DVOA Analysis
Football Outsiders' revolutionary metrics that break down every single play of the NFL season

Week 3 DVOA Ratings

by Aaron Schatz

It takes some getting used to seeing the Arizona Cardinals on the top of the Football Outsiders DVOA ratings, but there they are again this week after stomping San Francisco by a final score of 47-7. In the entire history of DVOA ratings, going all the way back to 1989, Arizona had never sat at No. 1 after a single week until last week. Now they have two weeks at No. 1, and they may be there for a while.

(Note: Sorry I didn't notice this last week, being No. 1 for the first time is a pretty big deal. The only teams that still have never been No. 1 in DVOA, not even for a single week, going back to 1989: Carolina, Detroit, Minnesota, and expansion Cleveland, though the original Browns were No. 1 a few times.)

Arizona is dominating in every way you can imagine, and certainly making our negative preseason forecast look pretty silly. The Cardinals rank third in offense, third in defense, and fifth in special teams. The Cardinals have put up a single-game rating over 60% in all three of their games so far. The Cardinals are one of only five teams to ever have a DVOA rating over 75% after three games. The others were the 1996 Packers, the 2007 Patriots, the 1991 Redskins, and the 2007 Steelers. That's three of the best teams in NFL history, and a team that faded down the stretch and then lost at home in the wild-card round. Fun ironies: the offensive coordinator of that Steelers team was Bruce Arians, and the opponent that handed the 2007 Steelers their first loss in Week 4 was... the Arizona Cardinals.

Of course, a lot of people have been asking if the Cardinals are truly "for real" given the quality of the teams they've played so far this year. On one hand, we know that big wins are a much better indicator of a great team than a string of close wins. On the other hand, opponent quality does matter. It's impressive to destroy a terrible team, but even more impressive to destroy an average team or, better yet, a quality rival. The three teams Arizona has beaten this year are a combined 1-5 in their other games. Those teams currently rank 29th, 31st, and 32nd in DVOA, and not just because of losing to the Cardinals.

Next week, we'll start slowly filtering in our opponent adjustments, which will gradually drop Arizona's rating unless their first three opponents turn out to be better than it seems right now. But there's a way to see the effect of Arizona's easy schedule before we put in the standard opponent adjustments after Week 4.

As you know, we have not only our DVOA ratings this early in the season but also our DAVE ratings, which combine our preseason projection with current DVOA to get a more accurate picture of how good we think teams truly are. Right now, the preseason projection makes up 60 percent of DAVE. What happens if we adjust the first three games of the season based not on the actual ratings for each team so far, but instead based on the DAVE ratings? That hopefully will give us the most accurate measurement of how well teams have played in the first three weeks, since we're considering their opponents based on both play so far and what we knew going into the season. (Since we don't do opponent adjustments in special teams, right now, the special teams ratings with this method will be the same as the regular special teams DVOA through three weeks.)

With these new "DAVE-adjusted" ratings, Arizona is still the No. 1 team in the league so far. However, the Cardinals have dropped from 76.4% to 59.1%. That's a rating that fits the best team in the league through three weeks, but doesn't rank them as one of the greatest teams in NFL history through three weeks. With this drop, the gap between the Cardinals and New England/Green Bay is much smaller. (Also, Green Bay is No. 2 in standard DVOA, but New England becomes No. 2 with this method because of how well Buffalo has played in its other two games.)

Arizona has the biggest gap between their actual "no schedule adjustments yet DVOA" and their "DAVE-adjusted DVOA," but not the only big gap. Tennessee falls through the floor because their first three opponents have also gone 1-5 in their other games. The Titans have played Tampa Bay, Cleveland, and Indianapolis. They drop from 11th so far in actual DVOA to 20th in these "DAVE-adjusted" ratings. Carolina also falls: it's only a three-spot drop from No. 8 to No. 11, but a much bigger gap in the actual rating which drops from 20.3% to 6.3%.

Which teams improve the most if we adjust their early performance for DAVE ratings of their first three opponents? Well, Chicago essentially is the opposite of Arizona; as bad as the Bears have been, they've played Green Bay, Arizona, and Seattle. They go from the worst team in the NFL by far to the worst team in the NFL by a much smaller amount. Other teams that have played a particularly tough schedule so far according to DAVE ratings include Kansas City, Jacksonville, San Francisco, and Baltimore.

The full "DAVE-adjusted DVOA" ratings appear in a second table at the bottom of the page, after the standard DVOA ratings table.

* * * * *

Last week, we made a change in our playoff odds simulation to account for the injury that will keep Tony Romo out for roughly half the season. This week, we've further changed the simulation to account similarly for the injury to Pittsburgh quarterback Ben Roethlisberger. As with Dallas, Pittsburgh was run through the playoff odds simulation with two different DAVE ratings. A rating that accounts for Michael Vick as the Pittsburgh quarterback is used in Weeks 4-7, and then in half of the simulations for Weeks 8-9. Roethlisberger is back with the full higher Pittsburgh rating as of Week 10.

We're sort of playing it by ear in the way we're accouting for these injuries in the simulation, as it wasn't originally written to do this and we haven't tested this method with similar injuries of the past. As each week goes by, the Dallas and Pittsburgh DVOA ratings will reflect what those teams are like with Brandon Weeden and Vick at quarterback rather than how good they are with Romo and Roethlisberger. This week's simulation may be slightly overrating the Cowboys and Steelers, because the Romo/Roethlisberger DAVE ratings work off the Weeden/Vick DAVE ratings rather than the other way around, and the Weeden/Vick DAVE ratings are partly based on how well the Cowboys and Steelers played in Weeks 1-2 with their original starting quarterbacks. We'll play with the method in future weeks to try to get the simulation as accurate as possible.

