Great stuff. The Fox contributions are great.

What in the devil was McCown's PAR? SF's pass defense was 53.9% as of last week. Even at an adjustment strength of 40%, that's still insane.

In Branch's defense, at least two of his incompletions were deflected. On one, it looked like he wasn't even the intended receiver.

Thing I like best about the new Fox column: no more sharing time with idiots like Skip Bayless from ESPN. All Aaron all the time. And giving a little explanation of each guy is sweet, like knowing that Jabar Gaffney caught 6 passes all for first downs and why Steven Jackson was actually pretty bad.

Thing I like least about the new Fox column: only get to see top and bottom 5 DPAR performances. That's all I really care about for WRs and RBs, but for QBs it's really nice to see how they all stacked up against each other.

I was forced with an ugly decision for a bye week fill-in for my fantasy team - Tim Rattay or Josh McCown?

I think you know who I chose. Thanks for the -2 points, loser.

I was forced with an ugly decision for a bye week fill-in for my fantasy team - Tim Rattay or Josh McCown?

I think you know who I chose. Thanks for the -2 points, loser.

Can someone explain this to me? In the article, McNabb's game was omitted from consideration for top 5 QB of the week on account of the Chiefs' poor pass defense. But McCown was rated as one of the top 5 even though he faced the 49ers pass defense. Now I thought the 49ers pass defense is pretty lousy and that the Chiefs' pass defense is at least better than the 49ers so I checked. As of 9/27/05, The Chiefs' pass defense has a value of 17.2% and the 49ers' has one of 53.9%. Doesn't that mean that between the two, McNabb's game should be considered and McCown's game be disregarded?

The article futhers added, confusingly, in the explanation :
"Players receive bonuses when they play well against good defenses, and they don't get rated as world-beaters when they shred the 49ers."

Doesn't that contradict the placement of McCown in the top 5 QB of the week??

Can someone explain this to me? In the article, McNabb's game was omitted from consideration for top 5 QB of the week on account of the Chiefs' poor pass defense. But McCown was rated as one of the top 5 even though he faced the 49ers pass defense. Now I thought the 49ers pass defense is pretty lousy and that the Chiefs' pass defense is at least better than the 49ers so I checked. As of 9/27/05, The Chiefs' pass defense has a value of 17.2% and the 49ers' has one of 53.9%. Doesn't that mean that between the two, McNabb's game should be considered and McCown's game be disregarded?

The article futhers added, confusingly, in the explanation :
"Players receive bonuses when they play well against good defenses, and they don't get rated as world-beaters when they shred the 49ers."

Doesn't that contradict the placement of McCown in the top 5 QB of the week??

Am I just a dork, or does anybody else here wish they had DPAR numbers for their Madden franchises? Sure, E. Palazzo and B. Dickus are putting up all kinds of gaudy numbers, but how good are they really.

Mike:

That's why I asked for McCown's PAR number. It's going to be insanely high.

Keep in mind that opponent adjustments are only at 40% strength, because KC could actually be worse than we think (probably not that likely) and SF could be better than we think (really not that likely).

If SF continues to stink up the joint, if Aaron would redo this several weeks from now, McCown's DPAR would be much lower. For now, though, it's smart to give them the benefit of the doubt.

Anyway, if you just ignore opponent adjustments, McCown did have a better game than McNabb. McNabb threw an interception, and McCown didn't.

well, it's good to know it just wasn't me screwing up and double posting.

Josh McCown: 16.6 PAR before adjustment, 12.4 DPAR after.

Donovan McNabb: 12.8 PAR before adjustment, 9.4 DPAR after.

And while we're at it:

Peyton Manning: 17.1 PAR before adjustment, 15.6 DPAR after.

Huh. That's not as high as I would've thought it.

So here's the question - did Philly attack a portion of KC's defense where they've been particularly vulnerable all year? The aggregate pass DVOA for SF is something like 4 times worse than KC's, yet the correction doesn't seem to be as large as you'd expect.

McNabb spent a lot of time in KC's territory in the latter half of the game - have they been really bad in the red zone?

Aaron,

How about including the PAR in your chart for everyone? It sounds like it might save you some work!

Pat, Thanks for the explanation. As you can see, I haven't gotten the grasp of the stats completely. If 49ers defense is really good or that McCown is really that good, I better grab him fast as I need a reliable backup. After seeing Griese's bottom 5 ranking, I don't have as much confidence in him as I did.

Matt, it is you. I clicked the refresh button more than necessary :).

