DVOA Analysis

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

Week 6 DVOA Ratings

by Aaron Schatz

Hey kids! I'm on the Amtrak Acela right now, winging my way to New York City for a meeting of ESPN the Magazine contributors. We're about 10 minutes from Penn Station, which means that there isn't enough time to get all the different team and player stats pages updated. In fact, there really isn't enough time to write a nice, long commentary on this week's DVOA ratings and where teams stand after six weeks. Nonetheless, I know you, the readers, demand that you get your DVOA ratings hot off the presses! So here are this week's two main tables for the purposes of fun and discussion.

As far as commentary, I feel like I should just tell everybody to read today's Varsity Numbers and just substitute the names of NFL teams for college teams. In this parity-filled season, every team seems to have some sort of weakness. Every team in the NFC already has two losses. Maybe the one team that doesn't seem to have a weakness right now is Pittsburgh (now with a real live quarterback!) and the Steelers do move into the top spot in DVOA this week. Fans may be a bit surprised to see the Giants moving into the second spot, but the Giants defense has been outstanding this season. That's remarkable considering the way that defense completely crumbled at the end of last season. Defensive coordinator Perry Fewell, of course, managed to put together the third-rated DVOA pass defense in Buffalo last year, and without Fewell, the Bills are dead last in defensive DVOA. Time to consider Fewell for some head coaching gigs?

Oddly, despite the parity, the top 11 teams in DVOA all have winning records, and San Diego is the only team in the top half of the ratings with a losing record.

I have no idea when I'll be able to get all the stats pages updated, it will be sometime today or tonight at the latest. However, FO Premium is already updated through Week 6, so Premium subscribers can see the splits and all the separate tables. We haven't had a chance to run playoff odds yet, but I'll also let everybody know when those go up. Until then, enjoy these numbers.

(Note: Stats pages and playoff odds are updated as of midnight Eastern, Wednesday October 20.)

* * * * *

These are the Football Outsiders team efficiency ratings through six weeks of 2010, 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.

Opponent adjustments are currently at 60 percent strength and will steadily grow stronger until Week 10. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

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 19 percent of DAVE for teams that have played six games and 27 percent of DAVE for teams that have played five games.