* * * * *

Once again in 2015, we have teamed up with EA Sports to bring Football Outsiders-branded player content to Madden 16 Ultimate Team. Each week, we'll be picking out a handful of players who starred in that week's games. Some of them will be well-known players who stood out in DVOA and DYAR. Others will be under-the-radar players who only stood out with advanced stats. We'll announce the players each Tuesday in the DVOA commentary article, and the players will be available in Madden Ultimate Team packs the following weekend. We will also tweet out images of these players from the @fboutsiders Twitter account on most Fridays. One player each week will only be available for 24 hours from the point these players enter packs on Friday.

The Football Outsiders stars for Week 3 are:

  • TE Greg Olsen, CAR (24-HOUR HERO): Led all tight ends with 60 DYAR in Week 3 (8-for-11, 134 yards, 2 TD).
  • RE Mike Daniels, GB: 1.5 sacks, 2 hurries, and run tackle for a loss.
  • LG Gabe Jackson, OAK: No sacks, hurries, or QB hits allowed; Oakland RB had 15 carries for 123 yards running left with 53 percent success rate.
  • RB Karlos Williams, BUF: 50 rushing DYAR, fourth among running backs in Week 3 (12 carries, 110 yards, TD). No. 1 RB in rushing DYAR through three weeks despite only 24 carries.
  • ROLB K.J. Wright, SEA: 10 total tackles including 4 that prevented third-down conversions.

* * * * *

[ad placeholder 3]

All stats pages are now updated through Week 3 of 2015 or will be in the next few minutes.

Some notes on the schedule for the next couple weeks. This is the week for the midseason update of our KUBIAK fantasy football projections. It's a project that takes a ton of man-hours to put together, and the fact that my computer seems to slowed to a crawl in the last few days isn't going to help things. It should be released on Friday afternoon. I'll be working hard to get it out as soon as possible. I know that means you can't use it for trades and waiver pick-ups this week, but we've never been able to do the necessary work that would automate this further.

Next week's update of the FO stats pages and posting of DVOA commentary is going to be even later than usual because of unavoidable scheduling conflicts. For all you Madden fans, we'll see if I can get a posting up earlier that will announce the Football Outsiders stars for Ultimate Team, but no promises. After next week, though, I'll be back to trying to get this all up by 6pm Eastern each week. Clearly not often succeeding, but at least trying.

I also will likely be taking next Tuesday off from my weekly appearances on the ESPN Fantasy Football Weekly Podcast, but I'll be back after Week 5 and for the rest of the season after that. I hope everyone is listening and enjoying the new podcasts this season. Remember you can get links to all the FO-related podcasts in our new podcasts section, which you can find at this link.

* * * * *

[ad placeholder 4]

These are the Football Outsiders team efficiency ratings through three weeks of 2015, 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 DVOA 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) and week of season. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

Please note that there are no opponent adjustments in DVOA until after Week 4. (It's still listed as DVOA instead of VOA because I don't feel like going through and changing all the tables manually.) In addition, our second weekly table which includes schedule strength, variation, and Estimated Wins will appear beginning after Week 4.

DAVE is a formula which combines our preseason projection with current DVOA to get a more accurate forecast of how a team will play the rest of the season. Right now, the preseason projection makes up 60 percent of DAVE. (This is a slight change from previous years, when the preseason projection made up 55 percent of DAVE after Week 3.)

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>

TEAM TOTAL
DVOA
LAST
WEEK
TOTAL
DAVE
RANK W-L OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
S.T.
DVOA
S.T.
RANK
1 ARI 76.4% 1 27.7% 2 3-0 37.8% 3 -32.3% 3 6.3% 5
2 GB 51.0% 5 25.7% 3 3-0 42.8% 1 -3.2% 9 4.9% 6
3 NE 47.2% 6 30.3% 1 3-0 42.0% 2 1.6% 15 6.9% 4
4 CIN 36.0% 4 18.2% 4 3-0 26.5% 5 -11.8% 7 -2.3% 22
5 BUF 28.8% 14 8.2% 8 2-1 26.5% 6 0.1% 12 2.4% 9
6 PIT 24.8% 3 11.6% 7 2-1 32.8% 4 3.7% 19 -4.2% 26
7 DEN 23.0% 7 18.1% 5 3-0 -24.3% 31 -39.6% 1 7.6% 3
8 CAR 20.3% 8 6.1% 11 3-0 10.7% 9 -21.0% 4 -11.3% 30
9 NYJ 18.5% 2 7.5% 10 2-1 -11.9% 22 -34.5% 2 -4.0% 24
10 ATL 14.7% 10 7.7% 9 3-0 23.2% 7 9.8% 25 1.3% 14
11 TEN 11.2% 9 -4.1% 21 1-2 4.3% 11 -16.3% 6 -9.4% 29
12 SEA 8.3% 23 17.3% 6 1-2 -4.1% 16 3.4% 18 15.8% 1
13 NYG 5.5% 18 0.3% 15 1-2 6.9% 10 3.2% 17 1.8% 12
14 OAK 5.1% 25 -4.0% 20 2-1 12.8% 8 11.4% 28 3.7% 7
15 DAL 0.9% 11 -4.3% 22 2-1 4.1% 12 3.8% 21 0.6% 17
16 MIN 0.7% 24 3.1% 13 2-1 3.0% 13 3.8% 20 1.6% 13
TEAM TOTAL
DVOA
LAST
WEEK
TOTAL
DAVE
RANK W-L OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
S.T.
DVOA
S.T.
RANK
17 STL -0.5% 16 4.6% 12 1-2 -11.8% 21 -8.1% 8 3.2% 8
18 PHI -2.5% 26 1.7% 14 1-2 -18.3% 27 -17.4% 5 -1.6% 21
19 WAS -3.8% 13 -13.3% 25 1-2 -4.0% 15 -0.9% 11 -0.7% 18
20 BAL -9.8% 17 -1.0% 16 0-3 -4.8% 17 7.3% 24 2.3% 11
21 KC -13.9% 12 -2.3% 17 1-2 -13.7% 23 2.5% 16 2.4% 10
22 IND -14.8% 29 -2.7% 18 1-2 -8.7% 19 1.5% 14 -4.6% 27
23 DET -18.4% 22 -5.0% 23 0-3 -14.9% 24 4.8% 22 1.3% 15
24 SD -18.5% 19 -3.8% 19 1-2 -3.5% 14 7.2% 23 -7.8% 28
25 CLE -23.1% 21 -15.2% 27 1-2 -22.9% 30 10.8% 26 10.6% 2
26 MIA -27.9% 15 -13.3% 26 1-2 -17.3% 26 11.3% 27 0.7% 16
27 JAC -29.3% 20 -21.4% 29 1-2 -11.2% 20 15.7% 29 -2.4% 23
28 HOU -30.9% 28 -15.5% 28 1-2 -17.0% 25 1.2% 13 -12.6% 32
29 NO -32.4% 31 -12.7% 24 0-3 -4.9% 18 26.4% 31 -1.2% 20
30 TB -37.5% 30 -23.9% 31 1-2 -38.4% 32 -2.1% 10 -1.1% 19
31 SF -51.4% 27 -22.7% 30 1-2 -21.9% 29 25.3% 30 -4.1% 25
32 CHI -70.2% 32 -32.8% 32 0-3 -21.6% 28 36.0% 32 -12.6% 31

