2010 Defense-Adjusted QB Performance (1 Viewer)

[Dinsy Ejotuz]

Using the info at PFR I started with a QB's Adjusted Yards/Attempt. This is (Passing Yards + 20 yards/TD - 45 yards/INT)/Attempts:

QB Age G AY/A Tom Brady 33 16 8.25 Philip Rivers 29 16 7.77 Aaron Rodgers 27 15 7.50 Ben Roethl 28 12 7.35 Vince Young 27 9 7.34 Michael Vick 30 12 7.29 Josh Freeman 22 16 6.94 Matt Schaub 29 16 6.74 Tony Romo 30 6 6.68 Matt Cassel 28 15 6.64 Peyton Manning 34 16 6.48 Kyle Orton 28 13 6.40 Joe Flacco 25 16 6.39 Matt Ryan 25 16 6.23 Eli Manning 29 16 6.09 Drew Brees 31 16 6.01 Jason Campbell 29 13 5.74 Matt Stafford 22 3 5.74 David Garrard 32 14 5.68 Ryan Fitz 28 13 5.68 Carson Palmer 31 16 5.54 Jay Cutler 27 15 5.50 Mark Sanchez 24 16 5.38 Donovan McNabb 34 13 5.33 Chad Henne 25 15 4.94 Matt Hasselbeck 35 14 4.86 Sam Bradford 23 16 4.73 Colt McCoy 23 8 4.73 Kevin Kolb 26 7 4.52Next I calculated each defense's AY/A against for the season:

Team	 AY/A GNB	 4.45 PIT	 4.69 CHI	 5.02 SDG	 5.09 BAL	 5.34 TAM	 5.41 ATL	 5.53 PHI	 5.70 CAR	 5.72 NWE	 5.75 KAN	 5.76 NYG	 5.81 STL	 5.91 NYJ	 5.93 TEN	 5.96 NOR	 5.99 MIN	 6.03 CIN	 6.17 ARI	 6.25 CLE	 6.30 IND	 6.40 SFO	 6.40 DET	 6.42 OAK	 6.53 MIA	 6.56 BUF	 6.63 DAL	 6.77 WAS	 6.95 SEA	 6.96 DEN	 7.67 JAX	 7.86 HOU	 7.96

Then I created an expected AY/A for each QB by taking a weighted average of his attempts vs the AY/A for each defense he played. So if Tom Brady threw 50% of his 2010 passes vs the Browns (AY/A = 6.30) and 50% vs the Eagles (AY/A = 5.70) his expected AY/A would be 6.00. This is similar to strength of schedule (lower = harder), except that it also accounts for the number of passes the QB threw against each defense:

QB	ExAY/AMatt Cassel 	 6.73 Peyton Manning 	 6.64 Philip Rivers 	 6.54 Vince Young 	 6.48 David Garrard 	 6.46 Eli Manning 	 6.34 Jason Campbell 	 6.34 Sam Bradford 	 6.29 Michael Vick 	 6.29 Matt Schaub 	 6.29 Matt Stafford 	 6.27 Tony Romo 	 6.24 Kevin Kolb 	 6.22 Kyle Orton 	 6.16 Mark Sanchez 	 6.11 Jay Cutler 	 6.09 Ben Roeth 	 6.07 Aaron Rodgers 	 6.05 Matt Hasselbeck	 6.04 Joe Flacco 	 6.03 Josh Freeman 	 5.96 Donovan McNabb 	 5.95 Chad Henne 	 5.90 Drew Brees 	 5.88 Ryan Fitz	 5.84 Tom Brady 	 5.80 Matt Ryan	 5.76 Carson Palmer 	 5.74 Colt McCoy 	 5.70

Finally, I took each QB's actual AY/A and subtracted his expected AY/A to get his net AY/A. This is a measure of the QB's performance adjusted for the defenses he played in 2010. It's probably more real-world based than fantasy based since it doesn't account for the number of attempts:

QB	 NAY/A Tom Brady 	 2.45 Aaron Rodgers 	 1.44 Ben Roethl 	 1.28 Philip Rivers 	 1.23 Michael Vick 	 0.99 Josh Freeman 	 0.99 Vince Young 	 0.86 Matt Ryan	 0.48 Matt Schaub 	 0.45 Tony Romo 	 0.43 Joe Flacco 	 0.35 Kyle Orton 	 0.24 Drew Brees 	 0.13 Matt Cassel 	 (0.09)Ryan Fitzpatrick (0.16)Peyton Manning 	 (0.16)Carson Palmer 	 (0.20)Eli Manning 	 (0.25)Matt Stafford 	 (0.53)Jay Cutler 	 (0.59)Jason Campbell 	 (0.60)Donovan McNabb 	 (0.62)Mark Sanchez 	 (0.72)David Garrard 	 (0.78)Chad Henne 	 (0.96)Colt McCoy 	 (0.97)Matt Hasselbeck  (1.18)Sam Bradford 	 (1.56)Kevin Kolb 	 (1.70)

A few initial thoughts/surprises...
[*]Tom Brady's 2010 really was other worldly

[*]Peyton Manning is starting to show wear

[*]Palmer was actually an average NFL QB (who faced a brutal schedule)

