RSS Feed
DVOA Analysis
Football Outsiders' revolutionary statistics that break down every single play of the NFL season
26 Sep 2006
Week 4 DVOA Ratings
by Aaron Schatz
Once again this week, we provide two sets of ratings -- in two entirely separate tables. The first table has VOA through the first three games, no opponent adjustments of course. The only differences between "NON-ADJ VOA" and "TOTAL VOA" are fumble luck and the weather/altitude effects on special teams.
The second table is the new rating that combines early-season performance with the preseason projection. We have a name for this now. One of the hardest parts of doing statistical analysis is naming your new statistics so that they make sense. A lot of people have criticized us because DVOA, DPAR, ALY, etc. just turns into a bunch of alphabet soup. My typical response to "Why did you name it DVOA" is "Well, I can't just name it 'Dave.'"
Anyway, here we are with a new statistic, and nobody can come up with a good name, so this time I really am naming it Dave.
No, seriously. It's called DAVE: DVOA Adjusted for Variation Early. For those curious, the formula this week is 67 percent projection, 33 percent 2006 VOA. The separate pages for offense, defense, and special teams will be all be updated soon, as will individual pages.
UPDATE: Commentary is finally available on FOXSports.com.
I look forward to the deluge of hate mail from New Orleans. Hey, statistics can't measure emotion, but for the other 31 teams, it's still more accurate to combine the first three games with the preseason projection. Honestly, it probably is more accurate for New Orleans too.
* * * * *
These are the Football Outsiders team efficiency ratings through Week 3 of 2006, measured by our proprietary Defense-adjusted Value Over Average (DVOA) system that breaks down every single play and compares a team's performance to the league average based on situation in order to determine value over average. (Explained further here.)
OFFENSE and DEFENSE VOA are adjusted to consider all fumbles, kept or lost, as equal value. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver/Mexico City) and week of season.
There are no opponent adjustments for the first three weeks of the season, which is why offense and defense are "VOA" and not "DVOA."
To save people some time, please use the following format for all complaints:
<team> is clearly ranked <too high/too low> because <reason unrelated to DVOA>. <subjective ranking system> is way better than this. <unrelated team-supporting or -denigrating comment, preferably with poor spelling and/or chat-acceptable spelling>
| ACTUAL 2006 NUMBERS | |||||||||||
| TEAM | TOTAL VOA |
LAST WEEK |
NON-ADJ VOA |
W-L | OFFENSE VOA |
OFF. RANK |
DEFENSE VOA |
DEF. RANK |
SPECIAL DVOA |
S.T. RANK |
|
| 1 | SD | 83.8% | 2 | 90.3% | 2-0 | 31.3% | 2 | -48.9% | 2 | 3.6% | 7 |
| 2 | BAL | 65.7% | 1 | 82.1% | 3-0 | -0.3% | 15 | -59.8% | 1 | 6.2% | 3 |
| 3 | CHI | 49.6% | 3 | 60.1% | 3-0 | 13.4% | 6 | -22.2% | 6 | 14.0% | 1 |
| 4 | PHI | 45.4% | 6 | 37.6% | 2-1 | 39.2% | 1 | -5.4% | 15 | 0.7% | 17 |
| 5 | NO | 34.9% | 11 | 33.3% | 3-0 | 1.9% | 13 | -29.5% | 3 | 3.5% | 8 |
| 6 | CIN | 25.3% | 4 | 29.4% | 3-0 | 0.6% | 14 | -21.4% | 8 | 3.3% | 10 |
| 7 | JAC | 18.9% | 5 | 23.0% | 2-1 | -1.9% | 17 | -25.6% | 4 | -4.9% | 26 |
| 8 | SEA | 18.8% | 8 | 17.6% | 3-0 | -0.6% | 16 | -22.0% | 7 | -2.5% | 21 |
| 9 | STL | 18.1% | 12 | 25.6% | 2-1 | -1.9% | 18 | -17.5% | 9 | 2.5% | 13 |
| 10 | IND | 14.1% | 13 | 15.8% | 3-0 | 25.2% | 3 | 15.9% | 29 | 4.8% | 4 |