 

A second table below presents each team's total 2015 DVOA with the single-game offensive and defensive DVOA for each game adjusted based on the DAVE rating of the opponent. Special teams ratings are not changed.

TEAM DAVE-ADJ
DVOA
ACTUAL
DVOA
RANK W-L DAVE-ADJ
OFFENSE
OFF.
RANK
DAVE-ADJ
DEFENSE
DEF.
RANK
S.T.
DVOA
S.T.
RANK
1 ARI 59.1% 76.4% 1 3-0 26.0% 4 -26.8% 3 6.3% 5
2 NE 48.4% 47.2% 3 3-0 38.3% 2 -3.2% 10 6.9% 4
3 GB 45.2% 51.0% 2 3-0 39.2% 1 -1.1% 13 4.9% 6
4 BUF 32.8% 28.8% 5 2-1 25.0% 5 -5.4% 8 2.4% 9
5 CIN 32.6% 36.0% 4 3-0 22.9% 7 -12.0% 6 -2.3% 22
6 PIT 27.1% 24.8% 6 2-1 32.9% 3 1.6% 16 -4.2% 26
7 DEN 17.1% 23.0% 7 3-0 -26.4% 31 -35.9% 1 7.6% 3
8 ATL 14.3% 14.7% 10 3-0 23.2% 6 10.1% 26 1.3% 14
9 SEA 13.5% 8.3% 12 1-2 -5.1% 19 -2.9% 11 15.8% 1
10 NYJ 11.1% 18.5% 9 2-1 -13.1% 24 -28.3% 2 -4.0% 24
11 CAR 6.3% 20.3% 8 3-0 3.1% 9 -14.6% 5 -11.3% 30
12 OAK 5.0% 5.1% 14 2-1 13.6% 8 12.2% 27 3.7% 7
13 DAL 4.0% 0.9% 15 2-1 1.4% 13 -1.9% 12 0.6% 17
14 NYG 2.3% 5.5% 13 1-2 2.6% 11 2.2% 17 1.8% 12
15 STL 1.5% -0.5% 17 1-2 -13.0% 23 -11.3% 7 3.2% 8
16 PHI -0.3% -2.5% 18 1-2 -16.5% 26 -17.9% 4 -1.6% 21
TEAM DAVE-ADJ
DVOA
ACTUAL
DVOA
RANK W-L DAVE-ADJ
OFFENSE
OFF.
RANK
DAVE-ADJ
DEFENSE
DEF.
RANK
S.T.
DVOA
S.T.
RANK
17 BAL -0.6% -9.8% 20 0-3 0.4% 14 3.3% 19 2.3% 11
18 KC -3.9% -13.9% 21 1-2 -6.1% 20 0.2% 15 2.4% 10
19 WAS -5.1% -3.8% 19 1-2 -4.5% 18 0.0% 14 -0.7% 18
20 TEN -5.1% 11.2% 11 1-2 0.2% 15 -4.0% 9 -9.4% 29
21 MIN -8.4% 0.7% 16 2-1 -3.1% 17 6.9% 24 1.6% 13
22 IND -9.7% -14.8% 22 1-2 1.5% 12 6.5% 23 -4.6% 27
23 DET -10.8% -18.4% 23 0-3 -7.8% 21 4.3% 21 1.3% 15
24 SD -13.3% -18.5% 24 1-2 -2.1% 16 3.4% 20 -7.8% 28
25 JAC -18.9% -29.3% 27 1-2 -10.8% 22 5.7% 22 -2.4% 23
26 CLE -21.8% -23.1% 25 1-2 -17.6% 28 14.9% 29 10.6% 2
27 NO -26.5% -32.4% 29 0-3 3.0% 10 28.3% 32 -1.2% 20
28 HOU -35.4% -30.9% 28 1-2 -14.3% 25 8.5% 25 -12.6% 32
29 MIA -35.4% -27.9% 26 1-2 -18.1% 29 18.0% 30 0.7% 16
30 SF -37.0% -51.4% 31 1-2 -19.2% 30 13.7% 28 -4.1% 25
31 TB -46.1% -37.5% 30 1-2 -42.4% 32 2.6% 18 -1.1% 19
32 CHI -51.7% -70.2% 32 0-3 -17.1% 27 22.0% 31 -12.6% 31

