Writers of Pro Football Prospectus 2008

16 Oct 2006

2006 Quick Reads: Week 6

Our weekly roundup of Sunday performances according to FO stats addresses two running backs who defy father time, two Super Bowl quarterbacks getting their games into gear after some problems early in the season, and a lot of wide receivers having very good days.

Posted by: Aaron Schatz on 16 Oct 2006

46 comments, Last at 20 Oct 2006, 3:29pm by morganja


by Jerry (not verified) :: Mon, 10/16/2006 - 5:09pm

For more on Foster, click here.

Now that's just cold!

by Moridin (not verified) :: Mon, 10/16/2006 - 5:10pm

But it made me laugh my butt off.

by Jerry (not verified) :: Mon, 10/16/2006 - 5:12pm

Yeah, even as a Panthers fan, I must admit, I got a good laugh out of it myself.

by J.R. (not verified) :: Mon, 10/16/2006 - 5:14pm

This Phillip Rivers can really play some football. I had my doubts when he was drafted but no longer. His accuracy is just amazing - I don't think any NFL QB is better. Makes every throw, right on the numbers. It's just, sick. I'll be shocked if he isn't the best player in the league within three years.

by JasonK (not verified) :: Mon, 10/16/2006 - 5:18pm

I'm surprised that Vick's rush DPAR is so low. Do his 4 fumbles count against his rushing DPAR, or his passing DPAR? (They each happened in the backfield on a sack/botched handoff.) Or were there just lots of 1-yd runs that I'm forgetting?

And I read that 4-yd pass to Shockey on 3rd&7 as a checkdown when nobody else was open. I don't think Eli really expected to convert the first down. (Note also that it was the only pass thrown to Shockey all game that didn't get a 1st down or TD.)

by Matthew Furtek (not verified) :: Mon, 10/16/2006 - 5:30pm

I swear Rivers is throwing exactly like Brees used to... at least on those intermediate routes... the deep outs, deep stops, and deep ins...

He looks a lot better on the longer throws.

by Becephalus (not verified) :: Mon, 10/16/2006 - 5:36pm

The Foster bit hade me spitting up my drink into the glass, gross.

by LyleNM (not verified) :: Mon, 10/16/2006 - 5:38pm

I'm pretty sure that Morris only fumbled once...even if it was at such a bad time as to be worth two fumbles.

by Podge (not verified) :: Mon, 10/16/2006 - 5:48pm

Just like to add myself to the list of people ignoring 99% of the article and praising the Deshaun Foster joke. Good job Aaron.

by Gerry (not verified) :: Mon, 10/16/2006 - 6:01pm

I was looking at the advanced stats for the year posted, and I was struck by the high catch % for Houston wideouts and running backs. Moulds, 88%. Andre, 72%. Walter, 75%. Cook, 78%. Dayne, 71%.

Are they all that sure-handed, or has Carr (and Sage!) been throwing extremely accurate and catchable balls this season?

But then, it does not appear to carry over to the TEs. Daniels, 54%. Putzier, 40%.

by Gerry (not verified) :: Mon, 10/16/2006 - 6:02pm

I mentioned this on another thread, but here looks like a good place too.

For the season, Tiki is #1 in DPAR and #7 in DVOA. Meanwhile, Brandon Jacobs is #7 in DPAR and #1 in DVOA.


by MRH (not verified) :: Mon, 10/16/2006 - 6:04pm

The Redskins seemed to outsmart themselves this week with overcomplicated play-calling.

At the risk of pulling a Foster here, let me repeat what I wrote in the Audibles thread:

For five years of watching Saunders’ play calls for the Chiefs I have compplained to my wife about this. There is a lot to like about al saunders as an OC but I’ve never believed in his “play-calling genius� that I’ve read about in the media from time to time.

One Redskin nailed it in this AM’s Boswell column in the WashPost:

“As soon as we start establishing something, then we go and do some tricky-dicky stuff,� said one veteran.

by Irishfan (not verified) :: Mon, 10/16/2006 - 6:12pm

I believe in DPAR I really do. However, can someone please explain to me how on earth Torry Holt is ranked 6th in WR this week?

154 yards and 3 TDs is good enough for 6th???

He is worth 4.4 DPAR - the last TD alone was worth 7 DPAR as that was a play that a replacement level player definitely does not make.

