1 INT
    A hot quarterback, 1:30 on the clock with a timeout, and only 40 yards needed for FG position. How did Carolina screw this up? Oh, right, a backwards toss play and a false start. Can't blame Jake for that. 5.3
    15. Jamie Martin
    16/31, 188 yards
    0 TDs, 0 INTs
    People criticized me for having Marc Bulger as a top 10 quarterback this year, but who doubts it now? 5.2
    16. Drew Bledsoe
    15/30, 183 yards
    1 TD, 0 INTs
    Buffalo defense also shoveled his walkway, cooked him dinner, and babysat his kids. 3.3
    17. Drew Brees
    4/6, 85 yards
    1 TD, 0 INTs
    Missed this game with flu, replaced with lifelike robot, the Acme Handoff-o-matic 3000. 1.6
    18. Michael Vick
    11/28, 154 yards
    1 TDs, 2 INTs
    4.2 DPAR rushing, -3.5 DPAR passing. So much for "Vick passes better when he runs well." I know we've been blaming the offensive coordinator but Vick himself has turned into Nuke LaLoosh. The WCO doesn't make him overthrow his receivers by five yards, and it doesn't make him throw a ball that bounces 10 yards in front of Peerless Price. Try breathing through your eyelids, Mike. 0.7
    19. Byron Leftwich
    9/20, 121 yards
    2 TDs, 0 INTs
    Gets up from more hits than Monty Python's Black Knight. 0.4
    20. Vinny Testaverde
    16/28, 176 yards
    1 TD, 2 INTs
    At least can take solace in fact that today I compared him to member of Manning family. -0.6
    21. Donovan McNabb
    20/35, 223 yards
    1 TD, 2 INTs
    Is this what the season would have been like without T.O.? Pitiful 2-for-8 on third downs with two sacks and two interceptions. -1.5
    22. Joey Harrington
    25/44, 361 yards
    2 TDs, 2 INTs
    Value significantly higher before adjusting for Vikings defense. Not his fault Lions can't snap the ball. -1.7
    23. Brett Favre
    30/44, 367 yards
    2 TDs, 3 INTs
    As a reader pointed out this weekend, complaints about annoying non-stop Favre worship by the announcers are themselves now becoming annoying and non-stop. -2.0
    24. Kyle Boller
    19/40, 210 yards
    1 TD, 2 INTs
    While Manning got the press, real battle of this game was to see who was for real, Indy defense or Boller. Boller did not win this competition. -4.1
    25. Brian Griese
    13/22, 118 yards
    1 TD, 1 INT
    A week after setting season low for passing yards allowed, Saints defense sets season low again. -4.2
    26. Jon Kitna
    16/32, 151 yards
    1 TD, 2 INTs
    Free Jon Kitna! Oh, whoops ... -5.1
    27. Jake Plummer
    23/41, 292 yards
    1 TD, 2 INTs
    Leads league in passes listed with no intended receiver (14). Explains a lot, doesn't it? -5.2
    28. Ken Dorsey
    20/38, 206 yards
    2 TDs, 4 INTs
    Come on, 49ers. Play Cody Pickett next week. What do you have to lose? -7.0
    29. Chris Chandler
    1/6, 1 yard
    0 TDs, 1 INT
    Mike Martz pinned the blame for Sunday's loss on Chandler, continuing his disturbing habit of taking absolutely no responsibility for anything that goes wrong with the Rams. "It's tragic for this football team, for that position to hold this whole football team hostage," said Martz. Gee, Mike, who makes roster decisions like naming a backup quarterback? -7.0
    30. Luke McCown
    11/27, 108 yards
    0 TDs, 1 INT
    The Browns could have played 40 quarters on Sunday and he was never going to get them into the end zone. -7.6
    31. Chad Hutchinson
    17/34, 168 yards
    0 TDs, 1 INT
    Followed up promising first appearance with total loss of accuracy -- just like his St. Louis minor-league days. -10.8

    How DPAR (Defense-adjusted Points Above Replacement) works

    The success of each play is judged based on yardage gained towards both a touchdown and a first down. Then each play gets compared to the NFL average on similar plays, based on down, distance, and other variables. Quarterbacks are judged not based on how many yards they get, but on how important those yards are in the context of the game.

    Ratings are also adjusted for the quality of the opposing defense. The quarterback's performance is then translated into an approximate number of actual points that such success (or failure) is worth when compared to a "replacement level" quarterback (defined as any quarterback named "Billy Joe").

    When all offensive, defensive, and special teams plays are added together for one team, the result comes very close to the actual difference between points scored and allowed.

    Among the advantages of this system:

    1. Gives value for first downs, which are not really included in any other QB rating system but are hugely important.

    2. Does not punish quarterbacks who are always in bad field position because of a poor defense, nor does it punish quarterbacks who are always stuck in third-and-long because of a poor running game.

    3. With enough data to begin including defensive adjustment, quarterbacks receive bonuses when they play well against good defenses, and they don't get rated as world-beaters when they shred the 49ers

    4. Includes both passing and rushing plays, which obviously helps a QB like Michael Vick.

    5. DPAR punishes quarterbacks for turnovers but also for fumbles that his own team recovers. Different kind of fumbles have different penalties depending on how often defense recovers for a turnover. Sacks are punished as well.

    6. 5-yard scramble on 3rd-and-10? Worthless!

    7. Actual points! Easy to understand!

    An even longer explanation of these numbers can be found here.

    Aaron Schatz is editor-in-chief of


  • Aaron