Comments

364 comments, Last at 06 Oct 2015, 4:42am

1 Re: Week 3 DVOA Ratings

Whoa - Eagles with a massive jump in defensive DVOA, from 12th to 5th in a single week. Lots of turnovers will do that! The NFC East is so wide open - all four teams bunched 13th through 19th in DVOA.

12 Re: Week 3 DVOA Ratings

Ha - all the Eagles have to do is making incredible plays on deflected passes and baiting the opposing WR's into surprise laterals to nobody and BAM! Top 3 defense.

27 Re: Week 3 DVOA Ratings

Man, they were SO conservative on offense. I know they had a bunch of injuries at the skill positions, but I'd be worried about what Bowles is trying to accomplish with that. They really didn't seem to have any intention of scoring early on, just running as many low risk plays as they could before punting...

67 Re: Week 3 DVOA Ratings

Did not get to watch the game, but some of that is Fitzpatrick's inability to accurately throw downfield. That pass to Devin Smith that got picked off should have been a touchdown.

2 Re: Week 3 DVOA Ratings

The Cardinals end their season as follows. Three of the games are at home, but there are no Bears or Niners in that mix. That Packers game might be for the #1 team in the NFC.

Thu, Dec 10 vs Vikings 7:25 PM
Sun, Dec 20 @ Eagles 12:00 PM
Sun, Dec 27 vs Packers 3:25 PM
Sun, Jan 3 vs Seahawks 3:25 PM

3 Re: Week 3 DVOA Ratings

Man - that's a fun schedule. All four of those teams still have legit playoff aspirations. I suspect it will be the make or break game for entire season for the Eagles, possibly for the Vikings, too...

119 Re: Week 3 DVOA Ratings

As much as I'd love to see the Vikes in the playoff hunt that far into the season, their schedule says 'nope'.

@Den, KC, @DET, @CHI, STL, @OAK, GB, @ATL, SEA, @ARI, CHI, NYG, @GB.

They need go 3-1 over the next four. That'd get them to 6-2. But after that Chicago game? Brutal. They'll win no more than 4 of those. So they have to go 3-1 over the next 4 games to have a chance at 10-6. I think that's a tall ask, unless the O-line and QB show some significant improvement.

125 Re: Week 3 DVOA Ratings

As soon as I saw the schedule this year, I knew there was way too much optimism about their chances. Even one of their weak opponents came in a brutal spot, opening game on the West Coast, late Monday Night kickoff.

Their only chance is to be home field defensive beasts, including against Rodgers, when he'll have a harder time doing his hard count mischief, and steal two or three on the road, with Peterson replaying 2012. It isn't an impossible task this weekend, and if they are ever going to win in Chicago, this is the year. They have won a lot of games in Detroit over the years, even when the Lions haven't been dregs. Winning this Sunday would be huge for their chances, in that it might give them some margin for error at home.

138 Re: Week 3 DVOA Ratings

Yeah - they're still a year or two away. I like what the current regime is doing as far as the talent on the field and the scheme, but they're still missing a few parts (starting with that line; in fact, I'd draft for both lines).

203 Re: Week 3 DVOA Ratings

If they think Bridgewater is their guy, then they must, above all else, adopt the First Commandment from the Book of St. Joseph of Gibbs, "Thy quarterback shall be made to feel comfortable in thy pocket, while holding thy ball, and peering down thy field. Forever and ever, Amen."

346 Re: Week 3 DVOA Ratings

Ryan Grigson disagrees. The recipe team for success is:
1. Franchise QB
2. Below replacement guys for the O-Line
3. Wrs with lots of speed and little experience
4. ?
5. PROFIT!

170 Re: Week 3 DVOA Ratings

"As much as I'd love to see the Vikes in the playoff hunt that far into the season, their schedule says 'nope'."

The last time that got said about the Vikings schedule was in 2012, when their post-bye schedule consisted of: at Chicago (7-2), at Green Bay (6-3), Chicago, at St Louis (3-5-1), at Houston (8-1), Green Bay. They ended up going 4-2 and made the playoffs over Chicago. Peterson willed them in that year, and it looks like he's back to his dominating ways again, although another MVP season is probably a bit too much to ask for.

6 Re: Week 3 DVOA Ratings

Hate that the Vikings had a late night West Coast opener, and a late season Thursday night road game against a quality opponent. Actually, I generally hate Thursday night games.

53 Re: Week 3 DVOA Ratings

Thursday night, Sunday night, Monday night...I hate 'em all. Give me a noon kickoff for every single game of the year and I'm a happy guy. I liked the primetime games a lot more back when I was in college and didn't have to worry about getting up early.

56 Re: Week 3 DVOA Ratings

I've always been blessed, or cursed, to be someone who feels great on 5 hours sleep, so that's never been an issue. I hate the Thursday night games because the quality of play tends to be so uneven.

140 Re: Week 3 DVOA Ratings

The Thursday games just ruin the flow of the NFL week. It sucks from a gambling and fantasy aspect. It sucks that 2 teams played 3 days before everybody else. I just hate it. Probably my least favorite change the NFL has made since I began watching in the early 80s.