| 11 | BUF | 12.7% | 9 | 7.0% | 1-2 | 3.6% | 10 | -2.4% | 17 | 6.8% | 2 |
| 12 | NYJ | 11.8% | 15 | 12.1% | 2-1 | 20.4% | 4 | 4.6% | 20 | -4.1% | 24 |
| 13 | WAS | 7.3% | 21 | 8.9% | 1-2 | 12.1% | 7 | 7.1% | 23 | 2.3% | 14 |
| 14 | DAL | 3.6% | 16 | 1.6% | 1-1 | 3.3% | 11 | -8.5% | 13 | -8.2% | 30 |
| 15 | NE | -0.6% | 14 | -2.6% | 2-1 | 9.9% | 8 | 14.0% | 27 | 3.4% | 9 |
| 16 | NYG | -2.5% | 18 | 3.0% | 1-2 | 14.7% | 5 | 16.2% | 30 | -0.9% | 18 |
| TEAM | TOTAL VOA |
LAST WEEK |
NON-ADJ VOA |
W-L | OFFENSE VOA |
OFF. RANK |
DEFENSE VOA |
DEF. RANK |
SPECIAL DVOA |
S.T. RANK |
|
| 17 | PIT | -3.8% | 20 | -12.5% | 1-2 | -20.3% | 28 | -23.8% | 5 | -7.3% | 28 |
| 18 | SF | -5.2% | 10 | -9.1% | 1-2 | 6.7% | 9 | 14.9% | 28 | 3.0% | 11 |
| 19 | MIN | -5.8% | 17 | -8.8% | 2-1 | -11.6% | 23 | -8.9% | 11 | -3.1% | 23 |
| 20 | ATL | -10.2% | 7 | -0.8% | 2-1 | -8.0% | 20 | -14.2% | 10 | -16.4% | 32 |
| 21 | KC | -10.3% | 19 | -8.3% | 0-2 | -20.1% | 27 | -8.6% | 12 | 1.2% | 16 |
| 22 | CAR | -11.4% | 23 | -13.3% | 1-2 | -15.7% | 24 | -2.6% | 16 | 1.7% | 15 |
| 23 | DEN | -15.2% | 24 | -13.4% | 2-1 | -8.0% | 21 | -0.3% | 18 | -7.4% | 29 |
| 24 | MIA | -18.4% | 26 | -4.9% | 1-2 | -19.7% | 26 | -5.7% | 14 | -4.4% | 25 |
| 25 | DET | -20.6% | 28 | -23.8% | 0-3 | -2.6% | 19 | 8.8% | 25 | -9.2% | 31 |
| 26 | GB | -22.5% | 22 | -20.2% | 1-2 | -8.5% | 22 | 7.0% | 22 | -7.0% | 27 |
| 27 | ARI | -30.5% | 25 | -24.3% | 1-2 | -15.8% | 25 | 11.9% | 26 | -2.8% | 22 |
| 28 | CLE | -31.1% | 27 | -37.4% | 0-3 | -26.3% | 29 | 8.6% | 24 | 3.8% | 6 |
| 29 | TB | -40.6% | 30 | -42.0% | 0-3 | -35.2% | 31 | 4.4% | 19 | -1.0% | 19 |
| 30 | HOU | -46.9% | 29 | -39.3% | 0-3 | 2.5% | 12 | 52.2% | 32 | 2.7% | 12 |
| 31 | TEN | -55.2% | 31 | -49.6% | 0-3 | -31.5% | 30 | 21.6% | 31 | -2.1% | 20 |
| 32 | OAK | -76.5% | 32 | -91.9% | 0-2 | -73.7% | 32 | 6.9% | 21 | 4.0% | 5 |
| DAVE (DVOA ADJUSTED FOR VARIATION EARLY) |
||||||||||
| TEAM | TOTAL VOA |
LAST WEEK |
W-L |
OFFENSE VOA |
OFF. RANK |
DEFENSE VOA |
DEF. RANK |
SPECIAL DVOA |
S.T. RANK |
|
| 1 | SD | 42.7% | 2 | 2-0 | 21.8% | 2 | -19.9% | 2 | 0.9% | 11 |
| 2 | SEA | 33.3% | 1 | 3-0 | 21.6% | 3 | -13.0% | 4 | -1.3% | 22 |
| 3 | CHI | 24.6% | 5 | 3-0 | 3.2% | 9 | -16.5% | 3 | 4.8% | 1 |
| 4 | PHI | 23.7% | 6 | 2-1 | 16.6% | 4 | -5.3% | 10 | 1.9% | 6 |
| 5 | IND | 23.6% | 4 | 3-0 | 29.3% | 1 | 5.2% | 27 | -0.5% | 19 |
| 6 | BAL | 22.3% | 7 | 3-0 | -2.8% | 16 | -23.7% | 1 | 1.4% | 8 |
| 7 | CIN | 21.7% | 3 | 3-0 | 15.5% | 5 | -5.3% | 9 | 0.9% | 12 |
| 8 | JAC | 9.6% | 9 | 2-1 | 1.5% | 11 | -9.5% | 5 | -1.3% | 23 |
| 9 | PIT | 9.6% | 10 | 1-2 | 3.1% | 10 | -8.6% | 6 | -2.0% | 25 |
| 10 | DEN | 7.6% | 11 | 2-1 | 10.4% | 7 | -0.5% | 18 | -3.3% | 29 |
| 11 | WAS | 5.0% | 17 | 1-2 | 1.5% | 12 | -1.2% | 16 | 2.3% | 4 |
| 12 | NE | 4.8% | 12 | 2-1 | 13.2% | 6 | 8.6% | 30 | 0.3% | 14 |
| 13 | CAR | 4.0% | 14 | 1-2 | 1.1% | 13 | -3.0% | 13 | -0.1% | 16 |
| 14 | ATL | 2.7% | 8 | 2-1 | -1.1% | 15 | -8.4% | 7 | -4.6% | 31 |
| 15 | KC | 1.9% | 13 | 0-2 | -4.4% | 18 | -6.6% | 8 | -0.3% | 17 |
| 16 | NYG | 1.0% | 15 | 1-2 | 5.3% | 8 | 3.9% | 24 | -0.5% | 18 |
| TEAM | TOTAL VOA |
LAST WEEK |
W-L |
OFFENSE VOA |
OFF. RANK |
DEFENSE VOA |
DEF. RANK |
SPECIAL DVOA |
S.T. RANK |
|
| 17 | DAL | 0.1% | 16 | 1-1 | -1.0% | 14 | -3.3% | 12 | -2.2% | 26 |
| 18 | MIA | -6.0% | 18 | 1-2 | -11.6% | 26 | -4.2% | 11 | 1.4% | 9 |
| 19 | NO | -8.7% | 31 | 3-0 | -10.1% | 25 | -0.2% | 19 | 1.2% | 10 |
| 20 | STL | -9.8% | 26 | 2-1 | -8.0% | 21 | 1.2% | 20 | -0.7% | 20 |