Aaron or somebody please explain? He caught every catchable ball thrown his way. Ask your gamecharters. He even caught some uncatchable ones.

by Jake (not verified) :: Mon, 10/16/2006 - 6:15pm


And yet the Chiefs still scored tons of points. I was fine with Saunders. He played to Trent and a no-name wideout corps strengths. The 'Skins are helmed by a guy with far fewer skills, Brunell and don't have the incredible line the Chiefs did. What did you expect?

by Yaguar (not verified) :: Mon, 10/16/2006 - 6:18pm

13: Just a lot of passes intended for him that weren't caught.

From watching the game, you can tell that those bad passes were generally Bulger's fault, not Holt's. But the DPAR computer doesn't watch games, and it can't make that sort of judgement. It just sees lots of incomplete passes intended for Holt.

by Irishfan (not verified) :: Mon, 10/16/2006 - 6:24pm

Re 15 - Point taken but Roy Williams 10-17 (58.8%) for 161 yards (average 16.8) and 1 touchdown gets a DPAR of 5.5.

Torry Holt 8-14 (57.1%) for 154 yards (average 19.3) and 3 touchdowns gets a DPAR of 4.4.

So an extra 1.7% of completion percentage is worth 1.1 DPAR AND 2 touchdowns.

Still Confused

by Pat (not verified) :: Mon, 10/16/2006 - 6:36pm

2 of Holt's touchdowns came from 10 yards or less. You can't give him too much of a boost for that: the only reason Roy Williams didn't get 3 touchdowns is because Detroit wanted Kevin Jones to run it in twice (he didn't get it once).

by Ryan (not verified) :: Mon, 10/16/2006 - 6:37pm

Perhaps because of the Defensive adjustments? Except..Seattle and Buff have had similar pass DVOA's this year...very peculiar. Maybe Buff is better against Roy's particular WR spot and Sea isn't as good against Holt's particular WR spot or something along those lines. Rember, DPAR accounts for opponent adjustments so similar games may have much different DPARs. Except...Sea has a higher DVOA Vs. both #1 and #2 receivers....so I'm pretty confused too..So..maybe it had something to do with the average gain on each particular play between the two..Maybe Williams was catching passes and gaining more yards when he would be expected to gain fewer..DPAR's so complex it's hard to know exactly. And remember, one big gain for 50 yards and one TD is worth less than 3 catches for 60 yards and 3 firstdowns, so maybe there was something along those lines going on. I didn't see a play of the Det Buf game so I wouldn't know what Williams actually did.

by Ryan (not verified) :: Mon, 10/16/2006 - 6:49pm

Yeah. I looked at the play-by-play and noted that Holt got 105 of his yards on two catches and just 49 more yards on his other 6 catches. So that would give an indication of how lopsided his game was; a couple of really big plays plus a bunch of tiny ones. Williams on the other hand had 8 catches of at least 10 yards. That would be why his DPAR is higher; he had more good plays and only a couple of worthless cathces while Holt had two monster catches and a bunch of tiny ones.

by kibbles (not verified) :: Mon, 10/16/2006 - 7:08pm

When in doubt about DPAR numbers, the answer is usually "third downs". I haven't checked the numbers, but I suspect that Williams had a much higher conversion rate on 3rd down. If all of Williams' non-catches came on 1st down, and all of Holt's came on 3rd down, that's the difference right there.

by ABW (not verified) :: Mon, 10/16/2006 - 7:11pm

Re: 20

That's exactly right. I've been trying to post that Williams had two big 3rd down conversions, and Holt had no catches on third down, but it keeps disappearing into the ether....

by ABW (not verified) :: Mon, 10/16/2006 - 7:12pm

Hey! It worked this time! The mysteries of the internet....

by Yaguar (not verified) :: Mon, 10/16/2006 - 7:23pm

Looking at the play-by-play:

Roy Williams had seven first downs, a touchdown, and two other catches. Holt had three first downs, three touchdowns, and two other catches. Two of Holt's touchdowns were short, which does not count much for DPAR because most offenses can get a score from close range.

If the Rams decided to run Steven Jackson in on those plays, Williams would have had the better day by any reasonable measurement.

Or, we could say it's due to FO's Anti-Rams bias.

by Irishfan (not verified) :: Mon, 10/16/2006 - 7:39pm

Re 21 - From Play by Play on NFL.com

3-7-SEA9 (9:42) (Shotgun) M.Bulger pass short left to T.Holt for 9 yards, TOUCHDOWN.