154 Re: Week 3 DVOA Ratings

I don't gamble or play fantasy football... but the games themselves are notoriously awful, too - there really is zero upside for fans, isn't there?

196 Re: Week 3 DVOA Ratings

I'm wondering if the league has done a better job than they expected on loading up at the end of the season. This is the Packers last five games.

Thurs Dec 3 @ Lions 7:25 PM
Sun, Dec 13 vs Cowboys 3:25 PM
Sun, Dec 20 @ Raiders 3:05 PM
Sun, Dec 27 @ Cardinals 3:25 PM
Sun, Jan 3 vs Vikings 12:00 PM

The Cowboys should have both Romo and Bryant back by then. The Raiders are no picnic anymore and having back-to-back West Coast road trips at Christmas sucks. (Would McCarthy keep the team on the left coast over Christmas?) The Packers could be playing the Vikings when both teams need a win (e.g., Packers for playoff position and Vikings for wild card) so they could move the game to Sunday night.

212 Re: Week 3 DVOA Ratings

McCarthy was quoted "We will be home for Christmas." But that may still be that they are set-up for the whole week out there, but give enough off time for everyone to get home for Christmas. Considering that a lot of players travel to where extended family lives, which isn't in WI, it's not that different.

5 Re: Week 3 DVOA Ratings

I like the Dave Adjusted DVOA concept. Dave is still your best theory of how well a team will do. But the new stat is better than straight VOA for assessing how well they have played so far. And as Dave converges to DVOA over the next six weeks, the adjusted DVOA will approach DVOA, too. It should all work together nicely.

11 Re: Week 3 DVOA Ratings

DVOA is loving Derrick Carr, too - 6th in DVOA, above Matt Ryan, Romo & Eli as well as early season success stories like Mariota and Tyrod Taylor.

343 Re: Week 3 DVOA Ratings

Well, teams do not have a lot of film on Amari Cooper and the new and improved Derek Carr yet. When they get enough film, I think there will be a correction on the offense DVOA coming. Still, it is a good sign that we are forcing opponents to adjust. Some of the increase will remain.

Now, if the defense could get closer to a zero DVOA, that would really help. I guess next year's project is the secondary.

9 Re: Week 3 DVOA Ratings

I'm more convinced than ever that the Vikings spent the afternoon before their first game at medical marajuana farmer's market.

13 Re: Week 3 DVOA Ratings

I've actually been processing DVOA/DAVE for opponent adjustments in a slightly more complicated way. I get an averaged value for DVOA and DAVE for each team (right now the ratio is about 50/50), then use those values to adjust each other in an iterative fashion.

I apply an opponent adjustment of 25% strength to each value to get corrected values, then I take the original values again but re-correct them with the corrected values, and then take the original values again and re-correct with corrected-corrected values, and so on (after 4 iterations or so it seems to converge enough to be trustworthy.)

I've also been correcting raw DVOA slightly for home advantage before doing all of that (using rules of thumb that home is worth about 2.7 points and 5% DVOA is worth about one point.)

I didn't know much about how VOA was processed into DVOA before I started doing this on my own due to wanting opponent adjusted DVOA before adjustments were usually done. I'm actually a little curious if you generally just adjust for opponents once through, or use an iterative system.

14 Re: Week 3 DVOA Ratings

I have watched each of Denver's games. The eye test says of their offense, "This is quite bad." The numbers here say, "This is the dregs of the league." Is this the most effective 31st-ranked offense in the history of DVOA?

I would be very interested to see the quarter or half splits for the Green Bay defense, and their splits for broken and non-broken plays, if that is even possible.

18 Re: Week 3 DVOA Ratings

Having the first-ranked defense is helping Denver quite a bit.

I can't see this as sustainable for them, though each game has looked incrementally less bad.

29 Re: Week 3 DVOA Ratings

I wonder what they're rank would be if you throw out Week 1. I've always wondered if some sort of outlier test would be helpful; or if you just throw out a team's best and worst performances.

That Week 1 was terrible. When we get the 'D' actually incorporated it will look worse as Baltimore's defense has gone to Sh*t since Suggs went out.

Week 2 was just merely below average, but actually had some nice spots. I thought they were fine in Week 3. Still the run game is an absolute disaster, but the passing game was above average. They're trending up, but you are right, they don't seem like the 2nd worst offense through three weeks.

I will note they've had a tough go with two road primetime games. So far this season there have been 10 primetime games; the home team is 7-3. The road wins are the Jets in Indy last Monday, and then the two Broncos wins.

32 Re: Week 3 DVOA Ratings

Apparently the players are teasing Peyton about leading the second worst ranked team in offense. They also call him Papa John which is pretty funny.

It's also hard to throw out outlying bad/good games when you have such a small sample to begin with. It might be appropriate under certain circumstances (like having Clausen as your QB) but it would be hard to draw the line.

38 Re: Week 3 DVOA Ratings

Just to confirm, I'm not advocating throwing games out this early, but maybe by midseason or so, or even at year-end, at some point when it makes sense.

Then again, I have no idea if this would improve the intended goal of DVOA.

51 Re: Week 3 DVOA Ratings

I figured you were talking about later in the season. But as that point there is also weighted DVOA which will lower the impact of some of those games.

The Broncos first game is going to look really bad regardless since Peyton clearly isn't capable of playing well under center with that line and Baltimore without Sugs doesn't get nearly enough pressure. Will just have to accept that it will be weighing down their offensive performance for a while.

52 Re: Week 3 DVOA Ratings

It's not about lessening the effect of early season games, because as you said wDVOA does that, but more lessening the impact of the 'they didn't show up' game.