| 21 | BUF | -10.7% | 20 | 1-2 | -11.9% | 28 | 3.0% | 22 | 4.2% | 2 |
| 22 | DET | -12.2% | 22 | 0-3 | -3.5% | 17 | 5.1% | 26 | -3.6% | 30 |
| 23 | NYJ | -13.3% | 27 | 2-1 | -5.5% | 19 | 8.1% | 29 | 0.2% | 15 |
| 24 | MIN | -14.3% | 23 | 2-1 | -14.3% | 29 | -1.7% | 14 | -1.7% | 24 |
| 25 | CLE | -15.8% | 21 | 0-3 | -16.4% | 30 | 1.4% | 21 | 2.0% | 5 |
| 26 | GB | -16.0% | 25 | 1-2 | -8.9% | 22 | 3.8% | 23 | -3.3% | 28 |
| 27 | TB | -17.5% | 19 | 0-3 | -17.0% | 31 | -0.7% | 17 | -1.3% | 21 |
| 28 | SF | -18.2% | 24 | 1-2 | -9.2% | 23 | 6.5% | 28 | -2.5% | 27 |
| 29 | ARI | -21.4% | 28 | 1-2 | -11.8% | 27 | 5.0% | 25 | -4.7% | 32 |
| 30 | TEN | -22.9% | 29 | 0-3 | -9.4% | 24 | 13.8% | 31 | 0.3% | 13 |
| 31 | HOU | -29.2% | 32 | 0-3 | -7.6% | 20 | 23.1% | 32 | 1.4% | 7 |
| 32 | OAK | -32.0% | 30 | 0-2 | -35.8% | 32 | -1.5% | 15 | 2.3% | 3 |
It's nice to see the 49ers are no longer the worst team in the NFL. You got to love DVOA if you are a fan of a team that gets glossed over on Sunday's.
I hope you're feeling better after your sick day, Aaron. And now that we have a Dave statistic, do we get a HAL?
I like that name.
What?
Also, I'm pretty sure this is not the right place for a post like this, but it's hard to believe any team could have beaten NO on Monday night.
This whole DVOA thing is a bunch of carp. You math geeks, with your numbers just don't understand REAL football. I mean, come on! Shirley the Raidahs MUST ranking and stats should have improved after their performance this weekend.
Underbus, you're not thinking clearly. ANY team could've beaten NO on Monday...
If they had ROBO-PUNTER!
NERDS!!!!
Look Dave, I can see you're really upset about this. I honestly think you ought to sit down calmly, take a stress pill, and think things over.
Denver is clearly ranked too low because your baised against teams that beat the Patriots. Ranking teams based on number of DUIs is way better than this. Dave is a stoopid name for a rating system anyway.
I'm going to go "Yay, Dave!" Because 9th is more palatable than 17th....
The Steelers are clearly ranked too low because without Roethlisburger under center they are a waaaayyyy better team. ESPN is way better than this. You guys are stoopid nerds! Batch 4eva!
For the teams that had a bye this week, and thus have played only two games, it appears that this week's formula (two parts projection, one part VOA) is applied. Shouldn't last week's formula
(80/20 split) be applied instead?
Wait a minute, if there are no opponent adjustments, then the teams with a bye in Week 3 (San Diego, Oakland, Dallas, and Kansas City) should still have the same DAVE as last week, using the 80/20 ratio, right? I mean, if 80/20 is the correct ratio to use with 2 weeks worth of results, shouldn't these 4 teams still be using 80/20, since they still only have 2 weeks of results?
Ah, somebody beat me to it. Well then, I'll ask this - Does the "pre-season" portion of the rating now have a new "QB injury" rating or whatever it's called for Tampa Bay, the way Kansas City and Oakland got one before?
I think the only thing that leaps off the screen for me regarding Dave is that he apparently has concluded that ten teams are worse than the Lions. How much does Dave drink, and does he go in drive-throughs naked?
The rating is based on "weeks ago" rather than "weeks of ratings," which is why the teams on bye change.
Cincinnati dropped from 44.7% in Week 2 to 25.3% now and won on the road. Recovering 5 of 6 fumbles sure is helpful.