Is it not a 3rd down conversion when its a TD?

Re 19

Holt had catches/result of

38 yards - 1st down
5 yards - 2nd and 5
9 yards - touchdown
4 yards - 1st down (red zone completion)
10 yards - touchdown
17 yards - 1st down
4 yards - 2nd and 6
67 yards - touchdown

He also drew a defensive holding penalty which resulted in another 1st down.

I still think his performance is worth more DPAR than Williams.

Pat - to say Williams didnt get two more TD's because they wanted to rush jones in - is that really the mike martz way?

by Irishfan (not verified) :: Mon, 10/16/2006 - 7:42pm

Re23 - 3rd and 7 from the 9 is hardly a rushing play and a receiver who scores a TD there is surely adding DPAR.

2nd and 8 from the 10 is hardly a 2 run gimme for Jackson either. I am not saying Bias - just the numbers look funny.

by Manning, E. (not verified) :: Mon, 10/16/2006 - 7:48pm

your numbers mean nothing, i might not be accurate but i win and winning is what this game is about.One day very soon im going to be in a position to win a big game, and i will. if you dont believe me just ask my brother or father. oh wait, no ask john elway.

by Irishfan (not verified) :: Mon, 10/16/2006 - 7:49pm

While I am at it Steve Smith had 6 first downs or touchdowns. Same as holt but 2 more of holts went for touchdowns. Admittedly 3 of smiths came on 3rd down against 1 for holt but does that really cause an extra 2.6 in DPAR when the 2 extra holt TDs are subtracted. I know this is where the D in DPAR comes into play but it still looks funny

by Jeff (not verified) :: Mon, 10/16/2006 - 8:08pm

re Williams vs. Holt

quantity matters as well. Williams had 2 extra catches (out of 3 additional attempts). Williams had 8 first down/TD catches, 1 success (8 yds on 1st-10) and 1 failure (1 yd on 1st-10). Holt had 6 first downs/TD catches, 1 success (5 yds on 1st-10) and 1 failure (4 yds on 1st-10). All else being equal, Williams had 2 more first down catches.

As for the TDs, a TD is not a significant value in itself, the catch and yards are. Holt's three TD catches could've been downed on the 1 yard line and he would not have lost a lot of his DPAR.

But also, the difference here is one point. It's not a huge amount.

by Mike (not verified) :: Mon, 10/16/2006 - 8:27pm

I called the Deshaun Foster joke about 5 minutes after the game was over. I knew Aaron would have some fun since Foster had actually be quite decent the last few weeks.

by gmc (not verified) :: Mon, 10/16/2006 - 8:40pm

It's about time Tennessee realized they weren't Steve McNair's team anymore, and stopped trying to prove that football should be played by throwing fourteen consecutive passes to fat tight ends standing on the line of scrimmage who then fall forward to gain yardage. It's about time they realized that if you have big fat tight ends, they can be used for blocking. It's about time they realized that with Travis Henry, Vince Young, Ben Troupe, and the Three Stooges at Wide Receiver (more properly two Stooges and a playing-hurt Drew Bennett), the team their offense most resembles is Atlanta.

Run the ball. And give Vince some choices; even bad decisions on the ground are better than letting the defense test his not-NFL-ready arm.

by Adam Gretz (not verified) :: Mon, 10/16/2006 - 9:01pm

I've been extremely rough on Nate Washington this season and i'm shocked, shocked I tell you, to see that he currently ranks 3rd among recievers in DVOA right now.

Not bad. He might rank first if he didn't have about 4 or 5 drops......

And can someone get the ball to Heath Miller?

by Josh (not verified) :: Mon, 10/16/2006 - 10:17pm

RE: 31-

The KC defense was all over Heath Miller yesterday. If you watch the Hines Ward touchdown catch, Miller was covered by a LB and the safety also keyed on him which is why Ward was so open. I haven't rewatched the whole game yet but I'd imagine they cheated on him during a lot of short yardage and goal-to-go situations. IIRC they kept him in to block more often than usual because Kendall Simmons was out and replaced by the rookie Chris Kemoeatu.