Like if you took out Arizona's 47-7 loss in '08 to New England, a game they basically decided not to play, it may have indicated Zona was slightly better than their final DVOA looked - or the infamous Patriots / Chiefs game last year (though that was early).

166 Re: Week 3 DVOA Ratings

Yeah, for rate stuff I wouldn't bother. There are enough plays by the team for the full season that making that kind of adjustment isn't going to be huge in the grand scheme (though you could skew one single game significantly).

I take the D in counting stuff like DYAR with a giant grain of salt - especially in single games - because among other reasons, you can make a solid argument every week about any number of opponents either under or out-performing their DVOA due to superb game planning, better than usual execution, terrible play by the other team (and all the opposites) and your eyes would absolutely not be lying to you.

But even though I agree that the Baltimore D Week 1 wasn't the same as the rest of the year and I expect it to be bad, I don't think that Denver's eventual offensive DVOA is going to suffer too much even if that terrible game becomes slightly more terrible. Especially since it's week one and de-weighted.

(Though now I *am* curious to see what kind of effect a manual week one adjustment done after the fact, 16 games later, might have... like if 2015 BAL D is 28th in DVOA but the same plays for the Den O are adjusted by the #1 D for just that week. Probably not very much, given the "it doesn't improve the numbers" statement, but I'm still curious.)

36 Re: Week 3 DVOA Ratings

Very interesting matchup on Sunday, even for non fans of the Vikings or Broncos, with two offensive Hall of Famers facing particular challenges. Clearly, the Vikings don't want their offensive line to try to win this game against these pass rushers, with these dbs and linebackers breaking on passes thrown by Bridgewater under duress. The Broncos are terrific on defense acorss the board, but they are better agaibst the pass. Time to find out if 28 can summon the 2012 season for 3 hours, and defeat all that a defense can bring at him, behind an o-line that is better at moving forard than stepping back.

On the other side, the Broncos o-line may be worse pass blockers than the Vikings, and Zimmer is a master of double A gap pressure, and he has some talented pass rushers, and, especially if Rhodes gets on the field, he has some dbs who can play tight. The Vikings are so-so against the run, but the Broncos are terrible at running. As usual, the oddsmakers make it tough, with most lines favoring Broncos a +6.5. The Broncos certainly have fewer things that need to go right in order to win, in Denver, so they should prevail. If Zimmer can fool longneck a time or two, however, while his blockers get confused a couple times, and 28 can do waht he has done so often against loaded boxes, Bridgwater may get to hit a a deep throw or two. Take the points, sez the degenerate romantic!

I'm a cynic, or at least a pessimist, however, so I'll stay away from this one.

81 Re: Week 3 DVOA Ratings

If Peterson is called to run up the gut, I suspect he'll be in for a long day. The Denver defense plugs those gaps pretty effectively. From what I've seen (i.e., completely non-scientific observation), the Broncos tend to give up more yardage on the ground when runs go wide; opponents have had some success sealing the edges.

91 Re: Week 3 DVOA Ratings

Oh, on those rare occasions when I put some skin in a Vikings game, it always is when I bet against them as favorites. I can always tell when the public has become too enamored with them, but I sometimes can't when I have.

169 Re: Week 3 DVOA Ratings

I tend to take the same approach with the Colts.

Still, even with Luck hurt and bad and the Jags getting 10, I have a very very hard time playing that game this week.

Or any of them, really. Last week was great, but I have a really hard time finding any game at all that I like this week.

15 Re: Week 3 DVOA Ratings

To point out how bad things are in SF coaching-wise, the Cards put this up on their own website:

http://www.azcardinals.com/news-and-events/article-2/Cardinals-Film-Room-Tyrann-Mathieus-Pick-Six/3f509408-e02f-47e0-b481-a951f6d6477a

Pathetic, they knew exactly what the niners were going to do. That they are confident enough to post this publicly speaks volumes about their opinion of the niners' coaching staff.

94 Re: Week 3 DVOA Ratings

This is the reason I have a hard time getting too worked up about Kap's horrific performance. The coaching staff and their gameplan are doing him absolutely no favors.

97 Re: Week 3 DVOA Ratings

Yeah, but that can be said about a lot (if not a majority) of QB's in the league. Ben McAdoo's idiotic system in NY is almost pathological determined to play away from Eli's strengths, but Eli still looks competent (if not like the All Pro he did in 2011.) The St. Louis Rams offense is god-awful in every way, but Foles hasn't melted down like Kaep. Some drop in performance under bad coaching is reasonable, Kaepernick's total collapse calls into question his value altogether. I personally think he's like RGIII and his success was almost entirely the product of great coaching...

118 Re: Week 3 DVOA Ratings

One game does not equal total collapse, if it continues then your comment would be more appropriate.

Edit: according to PFF Kap is being pressured on 43% of his pass attempts, which is crazy when you consider that quite a lot of his passes are off bootlegs or quick screens.

http://www.footballoutsiders.com/stat-analysis/2015/2014-pressure-plays-quarterbacks

According to this article the league average DVOA under pressure is 44% and 7.6 yards per pass, with pressure it’s -77% and 3.2 yards per pass.

As I say, he's getting very little help.

136 Re: Week 3 DVOA Ratings

Yes, as the players around him have declined.

I've said here before, I don't think that he's capable of transcending the players around him but he has shown that he can be productive when he has the help. He also looked better in the first two games, Sunday was atrocious but it isn't his career.

161 Re: Week 3 DVOA Ratings

Step one, find a mirror. Step two, look into it. You make very bold assertions and then get annoyed when people suggest that the picture might be more complicated than the image you describe.

