Writers of Pro Football Prospectus 2008

11 Dec 2006

2006 Quick Reads: Week 14

This week's best and worst performances by Football Outsiders stats. Yes, Mike Karney makes the top five running backs. Sorry to post late, but I was unable to get on the administration part of the site because of the comment spam attack we've been dealing with since last night. We're still trying to fix it.

Posted by: Aaron Schatz on 11 Dec 2006

56 comments, Last at 14 Dec 2006, 7:45pm by Jason Mulgrew aka The Mul Dawg aka J-Rocka


by throughthelookingglass (not verified) :: Mon, 12/11/2006 - 10:23pm

First! Karnage!

by Jeff (not verified) :: Mon, 12/11/2006 - 10:23pm


by Moridin (not verified) :: Mon, 12/11/2006 - 10:41pm

I can't believe Brad Johnson was basically a replacement level passer against Detroit. I figured Replacement level against that D was like 250+ yards.

Too bad it apparently was 'just enough' to keep his job for another week.

by hector (not verified) :: Mon, 12/11/2006 - 10:47pm

Edge 2005 above Sanders 1997? I'm sure the math jives but that doesn't seem right to me.

I think Miami fans feel bad about this Brees thing. But who knew? His shoulder must have been jelly when the Fish looked at it.

by hector (not verified) :: Mon, 12/11/2006 - 10:52pm

As for Losman's deep ball, wasn't Evans five yards clear of the DBs? (I know, I know, his pump fake too.) I'd like to think most of the 32 in the league could make that throw. I've also seen plenty of times this year where Evans had plenty of separation and Losman *didn't* make the accurate toss. I'm not sold on JP yet in any way, be it to win the Quarterback Challenge someday or to be a winning NFL QB.

by hector (not verified) :: Mon, 12/11/2006 - 10:58pm

I'm impressed with Hackett as I think most of us are, but I'm always curious if some No. 3 or No. 4 wideouts catch a high percentage of passes because QBs are less likely to force a ball to a non-primary target. If they're clearly open, chuck it, but it's not like Hasselbeck is throwing jump-balls to D.J. Hackett, right? Nonetheless, this guy can play.

by Will Allen (not verified) :: Mon, 12/11/2006 - 11:21pm

Ignoring time required for people to get adjusted to each other or a new system, I think it quite likely that the Giants would be substantially improved with Favre instead of Eli at QB, and the Packers would have at best two wins with Eli taking snaps. In fact, I think there are even fewer qbs than Aaron suggests who would have gotten as much out of the Packers' personnel as Favre has, and I say that as somebody ho gets as tired of Favreophila as anybody.

As longs as the guy has a good coach who isn't intimidated by him, Favre is well above average. He can still buy time in the pocket, and he can still make the difficult throws better than nearly anybody. Yes, his decision making is faulty at times, but surround him with good players and coaching, and I guarantee you he would easily still be one of the top five qbs in the league.

by Peter (not verified) :: Mon, 12/11/2006 - 11:23pm

Oh boy, DPAR and I have some issues.

by Jeremiah (not verified) :: Mon, 12/11/2006 - 11:26pm

I'm very surprised at Manning's DPAR, as I would have figured it to be far lower even with adjustment. How would it have looked if his receivers could actually do what they're paid to do and catch the ball?

Brees's number looks relatively small by comparison, considering the way he just shredded Dallas. Isn't their pass defense supposed to be pretty good too?

by DrewTS (not verified) :: Tue, 12/12/2006 - 12:49am

I love the Colts, but I simply can't believe that Peyton was the #3 QB this week. I spent the entire game watching the entire team shite the bed and what I thought was Peyton spraying passes, when the receivers weren't busy dropping hits to the hands. Maybe I'm just spoiled (see Bronco fans re: Elway/Plummer).

by db (not verified) :: Tue, 12/12/2006 - 1:06am

Trent Green has now had several bad games, including last week when he was arted as the #3 QB performance DPAR wise. Last week he had a good game right up until the Browns tied the game. We can all agree that the KC defense let down the offense but that kind of thing happens. With the game on the line Green gave up a sack and lost a fumble which insured OT. With possession in OT, he went 1 for 3, forcing a punt and allowing the Browns the opportunity to steal the game. Green never has been a clutch performer. He always finds a way to choke and he is without doubt one of the most over rated QBs in the league. Any chance KC had of making the playoffs went to the bench with Huard who is vastly superior to Green in all aspects of the game. The lack of a clutch rating in the QB stats renders them all but meaningless.

by Crushinator (not verified) :: Tue, 12/12/2006 - 1:13am

DPAR rewards multiple plays, so Peyton's high number of passes against a league-leading defense probably inflated the number.