RE: 9
I would like to point out that ranking teams based on number of DUIs is an objective ranking, rather than the template-required subjective ranking. :-P
I'm still amazed at how good the DAVE rating actually looks. If I was asked to rank teams right now, the only one I'd place differently is San Diego, and I can easily believe that stats know more about San Diego than I do, considering they just demolished Oakland and Tennessee.
That Baltimore game next week is going to be huge, though.
FO just keeps looking better in my eyes. Anyone who is brave enough to rank the Bills at 11 deserves a medal of some sort.
Comment 14 has a good point, so I've gone back and adjusted Tampa Bay for a rookie quarterback the rest of the season. Don't know how many more weeks I'll fiddle with the projection part of DAVE, but this change makes sense.
I assume that San Diego (which means "a whale's vagina") is going to drop once the D in DVOA kicks in. Correct?
i looked through DAVE and didn't have a WTF? moment. good job (i think).
these DVOA rankings are useless
So, what you're really saying is that DAVE is just short for DVOAAVE...
Plus, #8 stole my joke.
When do you expect to add home field adjustments to your DVOA ratings? Its alot easier to pass in the RCA dome than in the frozen tundra of Lambeau field. It would seem to me that that would have a rather large effect. Maybe not on the overall rating but teams that play in domes or in sunny Miami would have their offensive DVOA too high and their defensive DVOA too low. Without that things like the quarterback rankings are pretty meaningless.
So the Saints draft Reggie Bush and their defense improves. How did this happen?
re:18 - haha
Wait, is DAVE opponent ajusted, or should it be VAVE?
I'm also curious. Is DAVE opponent adjusted?
The 2006 performance is not, the preseason projection is. And starting next week, the 2006 performance will be opponent-adjusted also.
30: Cool. Thanks for the quick reply.
I'm surprised Minn dropped - it sure looked like they outplayed the Bears despite the loss.
Re:16 "The rating is based on “weeks ago� rather than “weeks of ratings,� which is why the teams on bye change."
But doesn't that seem incorrect? If we're not doing opponent adjustments yet, then our estimate of San Diego's quality so far should be exactly the same as it was a week ago. (And the same holds true for Oakland, Dallas, and Kansas City, obviously.) Our confidence in the pre-season projections should certainly go down a little each week, but that's only because we have more and more information about the current season each week. We shouldn't put more weight in San Diego's two games now just because it's a week later.
uh-oh.
Falcons are 2-1 and ranked 14th by DAVE.
Playin' with fire there buddy.
Here's my perspective on DAVE: It's not a better measure of the teams as we've seen them so far than the adjusted-VOA, it's a predictor of how they'll be when the year is over. So, right now, the Pittsburgh's offense is below average, and I don't just mean playing below-average, I mean the pieces as a unit are below-average right now. However, the preseason projection says that the offense won't be as bad as it has been for the whole season, so the Steelers offensive DVOA should be better at the end of the season. Similarly, DAVE gives a lower offensive DVOA to Philly because the preseason projection says that against an average schedule of defenses, the offense will not perform like they have against Houston and San Francisco. Of course, the projection could be wrong, and if the performance doesn't return to the projection, it will be reflected in DAVE as the projection becomes less important.
Is that a good way to explain the purpose of DAVE, Aaron? To combine what we know with what is expected and arrive at a happy median weighted mean?
I don't know what Aaron's stats say, but my take on the Vikings loss was that it was essentially a stand-off, much like their two wins. They lost, though,
and since DVOA (as opposed to VOA) doesn't kick in for another week, they don't get any credit for having played a top ranked team close. If they beat Buffalo on Sunday, they'll have defeated the teams rated 11 and 13 in total VOA for week three, while playing the number 3 team in total VOA for week three to a standoff, and it'll be interesting to see where 2006 DVOA places them.
It'll also be interesting to see the Vikings in relation to the Lions, in regards to DAVE as the Lions prepare to go into the Metrodome for week five, if DAVE is still being provided. If Vikings remain three slots lower than the Lions at that point, in terms of DAVE, well, I'll be looking for some DAVE supporters to put their money where their mouth is.
I do think that special teams rankings tend to be overlooked in evaluating VOA, DVOA, and DAVE, and once one considers this, moving the Vikings downward doesn't seem too off-base.
I think I can take a stab at #35. The porpuse is to combine what we know about the 2006 season with what we know about how the 2005 season ended and how each team was affected in the offseason (signing/losing free agents, coaching changes, age, experience, players returning from injury and rookies).
In the second table, shouldn't that be "Total DAVE," "Offense DAVE," "Defense DAVE," and "Special DAVE"?
#26: In PFP 2006 it is mentioned that an individual stadium-specific special teams adjustment has been made for Miami and San Francisco in addition to Denver. The adjustment is probably something that didn't become clear until analyzing all data from 1997-2005. Also, special teams has a number of elements that are almost completely opponent-independent. If you are going to look at something like whether teams pass better in the RCA Dome than Lambeau Field, you need to have a lot of data to decide that it is due to the stadium and not due to Al Harris or the Colts defense.