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 PIT 32.6% 4 31.0% 1 4-1 7.7% 14 -21.1% 3 3.8% 8
2 NYG 30.0% 5 26.5% 3 4-2 11.6% 11 -24.5% 1 -6.1% 30
3 NE 29.0% 1 27.1% 2 4-1 35.7% 1 14.3% 27 7.6% 4
4 TEN 27.6% 8 21.6% 5 4-2 2.9% 18 -21.2% 2 3.5% 9
5 NYJ 23.3% 3 21.0% 6 5-1 16.4% 6 -0.6% 15 6.3% 6
6 PHI 23.3% 10 21.8% 4 4-2 26.1% 4 -1.1% 14 -4.0% 26
7 KC 20.2% 2 14.6% 9 3-2 11.6% 10 -3.1% 12 5.4% 7
8 IND 14.3% 11 15.4% 8 4-2 26.6% 3 6.8% 21 -5.6% 28
9 NO 13.4% 14 12.7% 11 4-2 14.1% 8 -0.3% 17 -1.1% 17
10 ATL 13.2% 6 13.9% 10 4-2 9.7% 13 -5.0% 8 -1.5% 19
11 BAL 11.5% 12 15.6% 7 4-2 7.2% 15 -2.9% 13 1.5% 12
12 GB 11.3% 7 11.6% 12 3-3 14.6% 7 -3.5% 10 -6.7% 31
13 SD 10.6% 9 7.5% 13 2-4 17.4% 5 -12.5% 4 -19.3% 32
14 SEA 9.9% 13 5.4% 14 3-2 -4.3% 23 -3.7% 9 10.6% 1
15 HOU 0.7% 20 -3.1% 20 4-2 30.9% 2 28.5% 31 -1.7% 20
16 MIA -1.4% 22 2.3% 15 3-2 4.8% 17 1.6% 18 -4.7% 27
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 DAL -2.7% 15 -1.2% 18 1-4 13.1% 9 10.1% 24 -5.7% 29
18 CIN -2.8% 16 -2.8% 19 2-3 -3.1% 19 -0.3% 16 0.0% 16
19 WAS -4.2% 19 -0.9% 17 3-3 5.7% 16 10.0% 23 0.1% 15
20 DET -4.6% 18 -9.9% 23 1-5 -4.1% 22 7.8% 22 7.3% 5
21 MIN -5.6% 21 -0.1% 16 2-3 -17.1% 27 -9.1% 6 2.4% 10
22 DEN -7.5% 25 -7.2% 22 2-4 11.6% 12 15.5% 28 -3.5% 25
23 SF -8.2% 27 -10.1% 24 1-5 -13.1% 26 -6.8% 7 -1.9% 21
24 CHI -11.5% 24 -7.0% 21 4-2 -23.3% 30 -3.3% 11 8.6% 2
25 CLE -13.6% 23 -15.1% 25 1-5 -3.5% 20 10.9% 26 0.8% 13
26 TB -16.3% 17 -15.6% 26 3-2 -7.4% 24 6.8% 20 -2.1% 22
27 JAC -20.7% 26 -19.0% 27 3-3 -4.0% 21 24.3% 30 7.6% 3
28 STL -24.2% 28 -23.6% 28 3-3 -18.1% 28 2.6% 19 -3.5% 24
29 BUF -38.8% 30 -31.2% 29 0-5 -11.0% 25 30.1% 32 2.3% 11
30 OAK -39.5% 29 -37.8% 32 2-4 -20.8% 29 17.3% 29 -1.3% 18
31 CAR -42.3% 31 -32.9% 30 0-5 -49.1% 32 -10.0% 5 -3.2% 23
32 ARI -49.0% 32 -37.2% 31 3-2 -39.1% 31 10.2% 25 0.2% 14
  • ESTIMATED WINS uses a statistic known as "Forest Index" that emphasizes consistency as well as DVOA in the most important specific situations: red zone defense, first quarter offense, and performance in the second half when the score is close. It then projects a number of wins adjusted to a league-average schedule and a league-average rate of recovering fumbles. Teams that have had their bye week are projected as if they had played one game per week.
  • PAST SCHEDULE lists average DVOA of opponents played this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • FUTURE SCHEDULE lists average DVOA of opponents still left to play this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
  • VARIANCE measures the statistical variance of the team's weekly DVOA performance. Teams are ranked from most consistent (#1, lowest variance) to least consistent (#32, highest variance).