To try and return this to a more civil conversation; the line is garbage, there's only one halfway decent receiver and the defense is shipping huge numbers of points, how would you expect an average passer to perform? (And please try to consider more than just his worst game out of three)

171 Re: Week 3 DVOA Ratings

Karl, I haven't on multiple threads been accused of being a troll. I've been called an asshole, sure - but you're intractable. Discussions with you are very close to pointless - hence the trolls accusations.

Let's start with this: what would be proof to you that Kaepernick is current awful and will continue to be bad?

Here are things that you say are not proof, either by themselves or in combination:
Having the worst DYAR/DVOA in the league.
A three year long arc of decline.
Having only excelled under the guidance of a great coach who engineered an offense to his strengths.

Very simply and clearly: what would be the proof?

From there, we can test if Kaepernick meets your qualifications and further test is several other QB's we can all agree are bad and will continue to be bad into the future also meet those standards. Keep in mind, the argument isn't: "are there circumstance under which Kaepernick could be good," to which the answer is "even Jimmy Clausen & McCown looked good under Trestman" but "what proof would convince YOU that Kaepernick is currently playing awful and will continue to be bad into the future?"

It's very simple and I've asked you to do stuff like this before and you never, ever do it. Hence, the troll accusations...

185 Re: Week 3 DVOA Ratings

For what it is worth I tink you're both right. I'd certainy wager, at this point that Kaepernick is not going to be good, ever. Of course, that's the smart bet with nearly every qb, and this guy's trendline is very bad, as you note. On the other hand, his current situation is awful. What concerns me most is the limited insight I can gain, into his approach to his career, from media reports, which provides reason to think he just doesn't get it. He, still south of 30, though, so I wouldn't write him off, but sadly, for Niner fans like Karl, the best chance of salavaging his talent probably lies in getting the hell away from the Yorks, and to a coach with credibility to have a very frank conversation, that takes with him. He needs to humble himself, completely, to the profession, but that can be tough to do once you've already earned 30 million.

Like I said, however, I'm really guessing. This isn't Cutler in his 30s after multiple coaches.

190 Re: Week 3 DVOA Ratings

I'm not even saying he's good or going to be great. I think his various attributes add up to average, which is how he's paid.

My belief is that an average passer surrounded by crap will produce poorly, which is what we're seeing right now. My other point is that he did play much better in the first two weeks and some people are basing too much of their judgement on one really bad week, the worst of his career.

And for that I get called a troll.

197 Re: Week 3 DVOA Ratings

Hey, I'm at the point in my fandom where I was kind of hoping that Peterson would get out of Minnesota, because I wanted to see that rushing talent paired, if only for one more season, to go with the other single season, with an undeniably top level qb talent. In other words, I mostly just root for the most generic entertainment these days. I think watching Kaepernick play the position well, with his physical tools, is fun. I can't see how that can happen in San Jose, given the mess that is management. I thus root for Kaep to get out of there, to a place where me might maximize his talents.

219 Re: Week 3 DVOA Ratings

Yeah, I really think it isn't unrelated, strangely enough. They had a qb still able to really fling it, but mediocre to bad to very inconsistent, young (Harvin) receivers, and thus Peterson still drew all the attention, which Stubbleface acknowledged more than once. So Stubbleface has perhaps his best statistical season, at age 40.

Anyways, if the Vikings had done something really stupid, and cut Peterson for fear of his pr problems, and cap number, it would have been fun to see him go to Dallas, behind that line, with Romo and Bryant, uninjured versions.

259 Re: Week 3 DVOA Ratings

Aaaaaand once again, you refuse to outline any standard to see if it applies.

This is what I mean: just tell me what proof you would need and we can if it works for Kaep. If your terms make sense, you can win this debate! But you never, ever give out concrete terms...

269 Re: Week 3 DVOA Ratings

You colossal hypocrite. I asked you to define how an average qb would perform when surrounded by a crap line, crap receivers, with a crap defense and crap coaches. Did you answer? No.

And I did answer, I said I would take more than his worst game ever as evidence because one game is not suitable to describe anyone's career, an entirely reasonable position.

Being called a troll by you, the resident FO troll is water off my back. That you brought the nonsense from the deflate-gate thread into this just shows you up as who you really are. Was it you that made the absurd claim about teams not investing in LBs without even knowing who was playing at which position and then resorted to shouting, swearing and ranting at me because you were wrong? You have form here.

290 Re: Week 3 DVOA Ratings

Yes I did: I said Flacco, Foles, Cam all have situations of a crappiness equal to Kaepernick (I think all three have it far worse, actually) and all have DVOA/DYAR much, much better than him and have shown none of his issues. You could throw in Derrick Carr in you are really insistent that the defense and coaching be beyond the pale in evaluating a QB. I answered your question directly and brought it up even before you raised it. You can put Cutler in there too - again a player with a much, much better DVOA than Kaepernick's.

I have also repeatedly pointed out that it is not a single game I am taking issues with Kaepernick over: it is a season and 3 games. Last year, he was also awful, with a better coach, a better defense and skill position players who were a year removed from failure.

I would also say your logic is fairly circular, you want examples of an "average" QB on a team as bad as Kaepernick's to compare him to, but the fact of the matter is that teams with a competent QB's are almost never as bad as Kaepernick's. If you want to use a word like "average" I'd need clarity on what that means to you.

Also, the LB claim you did nothing to disprove it - it was the classic case of you refusing to offer a standard to see if a proposition was true and repeating demonstrably false claims.