I didn't see the game, but I saw the scores happening and the Colts had one of those scores in garbage time. It could also be a result of the Jags going prevent early

by RecoveringPackerFan (not verified) :: Tue, 12/12/2006 - 1:30am

Last year proved something obvious about Favre: he needs coaching and enough of a team to keep the game close to be anywhere close to good. I would still take Favre with a lead or in a close game over the vast majority of QBs; once down two scores, the inexcusably stupid throws replace the merely dangerous ones.

by Dagagad (not verified) :: Tue, 12/12/2006 - 1:55am

semi subtle trolling is still trolling.

by Dave Brude (not verified) :: Tue, 12/12/2006 - 2:32am

Edge over's Barry's 2000 yards compaign only proves the point that DPAR ratings aren't the end all be all as the outsiders themselves often say. Ya Edge was highly successful but teams were loading up to stop Peyton and company making it very easy for Edge to do what he does best. I mean Addai has an even more rediculous success rate than Edge did last season granted its on fewer carries but still.

by Dave Brude (not verified) :: Tue, 12/12/2006 - 2:34am

Also what is the average DPAR penalty for a fumble? -4?

by kevinNYC (not verified) :: Tue, 12/12/2006 - 2:38am

#7... The problem with your argument is the fact that GB and Favre are where they are (DVOA rankings) despite playing a craptacular schedule. Manning and the Giants are where they are having played one of the NFL's three toughest schedules. If Favre was as good as you say he is, the Packers should be better considering their opponents. The Giants certainly aren't where they are in spite of Eli or even 100% because of him. Saying that Favre would be better if this or that happened is pure conjecture, especially on this website.

I'm not surprised to see Brandon Jacobs as the bottom of the barrel RB this week. It's starting to approach the 2005 "Jacobs is coming in the game so he's definately getting the ball" problem with the Giants. There wasn't such a problem earlier in the year, but Petitgout's injury and general wear and tear of o-linemen have zapped their effectiveness. Coughlin and Hufnagel would be better off mixing in some passes in those short spots.

by Walter Lawrence (not verified) :: Tue, 12/12/2006 - 4:30am

Have you guys ever created a metric to examine DPAR per touch? That might give you something else to judge who has had the best season ever. It seems like San Diego goes out of its way to underuse LT, and maybe his DPAR per touch would be more than Priest Holmes in 2004.

by D (not verified) :: Tue, 12/12/2006 - 4:43am

If you check the individual stat pages you will see in addition to DPAR a stat called DVOA. DVOA, in this context, represents value per play where as DPAR is total value. It isn't exactly the same as a "DPAR per touch" stat because DVOA is compared to league average while DPAR is compared to replacement level and the unit used to measure DVOA is different than DPAR, but it is pretty close to what you are looking for.

by Yaguar (not verified) :: Tue, 12/12/2006 - 5:16am

17: "The problem with your argument is the fact that GB and Favre are where they are (DVOA rankings) despite playing a craptacular schedule."

Do you understand what the D in DVOA stands for?

by hooterrooter (not verified) :: Tue, 12/12/2006 - 9:02am

Can we start calling the Saints backfield BB Mack?

by BadgerT1000 (not verified) :: Tue, 12/12/2006 - 10:34am


The one cautionary note is that Favre's accuracy is decreasing. Once upon a time he could compensate for guys unable to get separation (Antonio Freeman, Bubba Franks) by drilling the ball into an area smaller then Matt Millen's brain. But those days are few and far between now.

I don't think it is any coincidence that Bubba Franks reduction in catches has tracked Favre's accuracy problem. Brett recognizes his issue and in the interests of avoiding the turnover won't force the ball to Franks who couldn't outrun Pat Williams in snowshoes.