I have nothing to do with the Football Outsiders staff, but those are my thoughts on why stadium-adjustments only exist for special teams so far.
However, it should be possible to say that adjustments to offensive VOA due to stadium effects are small, because if it had a large effect on play above the ability of the football players, it would show up even in limited data sets, just as field goals are obviously easier to kick in Denver, which I believe you can get evidence of by comparing the field goal percentage of away team kickers in Denver compared to the league-average percentage for a single season.
Drew Brees doesn't care about Atlanta people.
If it weren't for the existing baseball statistic(s), I would suggest that DIPS might be better than DAVE. DVOA Including PreSeason. Or DIPP, I guess... DVOA Including Preseason Projections.
Of course, that seems like a suggestion better suited to last week's discussion... whoops.
Just what do you think you're doing, Dave? I'm sure we can work this out.
*Daisy....daisy......*
27: I think that it's more Gary Gibbs and the massive personnel moves on defense that improved NOLA's defense. That and going against the Falcons passing offense.
Miami's defense is clearly ranked too high because they let Kerry Collins complete 58% of his passes despite 18-30% of his passes being completions if they'd been catchable. Throwing darts is way better than this. K3RR3 C0771N5 iZ tEH $$$CUK!1!
Interesting that DAVE has NFC teams in 3 of the top 4 slots. Are we seeing a conference change this year? I don't have stats on conference vs conference records. Anyone?
nfc 4
afc 3
45--I don't think so. The only two good teams from last year in the NFC got better, and Philly came back, but out of the top 10, 7 are AFC. On the other hand, the AFC is also home to the three worst teams, according to Dave. But still, out of those 7, Denver, Pitt, Cincinatti, Indy, San Diego, and Jacksonville will continue to be good if not great teams for the foreseeable future (I omitted Baltimore because of the old QB situation). Plus you gotta figure NE will rise back to top-10 status at some point.
Actually, the rankings look to be about right.
One thing that does stand out to me is the seemingly too-low placement of the Minnesota Vikings. The vikings beat the Redskins and the Panthers in their first two games (both of these teams are ranked above the Vikings) and came within a field goal of taking the Bears into Overtime -- the Bears are the #3 ranked team. (!) I believe the Vikings deserve to be higher.
It's cool what Aaron is doing -- I'm surprised that the folks running the SportsBooks in Las Vegas haven't bought the whole web site, and put Aaron under payroll.
In the second table, shouldn’t that be “Total DAVE,� “Offense DAVE,� “Defense DAVE,� and “Special DAVE�?
Defensive DAVE was my nickname when we played street football in middle school. (long arms and cat-like reflexes)
So DAVE is designed to predict DVOA for the entire season, but what would the mix be if it were trying to predict DVOA over the next 14 weeks?
What I love is how true suckiness in one year can totally overwhelm a good performance in the other year. Well I don't love it, actually, but look at Indy's 2006 ST ranking (4) and their DAVE ranking (19). How bad was last year's team? Waaaaaaayyy bad.
Buffalo, by contrast is 2 and 2. Way to go, Bills.
Even worse, though, is Indy's 2006 D: 29 this year, it was (IIRC) 8th last year, and combined they're... 27th. That 29th only counts for 33% and the 8th counts for 67%. Meaning that 29th is monstrously bad. Not quite sure how they can turn the switch on or off like last weekend vs Jax, but maybe they better leave it on all day each Sunday. Just in case they need it.
THANKS for the insult to the Broncos.
As a Donkeys fan, I actually was very pleased this week to see Denver quite poorly in the VOA and mediocre in the DAVE rankings. It really pissed me off when a ok win over the Pats jumped the Broncos up like crazy all over other power rankings. Thanks footballoutsiders for presenting such a solid and fair view, devoid of atleast a good portion of the supernatural hype surrounding sports journalism.
So DVOA says the Lions outplayed the Packers Sunday? Detroit moves up three slots, GB down four (down one in DAVE) - and that's without adjusting for the poor opposition. I'm surprised, especially given that Favre topped all QBs in DPAR and the Pack's special teams finally showed signs of life. Can you give us a breakdown? Is it the running game?
And wow that Texans' defense is bad!
Am I the only one who thinks Denver's defensive DVOA seems low? I thought they played St. Louis very well on defense, and they only gave up the points they did because of the great field position given away by Jake Plummer. They held KC to 6 points, and New England to 7. Other than the touchdown drive by the Pats, their defense was pretty much in control of that game, I thought. And for that, their ranking drops 3 spots? Based on the three games, I can't spot anything that would lead me to believe the Broncos defense is the 18th best in football. What am I missing?
I guess DVOA's adjustment for scoring lots of points in garbage time isn't nearly as strong enough; Giants with the 8th best offense in the league...I love it, thank god for garbage time and racking up ridiculously good numbers then.
Have you guys done DVOA by run/pass for this year yet? I'm confident the Giants split is something like 5th best run defense, 33rd worst pass defense (even Georgia has a better one). It truly shows no matter how you good you are at stopping the run, it don't mean nothing if the QB has 8 seconds to spot a guy who is standing wide open 25 yards downfield.