307 Re: Week 3 DVOA Ratings

That's an interesting way of describing your behaviour in that LB discussion. You said that those teams didn't prioritise linebackers and then I pointed out that they did and had each spent at least a 1st and 2nd there. Then you started shouting and swearing and said you were right because two of those teams weren't spending any cap room there, a point that looks comically absurd after Wagner and Kuechly were given such massive contracts which was obvious at the time. A classic case of you making a strong statement and then ranting at somebody else because they disagree with you.

160 Re: Week 3 DVOA Ratings

There's "productive when he has help" and there's "productive when surrounded by the greatest offense ever assembled and under the tutelage of one of the greatest coaches ever."

174 Re: Week 3 DVOA Ratings

What's hyperbole - that Matt Cassell has never been good without extreme help (true) or that the Patriots were the best offense ever with one of the greatest coaches ever (also true)?

184 Re: Week 3 DVOA Ratings

Hey man, I think Harbaugh is great and if he had kept coaching at the same level, after 10 more years he definitely would've been in some discussions about "greatest ever."

(Hence the universal agreement that the Yorks are a bigger problem than Tomsula or Kaepernick could ever be...)

195 Re: Week 3 DVOA Ratings

Why did the 49ers get rid of Harbaugh again? I think it's poetically hilarious that the team was godawful immediately before Harbaugh arrived, and now they're godawful immediately after he left.

199 Re: Week 3 DVOA Ratings

Harbaugh's an A-hole, and the Yorks are A-holes who are too stupid to grasp that when you have an A-hole who can make you a lot of money, then you just stay away from him, so his A-holery isn't something you have to observe.

183 Re: Week 3 DVOA Ratings

Cassel's ANYA that year was 4.26! That's the exact opposite of excellence.

I suspect you're talking about 2010 (KC was 10-6 and won the west), and I remember that year well. I wouldn't call it excellent. He did as excellently as a Matt Cassel can be expected to do, and didn't do anything catastrophic, but he was still only just good enough to give you hope and get you beat by any good team/coach/D. Very similar to peak Schaub and Dalton in that respect. He left a lot of opportunities on the field that year.

But he limited turnovers, the team played well around him, and he showed enough that he's still highly in demand as a backup QB not named McCown. He was better than replacement, but certainly nowhere near excellent, even at his absolute peak.

209 Re: Week 3 DVOA Ratings

Matt Cassel is a testament to what a great coaching staff and a loaded roster can do. His qbrs even with those chiefs teams were always below average and its pretty clear by now that Cassel is a solid backup, nothing more at this point.

Its one reason I believe BB deserves every bit of plaudit and adulation the media bestows on him. Again, that was Cassel's first start since high school. And he had to execute a modern passing game as opposed to scramble, heave, and read option wizardry that allowed Vince young and Tim Tebow to survive for a short while.

175 Re: Week 3 DVOA Ratings

It's a reasonable question but I don't see how Kap can be held responsible for Crabtree tearing his Achilles, Jonathan Martin being truly awful and Anquan Boldin, Joe Staley and Vernon Davis getting old. No matter how crap Kap is I don't think that he should be held accountable for the ravages of time.

188 Re: Week 3 DVOA Ratings

I agree 100% and I think so would everyone else.

Kaepernick's decline was no surprise to anyone because even when he was very successful, it came in large part due to simplified play design and easy reads (the coaching), and his running skill (and it being more of a surprise to people that weren't prepared for it). Even then, he was notorious for leaving plays on the field, not pulling the trigger on throws to open players that'd have gone for big gains, occasional mechanical breakdowns, and dropping the eye level and running as soon as his first read wasn't open (or even sometimes when it was) or he felt the pocket collapsing (even sometimes when it wasn't). This was all very obvious on the tape.

Now with a few years under his belt and plenty of tape on him and more of the league adjusting to read-option stuff and his speed, it is becoming apparent that the mental side of the game hasn't changed for him. We were encouraged in week one when he did show some awareness, better eye level, and good decisions, but again, that was just one week. Since then, it's more of the same.

And yes, it certainly looks worse and has poorer results when combined with all the stuff you list about other players. You aren't wrong to point that out (or to have hope for improvement). But it's also not entirely attributable to them, given that it's a continuation of the things that were evident even when he was succeeding.

182 Re: Week 3 DVOA Ratings

How much of that pressure is on him though? I read that he's being indecisive in the pocket, not finding open receivers and scrambling instead even on those occasions when the line provides him time. That sounds exactly like Wilson, and Wilson has been the 3rd-most, 2nd-most and most pressured QB in the last three years because of it.

256 Re: Week 3 DVOA Ratings

I think at least part of the difference is that this year, Wilson still has a great defense, and Kaep, not so much. They were not super different in performance over the last 2 years.

258 Re: Week 3 DVOA Ratings

2013, yes, but they had a 593-point difference in DYAR last year. That's like saying Rodgers and Flacco weren't super different in performance last year.

208 Re: Week 3 DVOA Ratings

The crux of the debate between chem and Karl seems to be: Karl- "Kaep is an average player". Chem - "Kaep sucks and probably won't be starting in the league very soon".

I'm curious where the rest of the readers fall. My personal opinion? He's probably average with some very strong highs and some pretty awful lows.

236 Re: Week 3 DVOA Ratings

I agree. He reminds me a lot of Falcons-era Michael Vick. They have the same combination of extreme strengths (mobility, arm strength) and extreme weaknesses (accuracy, decision making) that tend to cancel each other out, and the end result is a roughly average QB. Of course, an average QB can, at times, look anywhere from really good to really bad, depending on their supporting cast and the quality of the defenses they are facing.