But the difference in the Packers/49ers on Sunday was simply Favre vs. Smith. Favre had an "almost" pick but otherwise did a fine job. As you described, he kept plays alive by sliding in the pocket or on designed rollouts, made some really nice throws, and exploited a bad San Fran secondary.

Smith was dreadful. Vernon Davis was open MULTIPLE times and all Smith could manage was a 4 yard pass that Davis turned into a 52 yard TD. The Packer secondary is just as bad if not worse then the 49er group thanks to doofus A and doofus B in safeties Manuel and Collins. Yet Alex sprayed the ball around like a frat boy urinating his name in the snow.

So fans can b*tch, b*tch, b*tch about the announcers and Favre but until the Packers come up with a legit Option B folks just need to give it a rest.

by JasonK (not verified) :: Tue, 12/12/2006 - 10:35am


The Giants eventually did start throwing with Jacobs in the lineup. Shockey was wide open on his TD reception largely because the defense bit so hard on the play-fake to Jacobs.

But I agree that Petitgout's injury is most of what's driving Jacobs' diminished impact. The defenses are overloading the offense's right side, because they know that, if the run goes left, Whitfield isn't going to be able to move his man. Whitfield has been surprisingly effective in pass-blocking over the last few weeks, but he isn't nearly as good as Petitgout was in the run game.

by TGT (not verified) :: Tue, 12/12/2006 - 10:53am

@11: DVOA and DPAR take into account game score, field position, down and distance, and time on the clock. There's no official "clutch" component, but everything used to determine clutchness other than QB name is accounted for.

@15: If you read the outsiders frequently (or at all), you'd know that they believe there is still room for improvement in DPAR. They also have been trying to make it clear recently that a RB's DPAR and DVOA are really not representative of the RB in a vaccuum. They measure the entire offense on running plays. DPAR and DVOA do take into account how good defenses are overall, but they cannot take into account things like the defense playing 7 in the box (against the colts) or 8-9 in the box (against some of the old lions teams). If it's not in the play by play, it's not in DPAR/DVOA.

While Barry Sanders was head and shoulders above Edge as a RB, the sum of the results of his running plays in 1997 was not as good as the sum of the results of Edge's running plays in 2005.

by Chris Heinonen (not verified) :: Tue, 12/12/2006 - 11:12am

4, 15: I believe those are the best seasons since 1997, so it does not include the Barry Sanders 2000 yard season in them. At least I'm hoping it doesn't, since I'd think Sanders would be in the Top 10 with that year for sure, but I'm sure someone will correct me if I'm wrong.

by Andrew (not verified) :: Tue, 12/12/2006 - 11:31am

Did Mike Karney have more value than the entire Dallas offense combined?

Romo was negative value, Barber was negative value, Crayton was negative value (0-2 on receptions), Owens should be negative value (2-6 on receptions), Witten should be negative value (4-7 with an interception on receptions), Polite was negative value (0-1 on receptions)

by B (not verified) :: Tue, 12/12/2006 - 12:02pm

18: I'm not sure why you think LdT is underused. He has 257 rushes as of last week, which puts him behind only Chester Taylor and Larry Johnson. Add his receptions, and he's among the top-10 in RB touches.

by admin :: Tue, 12/12/2006 - 12:23pm

Actually, Barry Sanders' rushing value in 1997 does exceed Edge's rushing value in 2005. But the table is combined rushing and receiving value. Barry Sanders 1997 is like 12th or something in combined.

In addition, one thing we may need to do with past years is fiddle with baselines. For example, rushing was down in 1998 unless your name was "Terrell Davis." If we fiddle with baselines, Barry Sanders might pass those 2005 Edge + LJ seasons. But of course, this would make the argument that LT's 2006 season is not the greatest of all time even stronger.

by navin (not verified) :: Tue, 12/12/2006 - 12:30pm

I agree with the Badger.

Favre and Smith basically got to face equally bad defense. Favre's numbers look a little better than they were because of a few big plays Driver made for him and because of dropped interceptions--SF dropped two gimme ints. But for the most part he was very efficient with plenty of time to throw. The one time SF got pressure, they sacked him and forced a fumble, which Bill Leavy screwed up.