Something worth remembering in regards to PAR vs DPAR for McCown (and others) -- the defense adjustment is only at 40% strength right now. Go back and score this game in a couple months and McCown will rate lower.

Also Aaron,

What about creating a fantasy league based on DPAR? I've never liked the idea of fantasy football because of the way the stats are used. Maybe and innovation for the FOX site?

What about creating a fantasy league based on DPAR? I’ve never liked the idea of fantasy football because of the way the stats are used. Maybe and innovation for the FOX site?

That would be awesome.

Me, Sid, and some other guys whose FO handles I don't know (Dave Brown Fan Club?) made a Yahoo league where we tried to make fantasy value a little more dependent on real value. For example, QB's get penalized for incompletions and sacks, running backs have a penalty for each carry (set so that players who average more than 4.0 yds/carry come out positive), and more of the value of defenses is tied up in their points allowed. But having one based on DPAR would be way cooler.

Also, if you haven't yet, check out the PROTRADE site. They have "dividends" based on something similar to DPAR.

In general I do believe in the statistics you've put together Aaron. But I simply do not believe that Donovan McNabb going 33/48 (69% completion percentage) for 369 yards, 3 touchdowns, and 1 interception should not rank among the best 5 games by a QB for the week. And why Matt Hasselbeck, who passed for 242 yards on 26/39 (67%) and 1 touchdown, is ranked higher is just beyond me. I know you talk about difficult down-and-distance situations, but it's not McNabb's fault that he wasn't stuck with a lot of third and longs. The Eagles rushed for all of 28 yards! Talk about an off day...and that was on 17 carries. Shaun Alexander still had 98 yards on 20 carries! I know your numbers don't have an agenda, and that you put them out there exactly as they appear. But in this case your numbers stink. Plain and simple. McNabb (and Owens) were the Eagles' offense.

3. Bobby Engram SEA
A big day for this Football Outsiders favorite. Back in 2003, Engram had more value per play than any other receiver in the league, and the idea that he couldn't succeed as a starter was absurd.

Just thought I'd add that he played almost the entire game with a rib injury he took on a hit in the first play.

but it’s not McNabb’s fault that he wasn’t stuck with a lot of third and longs

Excuse me, please let us Seattle fans have our one bright moment for the week :p

On that note, even with the two losses this year, I feel like we're a stronger team this year than we were last. Our losses are frustrating but have a lot of positives in them.

Two strange points:

1 - The Redskins had *two* penalties called against them, and they were both for delay of game. A lot of my fellow fans on the ESPN boards were livid to the point that they wouldn't quit complaining about "the Baltimore game", "the Engram tripping", and "the Vinny 'Football Head' Testaverde" game.

2 - The Seahawks (3 takeaways, -3 T/O diff.) trail only Washington (2, -4) and Green Bay (1, -8) in fewest Opposing Turnovers. Seattle's lost 4 fumbles, and only given up two interceptions, though, and that's a pretty good sign.

Also, Washington went 13-18 on 3rd down, and anecdotally I would say a large majority of those conversions were for long yardage. I'm sure TMQ will rip us for that.

You forgot a couple of numbers. Hasselbeck: 2 carries, 16 yards, one for 10 yards on fourth-and-1. McNabb: 1 carry, 2 yards, fumble. Thus,

Hasselbeck: 9.4 pass + 0.8 rush = 10.2 DPAR.
McNabb: 11.5 pass - 2.2 rush = 9.3 DPAR.

Count me in on the "I'd like to see all the QBs" bandwagon and let me compliment the column in general.

but it’s not McNabb’s fault that he wasn’t stuck with a lot of third and longs.

It also wasn't Brian Westbrook's fault he wasn't handed the ball on every down. Should we credit him with the rushing title simply because he could've had the best performance of the day?

Defenses play different on third and long than they do on second and short. Someone who performs as well against a better defensive situation deserves more credit.

A hypothetical, but how many DPAR points would Marcus Pollard be worth if the TD had been good?

but it’s not McNabb’s fault that he wasn’t stuck with a lot of third and longs.

Actually, you could say it IS McNabb's fault. Because he completed passes on 1st and 2nd instead of getting sacked.

Which raises an interesting point. Which is more valuable, PAR-wise:

1st & 10 incomplete pass
2nd & 10 incomplete pass
3rd & 10 10-yd pass

OR

1st & 10 3-yd pass
2nd & 7 3-yd pass
3rd & 4 4-yd pass
(or something along these lines)

Or is that something that doesn't have a simple answer, but depends on field position, game time, etc...