TEAM TOTAL
DVOA
W-L NON-ADJ
TOT VOA
ESTIM.
WINS
RANK PAST
SCHED
RANK FUTURE
SCHED
RANK VAR. RANK
1 PIT 32.6% 4-1 31.6% 4.4 4 4.5% 8 -5.8% 24 10.8% 15
2 NYG 30.0% 4-2 26.9% 4.5 2 -2.6% 24 2.8% 12 26.4% 28
3 NE 29.0% 4-1 30.4% 4.3 5 -1.6% 20 1.5% 15 10.6% 14
4 TEN 27.6% 4-2 30.7% 4.5 3 -1.3% 19 5.8% 7 12.6% 17
5 NYJ 23.3% 5-1 32.0% 4.2 6 -2.1% 21 0.1% 18 4.9% 6
6 PHI 23.3% 4-2 31.4% 4.5 1 -2.2% 22 7.6% 3 8.2% 10
7 KC 20.2% 3-2 17.9% 4.0 7 0.7% 16 -16.2% 31 13.3% 18
8 IND 14.3% 4-2 14.0% 3.8 9 3.1% 12 5.3% 9 14.1% 19
9 NO 13.4% 4-2 14.6% 3.7 11 -18.0% 32 -3.5% 22 5.8% 8
10 ATL 13.2% 4-2 9.9% 3.6 12 -0.3% 18 -9.8% 28 14.4% 20
11 BAL 11.5% 4-2 7.9% 3.8 8 10.2% 2 -5.5% 23 4.4% 5
12 GB 11.3% 3-3 16.4% 3.8 10 -6.2% 26 5.7% 8 10.5% 13
13 SD 10.6% 2-4 16.9% 3.3 14 -17.2% 31 2.6% 14 24.1% 27
14 SEA 9.9% 3-2 8.9% 3.6 13 -8.2% 29 -13.8% 30 17.5% 22
15 HOU 0.7% 4-2 1.1% 2.9 16 3.0% 13 8.9% 1 21.1% 25
16 MIA -1.4% 3-2 1.9% 3.2 15 3.8% 10 1.2% 16 15.0% 21
TEAM TOTAL
DVOA
W-L NON-ADJ
TOT VOA
ESTIM.
WINS
RANK PAST
SCHED
RANK FUTURE
SCHED
RANK VAR. RANK
17 DAL -2.7% 1-4 -6.0% 2.7 21 1.4% 15 6.1% 6 17.6% 23
18 CIN -2.8% 2-3 1.4% 2.7 22 -6.3% 27 8.9% 2 8.2% 9
19 WAS -4.2% 3-3 -4.8% 2.7 20 3.8% 11 4.9% 10 4.3% 4
20 DET -4.6% 1-5 -6.4% 2.6 23 3.9% 9 -1.7% 21 9.6% 11
21 MIN -5.6% 2-3 -12.3% 2.8 19 5.6% 6 -1.3% 19 3.9% 2
22 DEN -7.5% 2-4 -6.6% 2.5 25 11.0% 1 -9.8% 29 4.0% 3
23 SF -8.2% 1-5 -15.3% 2.9 17 6.7% 5 -18.1% 32 30.0% 31
24 CHI -11.5% 4-2 -7.2% 2.2 27 0.3% 17 2.7% 13 30.0% 30
25 CLE -13.6% 1-5 -12.2% 2.4 26 9.7% 3 0.4% 17 1.9% 1
26 TB -16.3% 3-2 -15.2% 2.8 18 -2.5% 23 -6.5% 25 23.7% 26
27 JAC -20.7% 3-3 -25.9% 2.5 24 4.9% 7 3.4% 11 27.8% 29
28 STL -24.2% 3-3 -14.2% 2.0 28 -12.8% 30 -7.5% 27 18.1% 24
29 BUF -38.8% 0-5 -37.8% 0.8 32 8.3% 4 7.0% 5 5.6% 7
30 OAK -39.5% 2-4 -28.6% 1.2 30 -7.1% 28 7.1% 4 11.8% 16
31 CAR -42.3% 0-5 -42.9% 0.8 31 2.5% 14 -1.6% 20 10.4% 12
32 ARI -49.0% 3-2 -44.6% 1.2 29 -5.3% 25 -6.8% 26 33.2% 32

Comments

247 comments, Last at 30 Oct 2010, 7:08pm

1 Re: Week 6 DVOA Ratings

What is the record for combined 'good offense, bad defense' rating? Houston being +30.9% on offense and +28.5% on defense is pretty lopsided between D and O, but I'm sure there have been worse. Just curious.

15 Re: Week 6 DVOA Ratings

Without looking it up, I will guess that the top 5 has to include the Chiefs team from ... I don't know, 2003? coached by Vermeil, QB'ed by Trent Green, that lost to the Colts in the playoffs in a game which had something like 1 punt.

EDIT: I was wrong, that Chiefs team was 32.2% on offense, but only 7.5% on defense. I was also wrong in that the game featured ZERO punts, not one.

Here are all the teams I could find that had offensive and defensive DVOA over 10%, ordered by the lower of the two ratings. This year's Houston team would indeed be the most extreme ever, if their ratings stay the same for the rest of the year.

2010 HOU 30.9 / 28.5
2008 DEN 24.0 / 24.7
2004 MIN 18.4 / 23.0
2004 KC 32.9 / 17.1
2004 GB 18.3 / 16.4
2002 KC 38.0 / 15.1
2008 HOU 11.0 / 17.9
1997 CIN 14.6 / 14.2
2000 SF 17.8 / 14.4
2010 NE 35.7 / 14.3
2000 STL 25.9 / 13.5
1996 BAL 17.7 / 13.5
2010 DEN 11.6 / 15.5
1995 DEN 19.6 / 11.0
2008 NO 21.1 / 10.7
2008 ATL 12.9 / 10.6
2008 JAC 12.6 / 10.2
2010 DAL 13.1 / 10.1

65 Re: Week 6 DVOA Ratings

I was surprised Indy never made this list so I looked it up. The only 2 years over 10% on D in the Manning era were 1998 and 2001, which were also the 2 worst years on offense of the Manning era.

