Conference Championship DVOA Ratings

Conference Championship DVOA Ratings
Conference Championship DVOA Ratings
Photo: USA Today Sports Images

by Aaron Schatz

With only two teams left, and most teams not having played for three weeks, we're not going to run the big table with all 32 teams. New England's weighted DVOA is 33.2%, which ranks second behind New Orleans. The Giants' weighted DVOA is 14.1%, which ranks ninth.

Here are the ratings for the Conference Championship games. All four teams get positive DVOA, which is what happens with two very close games and strong positive opponent adjustments. These ratings aren't likely to soothe the hearts of any 49ers fans, as San Francisco comes out slightly higher in DVOA despite losing the game. I know some people might be surprised by the Patriots' offense rating higher than the Ravens' offense, but a lot of that is opponent adjustment and the rest of it is the running game, where the Patriots were much more efficient than the Ravens.

DVOA (with opponent adjustments)
NE 31% 32% -3% -3%
BAL 10% -8% -17% 1%
NYG 12% 10% 3% 5%
SF 15% -5% -22% -1%
VOAf (no opponent adjustments)
NE 10% 16% 3% -3%
BAL -2% 3% 6% 1%
NYG 2% -2% 1% 5%
SF 9% 0% -10% -1%

When it comes to evaluating the Super Bowl, we run into the same problems we had in both 2007 and 2008 (well, actually 2008 and 2009, but let's not get confused). How much sample size do we use to judge these teams? Can we judge them on just the last few weeks? Do we judge them on the whole season, even though the Giants have played so much better in January?

Usually, we're judging teams at this point by weighted DVOA. The Giants' weighted DVOA runs into problems because their big winning streak is only five games, but weighted DVOA gives almost full weight to the last eight weeks and a lot of weight to the four weeks before that, which means that weighted DVOA also incorporates the team's Week 10-15 slide.

Based on weighted DVOA, the Patriots are clear favorites in Super Bowl XLVI. (Our playoff odds report is based on weighted DVOA and gives New England a 61 percent chance of winning.) What if, instead, we only look at New York's five-game winning streak. Here's a look at what each team's DVOA rating would look like if we only included the last five games: Weeks 16-20 for the Giants and Weeks 15-20 for the Patriots. The ranks represent where these ratings would have ranked among all 32 teams for the 2011 regular season.

1 NE 37.7% 15-3 (5-0) 41.7% 1 12.5% 24 8.4% 1
2 NYG 36.8% 12-7 (5-0) 26.2% 5 -12.7% 2 -2.0% 27

Now, that's a lot closer. I think those ratings are a better match for how the public views this Super Bowl. In fact, these ratings may have the Patriots a little bit higher than conventional wisdom. I'm guessing most fans think the Giants have outplayed the Patriots over the last five weeks.

The Giants' late-season turnaround is the big reason why this year is being compared to 2007. However, there are a lot of reasons why this year is not a replay of 2007, and it's important that people understand the differences.

1) The Patriots aren't as good. The 2007 Patriots may have been the best regular-season team in NFL history. They were undefeated, remember? That's why they went into the Super Bowl as such huge favorites. The 2011 Patriots finished the year fourth in DVOA.

2) On the other hand, the Patriots aren't fading. The 2007 Patriots were absurdly dominant in the first half of the season, but were slowing down by the time they got to the playoffs. From Week 1 to Week 12, the Patriots had a DVOA rating higher than 49% in all 11 games. From Week 13 through the Super Bowl, the Patriots had a DVOA rating higher than 49% in only two games (Week 14 over Pittsburgh and the Divisional round win over Jacksonville). The Patriots won despite negative DVOA in both Week 13 and Week 15. This year's Patriots have stayed fairly consistent over the course of the season, with both good and average games throughout the year.

3) The Giants are better than in 2007. You may have seen me tweet this a couple times in the last week or so: The Giants are the first team in NFL history to make the Super Bowl after getting outscored by opponents in the regular season. They had a worse record than the 2007 team (9-7 instead of 10-6). Nonetheless, by our ratings, this year's Giants were the better team. The 2011 Giants were 12th in DVOA at 9.0%. The 2007 Giants were 16th in DVOA at 1.1%. A big reason: The 2011 Giants ranked third in the league in schedule strength.

The difference between 2007 and 2011 is really clear if we look at weighted DVOA. In 2007, the Patriots' weighted DVOA going into the Super Bowl was 43.5%. The Giants were at 3.0%, so the gap was over 40 percentage points. This year, the difference is less than half that.