[*]That looks like an industrial sized fork sticking out of Matt Hasselbeck

[*]Cassel benefited from an easy schedule, but he also did with it what an average NFL QB would have

[*]Ryan Fitzpatrick was middle of the pack

[*]Josh Freeman - believe the hype

[*]Welcome to the NFL Colt McCoy (lowest expected AY/A)

[*]Sam Bradford covered his low per attempt number with a ton of attempts

[*]The team that trades for Kolb could really regret it (this jibes with some other research I've done too - I think he'll be terrible)

[*]Kyle Orton is the cream of the FA QB class

[*]Vince Young is really underrated, and some team will take a chance on him despite the headcasery

 

[floods]

This is some great stuff man, thanks for putting it together.

I'm especially interested in the Kolb info, though I don't own him in any leagues, it's assumed he'll step in and improve whatever team he's traded to.

 

[hook014]

The Kolb sample size has to be pretty small.

Not sure if there is enough to make a good prognosis.

 

[Catbird]

Really interesting view and perspective to measure from. The results, it seems to me, are seriously team and situation determined, but with most guys in that same basic situation this year, you give us a very useful and different view. What it doesn't reflect is that most of these guys will face much of their same tough or easy schedule again this year and so will tend to repeat their '10 results (rather than score in accordance with your listing). Palmer, if he comes back in Cinci, will still face the Steelers and Raven defenses twice each and most likely still put up far worse numbers than either Manning, who come out just above and just below Carson with the most similar ability results to him on your final calculation. Most of your conclusions seem well founded, if not somewhat over-stated because the defenses faced will tend to repeaT. It would also make a big difference to me to know how much one 90 yard TD pass would change a ranking, and how much your final list fluxuates year to year.

Not to hijack, but an unrelated (and much less analytic) thought your list brought to mind is that there are only 4 starting QBs listed at 33 years or older, and 2 of those are McNabb (34) and Hasselbeck (35) who are about done. Brady (33) and Peyton (34) are the others. We think of QBs like Favre playing at 40, Marino and Elway playing at 38, Warren Moon at 42 and think the good ones can play forever. But in the newer (still faster) era, maybe 34-35ish is the more common back end for QBs? That would concur with your findings on Peyton (although his sieve of an OL in '10 might have us both downgrading him a little early)? But its a reason for me to cool some on the dynasty value of both the oldersters (despite the historic reality that some QBs can just play much longer than most and your conclusion that Brady was the best QB in existence last season).

 

[Dinsy Ejotuz]

The Kolb sample size has to be pretty small.Not sure if there is enough to make a good prognosis.

He threw 189 passes. That's more than enough for this sort of thing (where you're adjusting to account for the Ds).On the other hand, you can argue that as a young and relatively inexperienced QB he'll still get better. But I don't buy it. Four years with Andy Reid and he basically swallowed a golf ball when given the reins.

 

[Dinsy Ejotuz]

Really interesting view and perspective to measure from. The results, it seems to me, are seriously team and situation determined, but with most guys in that same basic situation this year, you give us a very useful and different view. What it doesn't reflect is that most of these guys will face much of their same tough or easy schedule again this year and so will tend to repeat their '10 results (rather than score in accordance with your listing). Palmer, if he comes back in Cinci, will still face the Steelers and Raven defenses twice each and most likely still put up far worse numbers than either Manning, who come out just above and just below Carson with the most similar ability results to him on your final calculation. Most of your conclusions seem well founded, if not somewhat over-stated because the defenses faced will tend to repeaT. It would also make a big difference to me to know how much one 90 yard TD pass would change a ranking, and how much your final list fluxuates year to year.

The idea is to get an idea of the quality of QB independent of the Ds they face or attempt-driven FF results.And like I said it's more real NFL than fantasy, but I do think that it's more useful for figuring out which QBs are talented than just looking at FF numbers or unadjusted stats. Most of the time situations change enough that talent wins out in the end and it's easy to see some of the cases where the situation isn't likely to change soon (the Browns and Bengals have to face the Steelers and Ravens 2x each year for example).

 

[Instinctive]

Great info!

Some things we'll want to account for:

Sam Bradford's offense was designed to dink and dunk, play it safe, and use passes as run plays. This definitely increased his attempts and decreased his average per attempt.

Kyle Orton played in the McDaniels offense last year, which is known to be incredibly good. See 2007 Brady, 2008 Cassel...2010 Orton. We should remember it took a full year in that system before Orton took off with it, so:

Bradford should be very very good NEXT season. This season will be the learning curve season in the McDaniels offense. I wouldn't expect too much of him, but his jump next year could be just like Orton's last year.

 

[Dinsy Ejotuz]

I haven't looked at other years, but I suspect that these numbers are a lot more stable than FF points. Certainly looking at the rankings you get the sense that it does a nice job of sorting guys out.

The group that I don't think this is especially useful for is rookie QBs. Just from pulling up career records of a handful of good ones they can really suck and it means next to nothing. But if a young QB doesn't show a big jump in his 2nd year as a starter it's usually a red flag.