241 Re: Week 6 DVOA Ratings

You may well be right, and of course some regression to the mean should be expected. On the other hand, the causative factors kind of suggest we should expect the split to get even more extreme: the offense has played 4 out of 6 games to daye with its starting left tackle suspended, has had the best receiver in football significantly limited by injury from Week 2 onwards, has had a pro bowl tight end gradually working his way back from injury, and has lost its talented, dangerous WR3-deep threat for several games. All of these guys should be pretty much full go after the bye, which means teams will not be able to emulate the Giants' game plan of crowding the box to stop the run, frequently fielding extra defensive linemen, and simply relying on the pass rush to get to Schaub before the Texans' banged-up receivers could get open deep. The defense, on the other hand, has just lost one of its few really excellent players to IR. The injury situations, past and present, suggest that if anything, the offense ought to get better and the defense, to the extent such a thing is possible, even worse.

I expect regression to win out, of course. But I really wouldn't be too surprised if this team is still at or very near the top of that list come season's end.

219 Re: Week 6 DVOA Ratings

Cincinnati from about 4 years ago has to be in this conversation. I recall a ton of 38-35 type games.

edit: Woops, somehow missed the giant wall o' rankings.

2 Re: Week 6 DVOA Ratings

How do the NYG get a D bump in VOA with the 24th ranked schedule. Shouldn't they get a hit, like the 4 teams behind them with the the 19-22 ranked schedule? Shouldn't NO get a bigger D hit to VOA with the 32nd ranked schedule than teams with harder schedules? How does SEA get a D bump with the 29th ranked schedule? Shouldn't that be a hit?

What exactly does the D in DVOA mean again?

Does it mean that teams that do well in DVOA vs VOA employ more frequen successful plays by down and distance vs. teams that employ fewer successful plays but get a larger proportion of big plays?

Because if that's the case, maybe you should rearrange the tables to put the adjustment from VOA to DVOA in the first table - away from the strength of schedule in the second. It's just confusing things (for me, at least).

21 Re: Week 6 DVOA Ratings

Kind of a stretch. Here's the DVOA rankings of the defenses they've played:

CAR - 5th
IND - 21st
TEN - 2nd
CHI - 11th
HOU - 31st
DET - 22nd

Mean defensive DVOA of Giants opponents: +1.72, which would rank 19th among NFL teams.

42 Re: Week 6 DVOA Ratings

This comes up very single week. More things go into DVOA than just opponent adjustments. There are also some fumble adjustments or something. I don't remember what specifically.

3 Re: Week 6 DVOA Ratings

it is remarkable how the bottom five in ST DVOA is a who's who of teams whose record should really be better. I know this is usually the case, but it is pretty stark so far this year...

4 Re: Week 6 DVOA Ratings

Looking forward to the TEN-PHI game next week (otherwise known as the quasi-backup QB bowl)

5 Re: Week 6 DVOA Ratings

Anybody who says they have useful insight as to who will win the NFC oughta' be makin' their omelettes with more conventional mushrooms. And fire their 300 pound Samoan attorney.

I'd say the snapshot of the Vikings is a fairly good representation; good on defense, above average special teams, bad on offense. The reason for optimism is that blitz pick-up is something that can improve significantly over the course of a half season, the receiving corps has obviously improved in the last 10 days and is scheduled to make further significant strides, and the schedule is getting much easier after week 8. The reason for pessimism is that their qb production has a high risk of collapsing at any point, for a variety of reasons, they can't afford another injury to a cornerback, and their offensive line in general has been substandard, I think mostly due to poor play at the center position.

I haven't watched Dallas a lot this year, and I probably am being influenced too much by their game against the Vikings, but their defensive rank seems low, and their offensive rank seems high.

63 Re: Week 6 DVOA Ratings

Will,

Very apt. You know, of course, that Hunter S Thompson was a big NFL fan (well, gambling on the NFL, which probably led to an actual appreciation of the game). In one of his books he mentioned a campaign flight with Nixon during which they talked football for a couple hours. What I'd pay to have seen that!