OTOH, Smith was horrible except for two drives. He threw interceptions on two plays where he had receivers wide open. One of them could possibly be attributed to miscommunication. Vernon Davis ran a post-corner route and was WIDE open around the 15 yard line. Smith thought that Davis was going to flatten his route off and threw the ball about 8 yards short, giving the beaten defender an easy interception. There were a few other plays like this, but they didn't end up with as bad of a result. Oh, also two of his picks were absolute killers, both were picked off in the red zone.

One last thing, not relating to GB-SF. Did anyone else notice that Tim Rattay actually gives Tampa a chance to win? What the heck was Gruden thinking the whole year. I really think a team like the Vikes or the Browns should go after Rattay this offseason. He may be immobile, but he's an efficient passer when he gets time to throw.

by kevinNYC (not verified) :: Tue, 12/12/2006 - 1:00pm

#20... What is your point? My argument is the fact it's hard to say Brett Favre would do better with another team with more talent when you consider his performances against teams with similar talent using DVOA ranking as the measure of talent. Yes, I realize it's a huge stretch to correlate DVOA ranking and team talent, but that's what I was doing. Favre's performance against those teams have been inconsistent... good performances like MIN and STL and bad performances like SEA and NYJ.

by 10K (not verified) :: Tue, 12/12/2006 - 1:17pm

You guys need to get off the Wilfork roughing call. Yes, it was a bad call. But what happened was the ref did not fully see it... meanwhile, Wilfork was rolling around on his back like a turtle trying to get a hand on Harrington so it appeared to the ref as a penalty... Plus, Harrington was hurt as a result of the play. Yes, a bad call, but it wasn't a blatantly bad call (as in the ref saw what we saw and still called it, rather the ref didn't fully see it, Wilfork's behavior was suspicious and Harrington got hurt, hence the call in order to protect QBs). As you said, it didn't matter... Likewise, the call on Bell later was fully viewed and a bad call. He had left his feet, tackled Brady with his arms, didn't drive through him as he was making the pass... Of course, that didn't matter either despite keeping the drive alive. Miami got the sack and stripped the ball, and Bell recovered it. How that affects Harrington being better rated by your system I don't know or care. But it sounds lame to claim a FG resulting from a roughing call bumped up Harrington and conversely, Miami's D completely dominating Matt Light and having good coverage mostly throughout hurt Brady. Boo hoo.

by Sophandros (not verified) :: Tue, 12/12/2006 - 1:20pm

21: No, only because it ignores the Saints' most potent weapon.

by MJK (not verified) :: Tue, 12/12/2006 - 1:30pm

Idle question, but I can't find it on the DVOA explanation page.

I know DVOA is based on "success points", and I know it includes time remaining in the game as an adjustment variable. But does it include time and score differential when defining a success? I.e., when a team is in a clock-killing situation, staying in bounds, or completing a pass for no yardage, is much better than going out of bounds (even if you get an extra yard) or throwing incomplete. Conversely, if you're behind and low on time, a 5 yard run on 1st and 10 that goes out of bounds is better than a 9 yard one that stays in bounds. Does DVOA change it's definition for "success" in such situations? How about DPAR? I.e. is a player penalized if they fail to get out of bounds when they need to stop the clock?

by Peter (not verified) :: Tue, 12/12/2006 - 1:49pm


DPAR has a pretty big weakness somewhere (haven't singled it out yet).

I saw a Tomlinson game against Oakland this year where he singlehandedly won the game for his team, but his DPAR was microscopic.

Something is amiss when a system disregards that performance in a victory and overvalues a performance like Peyton Manning's.

by Steve Sandvik (not verified) :: Tue, 12/12/2006 - 2:17pm

34--could you be less specific?

by Dev (not verified) :: Tue, 12/12/2006 - 2:22pm

RE: #22

"by drilling the ball into an area smaller then Matt Millen’s brain."