FYI: JJ Cooper does an OL review each week that usually posts on Tuesdays, but from what Steelers radio sideline reporter Craig Wolfley said during the game, Kemoeatu held his own.

by MRH (not verified) :: Mon, 10/16/2006 - 11:05pm

Re #14: How about giving the ball to Portis more than 14 times? How about pounding the ball with Portis, Betts, Duckett et al against a terrible run defense?

Look, Saunders ran a great offense in KC for the most part. Creative scheme, made stars of the rbs, worked the te in unlike the Martz version of the same offense, etc. That doesn't change my (seemingly Aaron's and the unnamed 'skin's) basic point: he sometimes gets away from basic football and calls things that are baffling. For example, vs. the Giants, Brunell running the faux option and throwing the forward option pass. When he's 4-0 and up two scores, run the "tricky-dicky" stuff. When he's 2-3 and the team with worst rushing defense in the league has rallied from 11 down, maybe he should put the Nixon playbook on the shelf and dial up some Eisenhower ground-pounding.

by clem (not verified) :: Mon, 10/16/2006 - 11:11pm

How did Kevin Jones miss the top rb list? (23-127 rushing, 6-36 receiving, 1 td) Weak D adjustment? Too many garbage yards? What was his DPAR anyhow?

by Jerry (not verified) :: Tue, 10/17/2006 - 5:17am

Re #32:

On the handful of plays where I watched Kemoeatu, he looked good.

by Sophandros (not verified) :: Tue, 10/17/2006 - 11:18am

So I take it that Grossman would take over the bottom slot for QBs?

by John A (not verified) :: Tue, 10/17/2006 - 11:21am

If you ran this again today, would Edge be the least valuable back?

by mactbone (not verified) :: Tue, 10/17/2006 - 11:52am

Edge is a loser league All-Star. The game last night might be one for the loser league Hall of Shame. Grossman has to be close to the worst performance by a QB and Edge racked up carries for no yardage or TDs.

by Pat (not verified) :: Tue, 10/17/2006 - 5:12pm

If you ran this again today, would Edge be the least valuable back?

By a wide margin.

by LyleNM (not verified) :: Tue, 10/17/2006 - 5:25pm

So how about getting Fox to correct how many fumbles Morris had? (It's still only one.)

by Oswlek (not verified) :: Wed, 10/18/2006 - 11:21am

Where is Rex Grossman?

by Wanker79 (not verified) :: Wed, 10/18/2006 - 12:36pm

Re: 41

Quick Reads is published before the MNF game. So nobody from Chicago or Arizon is listed.

by andrew apold (not verified) :: Wed, 10/18/2006 - 1:14pm

I'm assuming Edge will top Foster for least valuable...

by admin :: Wed, 10/18/2006 - 2:22pm

FOX blog now features articles on how bad Edge and Grossman were:


by Fiver (not verified) :: Wed, 10/18/2006 - 7:12pm

Seriously, how do you let Steve Smith get behind your entire defense when the Panthers are on their own 28-yard line and you're trying to get the ball back for a game-winning drive?

The answer is that Ed Reed screwed up. If you watch the replay, Rolle gives Smith plenty of cushion thinking Reed is going to be playing the deep middle zone, but Reed never drops back. He was playing man against someone going over the short middle. Reed stated in postgame interviews that he blew the defensive call.

Incidentally, it doesn't take a Steve Smith to score a TD on a play where one dback is bracketing him in a cushy zone on one side and the other dback is covering the wrong man. That particular play was probably the least impressive of Smith's catches in that game. More worthy of attention are his many great individual efforts on other catches that converted 3rd downs or change field position. The guy was on a mission.

by morganja (not verified) :: Fri, 10/20/2006 - 3:29pm

Looking at the once agin odd values placed on some performances by DPAR, it became a lot clearer to me where the biggest weakness of this statistical model lies. This model assumes that all offences are running the same system and therefore makes erroneous judgements off that misperception. Different systems make different demands on their players.

For people with a knowledge of football I give you exhibit one:

Bears 73 Redskins 0

exhibit two:

Falcons running attack in week 1.

Not having an answer to a scheme on any particular day is what leads to the most variance in scores. This model does nothing to factor that in. Nor does it have anything to say about whether a player does what the offense asks of him. Deshaun Foster is a case in point. He is asked to run into the center of the defense over and over and pick up two yards. That's what he does. All this Deshaun Foster bashing ignores the reality of what his offense is trying to do, which is set up the defense for big plays.