83 Re: Week 6 DVOA Ratings

Fear and Loathing on the Campaign Trail '72. IIRC, it was in the back of a limo (possibly in NH), and HST was only allowed there because he agreed to talk about nothing but football.

(I also like the Eagles)

113 Re: Week 6 DVOA Ratings

At some obscure level, I think Hunter appreciated Nixon, principally because ol' Milhous was incapable of completely hiding his essential mendacity. This likely prompted Thompson's observation, after seeing the '76 Democratic Convention, of, iirc, "When you get to know Carter, you're gonna love Nixon!"

182 Re: Week 6 DVOA Ratings

Not to get wildly off-topic, or violate the "no politics rule" (after all, it's history, not politics), but Nixon continues to be severely underrated.

His biggest negative, exposing the public to the fact that (surprise!) all politicians are power-hungry crooks and liars, should really be considered a huge positive. Despite being a Republican, his progressive credentials are impeccable, e.g. the EPA, OSHA, Clean Air Act, etc. Did more for the environment than any of his successors. Pulled the US out of an unpopular war with an ill-defined goal halfway around the world.

(I also like the Eagles)

92 Re: Week 6 DVOA Ratings

HST actually started out in sports journalism. Always had a great appreciation for the NFL and not just from a gambling angle. In fact his suicide note was titled Football Season is Over. According to his biographer, he always considered the Super Bowl the high water mark for the year.

6 Re: Week 6 DVOA Ratings

It looks like all the parity this year is because nobody can put together a good offense with a good defense. I guess you can count the Giants at 11th and 1st, but after them you have to go all the way down to the Packers and Chargers before you hit someone with a top 10 offense and defense. And those three teams are the three worst special teams in the league.

Former Eagles Fan. Go JETS!

10 Re: Week 6 DVOA Ratings

I would argue that perhaps PIT has put together Good O, Good D, and S.T. This week, they're OFF DVOA is in the postive side of the average range, whereas last week it was just on the other side of the negative - they've swung 10 percentage points from last week. I suspect the first four games are going to cause DVOA to underestimate PIT's offense for a while.

137 Re: Week 6 DVOA Ratings

About that parity - note how many AFC teams are in the top half, and how many NFC teams are not.

6 NFC teams in the top 16 by DVOA. There may not be any real stand-out individual teams, but at this rate the Pro Bowl will be a massive snoozer.

Or...even more of a snoozer than it usually is.

7 Re: Week 6 DVOA Ratings

I'm a little confused by the NE defensive ranking this week. As a thoughtful observer, I have the sense that the NE defense is increasingly showing signs of improvement. It's not consistent, but one would imagine some evidence of trending improvement in the numbers. If NE goes out and actually has an impressive defensive showing in the next few weeks, and then proceeds to build on its improvement, I have the sense that DVOA will be quite slow to track the turnaround. I wonder if this is a more general problem, that DVOA fails to distinguish between teams that are simply weak and perhaps even trending down, such as say Jacksonville in the secondary, and teams that are showing on-field evidence of improvement but still have a ways to go. It seems to me that the positive evidence from the Patriots is how they are periodically able to string together multiple series of strong defensive play, as opposed to a more distributed performance in line with the overall stats that they have been producing.

27 Re: Week 6 DVOA Ratings

The problem right now is that the NE defense goes long stretches of the game doing the worst thing imaginable, i.e. allowing opponents to convert 3rd and long. How many times did the Ravens do that on Sunday?

In many ways, the Patriots defense is already pretty good. But they're not getting credit for putting the opponent in 3rd and long when they turn around and cough of the big gainer to allow a 1st down.

At least, that's my read on the situation.

145 Re: Week 6 DVOA Ratings

Right, but hasn't it been shown that 3rd down DVOA being significantly different from other-down DVOA tends to regress to that other-down number?

151 Re: Week 6 DVOA Ratings

I thought that phenomenon was a year-to-year thing, and that you wouldn't necessarily expect it to correct itself within a season.