Brilliant in so many ways! Clearly, what Millen needs is a football trepanning delivered by Favre.

by Sophandros (not verified) :: Tue, 12/12/2006 - 2:26pm

26: Well, he DID score more points than they did...

by Wanker79 (not verified) :: Tue, 12/12/2006 - 2:41pm

Re: 34

Without further details, I'm going to assume that you (like many other people) are confusing DPAR and DVOA. DPAR is a cumulative stat. So if Tomlinson's performance didn't account for a great deal of yards, but those relatively few yards came at very valuable times, DPAR would be low but DVOA would be high.

by Crushinator (not verified) :: Tue, 12/12/2006 - 2:51pm


I think it's because the Chargers really have tried to spare him a lot of carries. Turner's been getting the rock a lot to try and not burn out LdT - As opposed to Larry Johnson who has 5 more rushing yards than Tomlinson on 50 more carries, or Chester Taylor, who only seems to get a rest when he's injured.

by Pat (not verified) :: Tue, 12/12/2006 - 2:53pm

No, it was the Oakland-San Diego game two weeks ago. Yeah, Tomlinson won the game for them. They also only won by 7 points, and only scored 21 in total, and Tomlinson had 4.6 DPAR that game. A replacement-level team would score ~15 points against Oakland's defense (an average team would score ~19 points versus them). San Diego scored 21. Virtually all of their points above replacement level came from Tomlinson.

So yeah, DPAR said that Tomlinson single-handedly won that game for them as well.

It should also be noted that 4.6 DPAR/game is pretty damn good. Unfortunately for Tomlinson, that was a banner week for running backs. Remember that was the week that Joseph Addai got to run an entire game versus a nickel defense that DVOA hadn't realized was falling apart as a run defense yet.

by Carlos (not verified) :: Tue, 12/12/2006 - 3:04pm

Ladell Betts had a lot of yards but many of them came on carries of two yards or fewer

Scratching my head trying to figure out what you mean. Maybe you mean, "Betts had a lot of yards, but he also had a lot of carries for two yards or fewer." But there's no way given that he had 171 yards on 33 carries that he gained what any native speaker would call "many" of those yards on short carries.

I don't know all the ins and outs of your formula, but my guess is that Betts didn't rank b/c he didn't score a TD, not because of short carries. Seriously, what kind of metric says Karney was 2d most valuable when all he did was "poach" TDs, but Betts who ran 33 times for more than 5 yards per carry isn't even in the top 10.

Ladell Betts had exactly the kind game I'd want from a back. Sure, a third of his carries went for less than 3 yards, but he had only 2 carries for loss, and he had 2/3 of his carries for 3+ yards, including about a quarter than went for 10-20 yards and didn't get any yardage padding on a long run (no runds over 20 yards).

You might want to rethink the metrics.

by Richie (not verified) :: Tue, 12/12/2006 - 3:32pm

You might want to rethink the metrics.

hehe. I always love these arguments. "You are 'wrong' on Betts, so the system sucks."

by kubiwan (not verified) :: Tue, 12/12/2006 - 6:02pm

the greatest running back season since 1997 belonged to Priest Holmes in 2002

Oh what might have been! This was the year he missed the last 2 1/3 games of the season after hurting his hip. He was on a 100 DPAR pace before then.

by Chris M (not verified) :: Tue, 12/12/2006 - 6:52pm

Regarding the comment on Vince Young - doesn't DPAR implicitly include "clutch" plays?

by Carlos (not verified) :: Tue, 12/12/2006 - 7:32pm

hehe. I always love these arguments. “You are ‘wrong’ on Betts, so the system sucks.�

My bad. I forgot that the FO metrics are perfect. Gosh, where did I get the idea that they're still in their infancy?

Also, note that I never said anything like "the system sucks." I made very specific observations on Betts' day and the strangeness of awarding lots of "value" to a fullback who poaches a few TDs, and made a very specific comment: TDs may be overvalued.

I really enjoy this site, but I could do without the insipid comments like "hehe." What's the point in that?

by Dave Brude (not verified) :: Tue, 12/12/2006 - 8:58pm

Still wondering what the DPAR penalty is for fumbles. I'm curious to see where RB's would rank without fumbles factored in.

by DrewTS (not verified) :: Tue, 12/12/2006 - 11:50pm

Re 46

I think the fumble penalty changes depending on where the fumble occurs (i.e. in the backfield vs 40 yards downfield). I could be wrong though. I remember it being explained somewhere. I'm sorry I couldn't be less helpful.

by Pat (not verified) :: Wed, 12/13/2006 - 12:08pm

I don’t know all the ins and outs of your formula, but my guess is that Betts didn’t rank b/c he didn’t score a TD,

Well, given that the title of the article pretty much says "TDs are overvalued" I'm guessing the answer to that is "no."