Also helps explain the 4th quarter - teams try to run up the middle, and it's the only thing we can do well defensively.
I'm sure this has been asked before but is there a section of the site where you can see individual game DVOA scores for all teams? That would be pretty awesome :).
I agree with 54 on Denvers defence rating although I have two possible rationalities.
1) sacks, interceptions, Forced Fumbles: Denver has been tragically low on each.
2) I think that they are going to be a team that does poorly until harder teams weigh more significantly into their schedule. The mentality seems to be that a close win is a win, and a big win is a win. Thus I think that they will have some losses to bad teams (Stl etc) and won't dominate a great deal in any of their games, but should be able to actualize a fair number of wins.
Interesting looking at the top of the rankings. Philly and Indy are your greatest shows on turf, with elite offenses and defenses that are respectively mediocre and bad. Everyone else in the top 10 is doing it with defense, with only the Chargers and Bears showing much life offensively. The Bengals offensive VOA, in particular, is almost shockingly low.
Shouldn't DAVE be VAVE - given that there's no defense adjustment yet?
I'd have preferred VIP: VOA Including Projections.
Yikes, Green Bay is on course for the worst rushing DVOA of all time - a title held by last year's Cardinals. I'm guessing the formulae didn't appreciate the three fumbles, two first downs and 50-odd total yards vs Detroit. What's the worst single-game rushing DVOA total?
We got a bit of New Orleans hate mail saying we're biased towards the Falcons and Vick.
Yeah...
Even if the Vikings aren't as good as their record, and their ranking is sensible, doesn't this mean the Bears should take a hit after last week's game? The defense was still strong, but both the passing game and running game struggled... and we haven't adjusted for the Vikes D yet.
They did take a hit. The offense went from 21.5% to 13.4%, implying something like a -2.5% VOA on offense for that game.
This whole DAVE thing is a waste of time.
By next week we will have 25% of the season in the books. Then you can throw out last years numbers as an indication of this year, for sure.
I made a point BEFORE this season started that some teams, the Rams defense for one, changed over so much. Your 2006 projections, however, did a poor job of taking that into account.
Now you want to use those same lousy projections to influence the current rankings. BAD IDEA.
max #64:
Rams VOA for defense by week.
Week 1: -40.5%
Week 2: +9.3%
Week 3: -21.3%
That's a lot of variance there. It wouldn't be surprising to see the Rams play closer to their +10.6% projection in many more games, as they did in Week 2.
Do the projections for the Saints take into account that they had nine road games and that for the seven home games they probably had a smaller home field advantage?
Re #56
Yes. It's called PFP 2007, to be released next summer. The current edition, PFP 2006, contains offensive and defensive DVOA for each team for each game in the 2005 season.
It would be surprising to see the Rams play closer to their +10.6% projection. You see, they now have Haslett and Venturi coaching the defense. Bad move. Watch the season progress and notice that the middle of the defense takes itself out of the play. Watch Gore's run in week 2 to see the MLB run himself right out of the play. Watch the tackle's techhnique disappear. (True story: after being chewed out for letting the O-line get under them on short yardage plays, the Saints' tackles once actually dove onto the ground on a goal line play in order to get under the O-line).
Saints D will be much improved without H and V, as seen on Monday night. And what about the improvement on the Saints O, get up to the line, run the motions and snap the ball crisply, haven't seen that in years. Oh, and collapse the whole side of the defense for a running play, nice.
68:
For defense, negative values are good, positive are bad. So unless I am misinterpreting your post, I think you should be saying that it would NOT be surprising to see the Rams play closer to their +10.6% projection.
I don't really know what it means (if anything), but here are the 3 teams with the most positive difference between VOA and DAVE and the 3 teams with most negative difference between VOA and DAVE. You may be able to think of it as the teams playing the most above / below expectation (but I'm not really sure about that):
Total
1. NO ... 43.60%
2. BAL ... 43.40%
3. SD ... 41.10%
30. TB ... -23.10%
31. TEN ... -32.30%
32. OAK ... -44.50%
Offense
1. NYJ ... 25.90%
2. PHI ... 22.60%
3. SF ... 15.90%
30. SEA ... -22.20%
31. PIT ... -23.40%
32. OAK ... -37.90%
Defense
1. BAL ... -36.10%
2. NO ... -29.30%
3. SD ... -48.90%
30. IND ... 10.70%
31. NYG ... 12.30%
32. HOU ... 28.10%
Special Teams
1. CHI ... 9.20%
2. SF ... 5.50%
3. IND ... 5.30%
30. DET ... -5.60%
31. DAL ... -6.00%
32. ATL ... -11.80%
48 -
The current numbers don't reflect opponent adjustment yet. The Vikings low ranking is really in large part of them not having really dominated any team just yet. They'll probably get a couple of spots bump when opponents are added in, depending on how they do the next few games.
It just currently considers a 3 point win over the Panthers about as meaningful as a 3 point win over the Texans. When its adjusted, it won't.
Wow, I should learn to proof-read.