46 Re: Week 6 DVOA Ratings

Pretty much what RickD said and I'll add the following: The Ravens are a middle-of-the-pack offense according to DVOA. It looked to me like the Ravens walked all over them all game long, until the last five drives, which went punt, punt, punt, punt, punt.

My subjective opinion is that the Pats D has improved from "Getting beaten like a rented mule and needing the offense to win a shootout" to "Getting beaten like a rented mule most of the game but making just enough plays to pull it out". That difference isn't quantifiable, nor, statistically, can you use it to demonstrate an improvement trend.

I find that last game hugely encouraging, however. They're just going to have to improve by a lot more before DVOA notices it. Baby steps. They just might have an average D (in Weighted DVOA) by the end of the year...

198 Re: Week 6 DVOA Ratings

DVOA does not track trends. It weights each play of the season compared to the average for that down/distance/etc. It does not care whether the play occurred in week 1, week 6, or week 17. If you add more above average plays than below in a certain week, that will make DVOA go up that week.

Weighted DVOA, which will show up later in the season tracks a teams trend, by discounting plays from earlier games.

8 Re: Week 6 DVOA Ratings

I just dont get DVOA. As a Steelers fan, I was more impressed with their victories over Tennessee and Atlanta. Those wins werent enough to get them into first place, but beating up on a hapless Cleveland team was?

13 Re: Week 6 DVOA Ratings

In reply to by Vicious Chicke… (not verified)

You are reading too much into rankings. This most recent game bumped Pitt up from ~25 to ~32 (which seems reasonable considering how well they played) while the three teams ahead of them in week 5 all had lowered their DVOA.

17 Re: Week 6 DVOA Ratings

Rankings just don't matter for DVOA, even if we've been trained to look at stats that way our whole lives. I just kind of group them in tiers and call it a day. DVOA doesn't really think there's much measurable difference between PIT's and TEN's defense, but they have to put one as #1 and one as #2 even though that ordering doesn't mean anything.

Also, don't forget that early season numbers are just wonky. If/when they get recalibrated at the end of the year with proper "D" values, they'll probably look better.

29 Re: Week 6 DVOA Ratings

In reply to by Vicious Chicke… (not verified)

I think this might be a Guts vs. Stomps thing. Apparently Football Outsiders give significant weight to large margin victories over bad teams, as opposed to closer victories over decent/good teams.

39 Re: Week 6 DVOA Ratings

Actually, it is not a Guts vs. Stomps thing. DVOA is based strictly on performance on a play-by-play basis, not a game-by-game basis.

52 Re: Week 6 DVOA Ratings

It doesn't. In fact, it doesn't factor in victory at all. But it does factor in every scoring play a team makes (and allows) in each game. "Stomps" happen to include a lot more of those than "Guts," and DVOA doles out appropriate credit.

121 Re: Week 6 DVOA Ratings

That was a great article. Hey Aaron, how about an update? I have been wondering if the Stomps are better phenomenon may be related to team health. Teams that win easily may be more willing to let guys rest a bit more, and may be a little less desperate for that play on the very edge.

I also wonder if strong teams in crappy divisions or with particularly weak schedules come out ahead vs teams that had to beat a lot of tough opponents. This kinda seems to be the implication of the Stomps, but not quite.

60 Re: Week 6 DVOA Ratings

DVOA will basically rate close wins (and losses, for that matter) as draws.* Teams don't get extra credit for "knowing how to win." Suppose that long run Pittsburgh had to beat the Falcons in overtime had instead been a pick-six for Atlanta; PIT's DVOA for that game would be hardly any different, because that wouldn't change anything about the prior 120-ish plays. Convincing wins over bad teams are usually more impressive than close wins against good teams -- especially this year, when no teams are really that good.

* As another commenter noted, DVOA really doesn't consider whether a team wins, loses, or draws in a given game. It judges every play by the circumstances of the game at that time, regardless of what the final result is. It's just faster to say "DVOA rates close wins as draws" than to say "the total value of every play in a close win will be very similar to the total value of a close loss or draw, all else being equal."

79 Re: Week 6 DVOA Ratings

In reply to by Vincent Verhei

I see, but I'm not sure if I totally agree. Yes, the Steelers beat up on the Browns but they were supposed to.