Ladell Betts had exactly the kind game I’d want from a back. Sure, a third of his carries went for less than 3 yards, but he had only 2 carries for loss, and he had 2/3 of his carries for 3+ yards, including about a quarter than went for 10-20 yards and didn’t get any yardage padding on a long run (no runds over 20 yards).

First point: he had three tackles for loss, not two - a 3 yard loss on 2nd and 6, a 2 yard loss on 1st and 10, and a 3 yard loss on 1st and 10.

But carries for 2 yards or less - unless they come when you need short yardage - are carries for loss. Yay, they give you positive yardage. But they also make it less likely for you to sustain a drive - you're more likely to get a new first down from 1st and 10 than you are from 2nd and 8.

Want a funny statistic? Every single one of Washington's drives that stalled (i.e. every one except that Randle El touchdown and the two interceptions) began its final series with a Betts run for 3 yards or less. And one interception came partially because Betts had a three-yard loss, putting them in a third and long.

I'm not saying Betts was bad. I'm not even saying Betts was average. But he didn't light up the place.

Seriously, what kind of metric says Karney was 2d most valuable

Well, it is a lot harder to run on Dallas than it is to run on Philadelphia.

I do agree, though, that I was surprised to see Karney so high - but when you look at it, every single one of Karney's receptions and catches were a big deal - 9 yards on 1st and 10, 8 yards on 2nd and 1, 12 yards on 1st and 10, 9 yards on 1st and 10, 4 yards on 2nd and 1, plus the two touchdowns. If the game had been full of plays like that, New Orleans would've won like 80-17. Dallas never stopped Karney for what could possibly be considered a negative play for New Orleans.

by Carlos (not verified) :: Wed, 12/13/2006 - 3:51pm

But they also make it less likely for you to sustain a drive

True enough. But when 1/3 of your runs are

by Wanker79 (not verified) :: Wed, 12/13/2006 - 7:16pm

In order for a play to be considered successful it needs to gain 45% on 1st down, 60% on 2nd down, and 100% on 3rd and 4th down.

That means that only about 1/a of Betts' 36 plays (13 of them) are considered successful.

Over 1/3 of his 36 plays (13 of them) went for less than 25% of the yardage needed, regardless of down. That includes four gains of 2 on 1st & 10 and one on 2nd & 10; a gain of 3 on 2nd & 13; a gain of 5 on 3rd & 31; two gains of 1 on 1st & 10; and incomplete pass on 1st & 10; and losses of 2 and 3 yards on 1st & 10 and a loss of 3 on 2nd & 6.

And the other 1/3 of his plays fell somewhere between miserable failures and successes (like 4 gains of 4 on 1st & 10)

Is that really "exactly the kind game I’d want from a back"?!?

by Wanker79 (not verified) :: Wed, 12/13/2006 - 7:19pm

Re: 46

I don't know about DPAR, but this is from the DVOA explaination:

"...a fumble is worth anywhere from -1.70 to -3.98 points depending on how often a fumble in that situation is lost to the defense - no matter who actually recovers the fumble."

by Pat (not verified) :: Wed, 12/13/2006 - 8:54pm

True enough. But when 1/3 of your runs are

I'm guessing you were going to say "between 10 and 20 yards"? Or something close?

Keep in mind what #50 said: a third of Betts's runs were failures. He's got a third that are good successes, but those essentially even out between the two. So you're really talking about about 11 plays above average or so - and again, this is against a team where the average back succeeds more than he fails.

That's kindof where Karney stands out. Karney had 8 plays. Absolutely none of them could be considered failures of any sort - not one incomplete pass, not one play that put the team in worse position than before. And that was against a defense where the average back fails more than he succeeds - especially passing.

by Jason Mulgrew aka The Mul Dawg aka J-Rocka (not verified) :: Thu, 12/14/2006 - 2:08am

Priest Holmes used to be good.

by masocc (not verified) :: Thu, 12/14/2006 - 8:24am

Jason Mulgrew used to be funny.

by Yaguar (not verified) :: Thu, 12/14/2006 - 8:43am

He did?

by Jason Mulgrew aka The Mul Dawg aka J-Rocka (not verified) :: Thu, 12/14/2006 - 7:45pm

More funny than you ever were, Yaguar.