Defense
3. SD ... -29.00%
32. HOU ... 29.10%
With San Diego, I read the Defensive VOA column instead of the difference column.
69, you're right. Point obviously is that the paid pundits who think Haslett is going to be an improvement are wrong.
The DAVE ranking looks pretty solid to me. Personally, I would know Seattle down a few spots (to like 5th), I'd flip Washington (to 17th) and Dallas (to 11th), and I'd bump Minny up a few spots (to like the 20-18 range). But that's really just nip-picking. It's nice to see Oakland has finally taken their rightful place.
Re: 61
People complaining about NO being ranked 19th are morons. I bet if you had told Saints fans before the season that their team would be solidly middle-of-the-pack this year, they would have been ecstatic.
Wait, people thought Haslett was going to be an improvement? Didn't they see his performance in New Orleands. I guess it falls under the peter principal where a failed head coach makes a good coordinator.
*Personally, I would move Seattle down...*
I suck at being coherent today.
Just to let everyone know, my ESPN News bit was cancelled today because of the T.O. situation, so I'll try to take the extra time to put together a mailbag answering some of the DVOA questions from the past couple of weeks.
Will the defense stats be broken down into passing and rushing at some point?
They are. They're on the "TEAM OFFENSE/TEAM DEFENSE" pages.
Re: 34
Nah, when you have got the worst special teams in the league, it's hard to be any better than 15th in the league overall. FGs, FG protection, and punt protection have been miserable so far. Actually, kickoff coverage and kickoff returns have been pretty good, but not so good as to offset the abject failures in the other parts of the game.
Re 64:
Max, I think the point of DAVE is that, while specific teams may have seen a lot of turnover and so the projections for them will be bad, they do a better job for MOST teams than just looking at VOA from games 1-3. I think the preseason projections missed horribly on one or two teams last year, but they did better on most teams than most prognosticators. If you're right, and they do a lousy job this year, Aaron will re-tune the formula to make them better, or de-emphasize them. That's the beauty of statistics--the more information you have, the better your predictors become, usually. But there will always be outliers.
Re: 21, isn't there an argument for a reverse adjustment for the Jets? The projected DVOA presumably was influenced heavily by the offensive performance last year under wee Brooks and the expectation of damaged goods at QB this year. I would be suprised if the system would project the Jets to be 31st in DVOA with close to a pre-injury Pennington at QB.
The projected DVOA presumably was influenced heavily by the offensive performance last year under wee Brooks and the expectation of damaged goods at QB this year
No, not really. Pennington was used as the starting QB for the Jets. He does have an injury history, though, which can't be ignored: while they're playing well now, they've got worse odds for having a healthy QB throughout the season than the average team.
Why is DAVE so down on Arizona's special teams? They ranked in the high 20s last year, 22nd through three games this year.....
This isn't a criticism or a quibble; I'm just interested to know what it is that the numbers don't like.
The only thing that kept them afloat last year was Rackers' godlike FG performance. Most likely it's just regression to the mean - projecting Rackers to be 'average Rackers' rather than 'godlike Rackers' would make the Cardinals special teams god-awful.
Their kickoff/punt coverage and kick returns and punt returns were bad last year. Looks pretty similar this year, although the kickoff coverage looks better. So far.
Oakland is clearly ranked too low because weve won 3 titles. Everyone's ranking system is way better than this. ALL YOU SMUCKS FEAR TEH RAIDERS!!!!!!
Oakland is clearly ranked too high because OSU, USC, and Auburn are clearly better.
#87 wins. I don't know what you win, but if there was a prize, you would win it.
Yeah! I win a big fat Box-O-Nothin', hurray for me!
Any word on the FOX rankings w/ comments? Or have those been TO'd as well?
So are the DAVE rankings going to be the FOX rankings all year or will DVOA be the main ranking once the D kicks in next week?
I'm sorry. DAVE? I'm afraid I can't do that.
At first I thought Aaron meant it was my fault ... but I've only charted half of one Saints' game this season ...
I've got a general question, when DVOA becomes available, will the DVOA numbers for the first 3 weeks be retro-adjusted? And will those numbers be available somewhere? I forget how it usually works.
Has anyone noticed that the Eagles are #1 in Rushing Offense?
I'm an Eagles fan, but I'm not crazy. I watched that Giants game. #1 in rushing? Not by a long shot.
#1 in stiff-arming weak SF safeties, maybe.
Pat (and generally), I understand the probability of Rackers regressing to the mean. But in that case, why don't the "god-awful" coverage and return teams regress towards the mean, too? Is there a significant difference in the (downward) performance of a field-goal kicker from one season to the next, vis-Ã -vis a kick return unit (which is, in theory, likely to undergo personnel changes)?
Late response to the response to my comment about field adjustments.
Even if it would be hard to get specific field adjustments, you could have adjustments for the type of field and playing conditions:
Indoors
Outside no precipitation/wind
Outside rain/low wind
Outside no precipitation/high winds
Outside snow/high wind
etc.
95: Racker's performance in 2005 was considerabally better than it was in any other year of his carreer, and thus likely to decline. The performance of the Cardinals coverage teams wasn't considerabally worse than other years, and thus not likely to improve.