I guess what I am saying is that the immeasurables (like gutting out 3 wins with your 3rd and 4th string QB's against pretty good teams) means more to me than whooping a perenial loser.

Stats are fun to play with, but I dont think anything can really beat the eyeball test.

91 Re: Week 6 DVOA Ratings

And can your eyeball test results be broken down into offense and defense, down, distance, game situation, passing vs. rushing, etc., to tell us not just who will win, but why they will win?

153 Re: Week 6 DVOA Ratings

Well, not exactly....but I would hazard a guess that somewhat knowledgable football fan could pick the winner of a game just as reliably as DVOA. Is that what DVOA is for? You are saying that the team with a higher DVOA should beat a team with lower DVOA every time...has that ever been tested? What is DVOA's win% per se?

199 Re: Week 6 DVOA Ratings

In reply to by Vicious Chicke… (not verified)

No. The team with the higher DVOA is more likely to win. If there's a big delta in DVOA, the win is more likely. If there's a small delta, it's a crap shoot. Nothing can predict the actual winner every time. There's too much luck.

208 Re: Week 6 DVOA Ratings

In reply to by Vicious Chicke… (not verified)

At DVOA's core, it's a useful tool for comparing different aspects of teams to find out which one is better at certain things than others. Saying that a Team ranked 3rd in DVOA will beat a team ranked 7th in DVOA 90% of the time is a real stretch though because those rankings don't tell you much. Their could be a group of teams with very close DVOA scores from 3rd to 7th and have a very small statistical difference between them, but the rankings will make it seem that they are quite far apart.

Another issue with using the ranking to compare teams is that it doesn't take into account their individual strengths and weaknesses. For example, if a team has a strong positive offense DVOA and it plays a strong positive defense dvoa, the expectations is that the offense of that team with the high offense DVOA will succeed at making positive plays very often throughout the game. However, items like special teams and the relative ranks of the the other offense and defense in the stadium can also impact the outcome of the game. This means that DVOA rankings can not be used as a reliable predictor of future onfield performance for an entire team.

What it does mean, though, is that when you compare offense to defense, it IS reliable (to as reasonable statistical degree, the game of football does include the possibility of luck influencing the outcome of games, not limited to fumble bounces, officiating, weather favoring one team over another, time and location of games, etc) in predicting wether one will be successful against the other.

215 Re: Week 6 DVOA Ratings

"What it does mean, though, is that when you compare offense to defense, it IS reliable (to as reasonable statistical degree, the game of football does include the possibility of luck influencing the outcome of games, not limited to fumble bounces, officiating, weather favoring one team over another, time and location of games, etc) in predicting wether one will be successful against the other."

Is there any evidence, other than intuitive and anecdotal to support this. I mean, ideally this is what we assume when we look at DVOA, but is there support for this. To be able to demonstrate this, would be a huge plus for FO, but I don't recall this ever being posited or tested. Additionally, this is where the discussion of heterogeneity of variance and different means last week comes into play. Since the the scales of measurement may differ, at certain points it may not be possible to determine whether a teams OFF is greater than the oppositions DEF.

149 Re: Week 6 DVOA Ratings

Of course not. Why dont you read what I wrote before making comments.

I was saying that in Pittsburgh's games (which yes smartass I watched them all), I found their wins over Atlanta and Tennessee to be more impressive than their win over Cleveland. Yet here the Cleveland win seems to be weighed more heavily because they blew them out.

157 Re: Week 6 DVOA Ratings

PIT's offensive DVOA improved markedly from week 5(4) to 6. DEF DVOA fell off slightly over the same time period. The ATL and TEN wins were impressive from a defensive perspective. The offense, not so much. The defense did not look as dominant against CLE. The marked improvement in the offense (AEB by the blowout) is indeed the key to the move.

94 Re: Week 6 DVOA Ratings

Well, the idea behind this site is to design and test stats to assist your eyeballs, so that a) when you see something you believe, but can't understand, they can help, and b) when you see something you don't believe, you can check to see if there's something you're missing. The stats here aren't trying to replace what you see; they're simply here to help tell the story.

Unfortunately, if what you want is purely subjective in nature, there isn't going to be much here that will please you.