To generalize the question:

Which is more valuable, several failures followed by a large success OR several small successes (which add up to large success)?

I would guess the second is better, because the offense is more consistent.

Re: #18 -- "Dave Brown Fan Club" is me. I don't post here that often, so I figured my screen name wouldn't be recognized anyway.

I think the league is pretty cool. We've got it set up so that QBs get rewarded for having above average Completion Percentages and Yards Per Attempt (and are penalized for being inefficient in those categories) -- ditto WRs and RBs with regard to Yards Per Catch/Carry. A few other bells and whistles, too.

It's pretty complicated. No point in going into it now, but you can bet I'll post a thesis next year when I set up another league (or two) like it.

DavidH: A successful play is usually one that nets 40% of yardage needed for a first down on 1st down, or 60% on second down or 100% on third down, so in your example the first two completed passes would be considered failures. Of course gaining a little yardage is better than getting nothing. Also, since DVOA compares to league averages in down and distance, I think 3 yards on 1st and 10 would be considered below average. If I had to choose, I'd take two incompletions and a 10 yard gain, because that offense seems to have a better chance of getting a big play. Now, if they were rushing downs instead of passing downs, I'd rather have the constient small gains, but in the modern NFL you throw the ball to get a lead and run to maintain it.

The other thing to remember is what I mentioned earlier. Defenses play different on 3rd and 10 than they are on 1st and 10. This is why using "league-average production" is such a good idea. So getting 12 yards on 3rd and 10 could actually be harder than getting 12 yards on 1st and 10.

Re: A fantasy league. Yeah, that would be cool. A couple thoughts on this theoretical league:

Not DPAR, but just PAR. Part of fantasy football is picking who is going to be better based on their opponent.

Second, not necessarily PAR either. Instead of subtracting the value of a replacement player, just leave it in. So PAR becomes Points. Each player score is how many points (according to the calculations they contributed to the team's win). To do it properly it would have to be on a play-by-play basis like DPAR is, instead of just accumulated stats.

Oh, and with their QB and both RBs on the bottom list, is it safe to say that San Francisco's offense had a bad day?

This is interesting. Assume the following trend for two quaterbacks.

QB1:
1-10: InComplete.
2-10: Complete > 10

QB2:
1-10: Incomplete
2-10: Incomplete
3-10: Complete for 10 yards.

Who is going to be ranked higher?

The fact that Eagles threw over 75% of time and McNaab still didnt get into many 3rd and long situations, should be considered as an achivement. It means he was consistantly getting yards in 1st and 2nd downs.

#31
That's the point. Sorry, Aaron, for going overboard on my last post. But the emphasis on 3rd down achievement in PAR does diminish the performance of players/teams who do consistently well on 1st and 2nd down. The Eagles in that game had no running attack to speak of, and virtually every point they put up resulted, in some part, because of McNabb's contributions on offense. He wasn't faced with 3rd and longs because he completed passes on first and second down for positive yardage.

Guys, McNabb's passing was worth a TD and a half. That's pretty damn good. Remember, if he simply hadn't fumbled he be on the top 5 list, but he did. He also had an interception which probably knocked his PAR down a couple of points as well.

What do you want McNabb to get for not screwing up on 1st and 2nd down? 20 pts above replacement? Throwing for 4 yards on first, second, and third down is an only somewhat above average performance on a play by play basis. The fact that McNabb managed to convert steady but unspectacular play-by-play results into a large number of points on the board probably has a lot to do with the playcalling and the defense, and that gets reflected in the numbers not being ridiculously high for McNabb.

I guess I don't see an 11.5 passing DPAR as undervaluing McNabb's game.

should be considered as an achivement.

It is. It's not like DPAR says "McNabb sucked!" 9.3 DPAR scaled to a season puts McNabb second only to Manning last year. It's phenomenal. Keep that in perspective.

The difference is that it's harder to get a first down on 3rd and 10 than it is on 1st and 10 because the defense plays tighter, and you also have no room for error. What if it isn't Hasselbeck's fault that he got into 3rd and long situations either (it isn't - Alexander wasn't running well)?

If it was Hasselbeck who put them in 3rd and 10, then things would even out (because 1st and 2nd down would be negative plays for him). But a QB gaining 10 yards on 3rd and 10 is more impressive than gaining 10 yards on 1st and 10 because you can not risk failure on 3rd and 10. You can on 1st and 10.