I’m an Eagles fan, but I’m not crazy. I watched that Giants game. #1 in rushing? Not by a long shot.
They only played 1/3 of the season against the Giants. The other 2/3 of the season so far, they've rushed extremely well. However, those were against defenses that are going to be very bad against the run. The Giants, however, are very likely going to be one of the best in the league against the run.
For comparison, in the Seahawks versus the Giants, Shaun Alexander rushed 20 times for 47 yards that game, and Morris 15 times for 18 yards. They ended up just barely over 100 yards total for the team.
No one runs well against the Giants.
Re: 98
And as someone pointed out earlier (I'm too lazy to try to track down the specific post to give due credit), that may be one of the reasons for NY's 4th quarter proficiency. By that time, teams are trying to run out the clock and that plays into the Giants' largest strength.
Is anyone else starting to freak out because there's no commentary yet? I think I'm addicted. Can we start a support group?
I'm procrastinating, wondering why the 'Sports Bet' ad uses a photo of Brady Poppinga for its 'Who'll score next' tagline (answer: whichever tight end Poppinga's meant to be covering?), posting comments about the Cardinals' special teams for chrissakes....
I'll join the support group.
milo,
The Rams DC last 2 years was Larry Marmie. That guy was totally incompetent. How did Haslett do in Pittsburgh as a DC? That's the real comparison.
I am with Zip at #100. Sickdays are no excuse...
#103
I'm not criticizing Aaron, he can take all the sick days he wants. But I reserve the right to freak out when FO content is late, because I'm dependent like that.
It isn't that I am criticizing him, I just know that when I am out on a sick day, the things that are due are still due.
it's interesting to see how the nfc west is kinda close, with seattle and st. louis almost interchangable in the raw numbers. i had thought that many were overstating the lack of challenge to the seahawks from the teams therein, but i think the divisional games could end up being pretty tight later in the year, maybe with the rams sneaking away with the divisional title.
The new name makes me think of when Jim from the Office (in the commentary) says that he believes that DVD stands for the guy who invented it, "Digital Video Dave".
Actually, the editing of the commentary over at FOX was delayed by all the T.O. michigas. It's finally up here:
http://msn.foxsports.com/nfl/story/6006820
Meanwhile, all team stats pages, individual stats pages, and -- for the first time in 2006 -- adjusted line yards stats are now updated.
Speaking of delays, Scramble will be delayed until tomorrow.
And this is me, telling you with words, thanks for getting it done for us all.
I USED THE DAVE WEEK 3 RATINGS ON MY WEEKLY FOOTBALL PICKS AND ONLY MISSED 2OF16, JETS OVER BILLS ,AND NEW ORLEANS OVER ATLANTA WERE THE MISSED PICKS PREVIOUSLY I USED ESPNS EXPERT PICK AND POWER RANKINGS AND MISSED 4 FLUKE ? MAYBE STAY TUNED.
Aaron, please pass the word to the Fox people that they have the wrong logo for the Ravens.
98, 99: Thank you both for responding. I always enjoy what you add to these threads.
I'm still stinging from that Eagles collapse against NY. Our beloved Andy is one of the least likely to run late in a game to consolidate a lead (last in the league in 2005 at 43% according to PFP).
Here's why I'm still surprised by the #1 ranking. I don't expect it to hold up. I've always felt that the Eagles running game is a counter-punch, opportunistically called. It's not a traditional #1 running game, to be sure.
Using the available stats at NFL.com that includes McNabb's 5 kneel-downs and 6 scrambles, the Eagles are tied for 20th in rushing attempts per game with Indianapolis. (They're at best 21st in attempts when you remove Manning's 3 kneel-downs and 2 rushing plays).
DVOA will most likely straighten things out for us next week. 'Cause here's what my birds did over the last 3 weeks on the ground:
Week 1 (vs. HST) 26-123 / 4.73 (QB-adjusted)
Week 2 (vs. NYG) 26-76 / 2.92 (QB-adjusted)
Week 3 (vs. SF) 14-149 / 10.64 (QB-adjusted)
In all cases they had 2nd half leads:
Week 1 (17 pass, 19 run)
Week 2 (14 pass, 14 run)
Week 3 (10 pass, 7 run)
Andy's at least willing to call the run almost half the time now when leading in the 2nd half.
Maybe their week 2 rushing performance against the great Giants run defense was a good performance. It still felt like it cost them the game, and that's going to sting for a while.
I'm just happy to have McNabb back being McNabb. But our last pre-season game is this Monday night.
...and though the withdrawal symptoms hit a bit of a whining note there, we still appreciate the hard work.
And truly without a hint of sarcasm: I love the rare moments around here when - FOX and ESPN deals aside - personal stuff happens that reminds me of the first months, when article deadlines were at the mercy of Mirinae Fays sleeping patterns.
Should we tell Brad there were only 14 games this week, or would that spoil his mood?
Thanks are in order for the whole site, and the whole concept, sure you're getting paid for it, but instead of paying FO for something wonderful like the DVOA and the wonderful commentary you guys provide.