4) New York's improvement is very different. In 2007, the defense improved a bit in the late-season run, but the big change was the sudden maturation of Eli Manning and the improvement of the offense. This year, the offense has improved a little bit, but for the most part the Giants have been this good all year. Back in 2007, we thought Eli Manning was an average quarterback and a disappointment as number-one overall pick. But in 2011, we know Eli Manning has been one of the top quarterbacks in the league for the last three or four seasons. The Giants offense ranked seventh in offensive DVOA this season, and fourth in pass offense DVOA. The dramatic turnaround this year has been the defense. Through Week 15, the Giants' defense ranked 22nd in DVOA. Since then, as noted in the table above, they've played at a level that would have made them one of the top defenses in the league during the regular season.

5) The Giants are less of a surprise. In 2007, we sat here saying, "look, it's nice that the Giants keep winning games by three points, but teams that were mediocre during the regular season just don't win the Super Bowl." Well, they did, and the next year the Arizona Cardinals almost did. It's a lot easier to believe in the Giants' turnaround when we've seen similar things in recent years.

Of course, that's not going to make Patriots fans feel any better if they lose.


206 comments, Last at 02 Feb 2012, 8:04pm

118 Re: Conference Championship DVOA Ratings

How many things are wrong in this sentence?

Finally, if The Giants complete this run and win the Super Bowl, then they would have completed possibly the greatest playoff sweep in the history of the league defeating a potent Atlanta team, a tremendous defending champion in the Packers, a strong 2nd seed in the Niners, and arguably the best organization in league history at it's apex in The Patriots.

In comparison with the most obvious example, the 2007 Giants, you've already lost.
The 2007 Giants had to win three road games to make the Super Bowl. And the claim that the Patriots are at their apex now is laughable. They were much better 4 years ago.

128 Re: Conference Championship DVOA Ratings

Well we'll agree to disagree on this one. The Giants beat lesser competition in 2007 to get to the Super Bowl. It seems to me that both the Packers and 49'ers from 2011 are better than the Packers and Cowboys from that season. And I still think the Pats are in the middle of their historical height. Sure they were a better squad in 2007 but they still went 13-3 this season and have Brady and Belichick at the height of their powers. Overall the 2011 Giants' road to a championship is tougher in my opinion than the path in 2007. And from a historical perspective the 2011 Falcons, Packers, Niners and Patriots would be a hell of a feat.

186 Re: Conference Championship DVOA Ratings

You don't remember being stunned that the Giants won in 2007? I do. So do the Giant fans I watched the game with. No one expected them the Packers to lose. Can you honestly say that this year, with this defense, you were that confident?

138 Re: Conference Championship DVOA Ratings

I was talking about the 2006-2011 Giants as a whole, being the only team to make this improbable run, and not once but twice now. 2007 was obviously the upset of the century, and the 2011 run appears to be similar in nature, if not nearly as impressive.

145 Re: Conference Championship DVOA Ratings

Atlanta can't beat quality opponents, the Packers were an excellent team but not as good as their record and not much stronger than a typical #1 seed, if at all, the 49ers struck me as a distinctly weak #2 seed and this is really not one of the Patriots' best teams (clearly behind 2003, 2004 and 2007, at the very least). It's still a very impressive run, but I don't think it's historic.

91 Re: Conference Championship DVOA Ratings

I'm not sure you need to alter the model to take into account something that is "amazing." The whole point this being the "greatest playoff sweep in the history of the league" as you put it, is that it IS an unprecedented event. If you start messing with your system to account for every outlier, you end up with a messed up system.

170 Re: Conference Championship DVOA Ratings

I agree that you don't want to end up with a "messed up system", however you also don't want to simply say the lesser teams are winning without understanding why or if they are in fact the lesser teams. It is just lazy to conclude that DVOA is objectively correct (since it is a model). The point is that any of these models are merely a snapshot of an ever-changing reality - that is after all why the stat is weighted to favor recent performance over earlier performance. But these weights are random and don't necessarily explain what happens on a particular gameday. The Giants have raised their level, this is certainly true, but to call them poor or mediocre because of the regular season they had is to fail to appreciate their current level of play. This is a team that dominated The Packers who were an awesome team most of the season. They then defeated the darling of stat-heads, the 49ers, on their home field, and essentially shut out the Falcons. The chance that this string of performances is a fluke rather than the performance of a superior team is highly unlikely. And if they go on to beat the Patriots then it is my contention that they have had one of the great postseasons. For the purposes of understanding who are the great teams of all time, that would beat anybody on a field at their peak, I throw my hat in with the teams that have achieved a string of dominant performances when the games matter the most.

94 Re: Conference Championship DVOA Ratings

Everyone has pieces of the truth. Passing dominant outcomes are more variable than running dominant outcomes. The internet and other technologies make each team smarter about the other. Free agency levels the playing field. More teams and more games yield more chances for variation.

But whatever it is, I can't complain. The worst position major league baseball ever got in was the sustained impact of the Yankees (1920s-early 60s) dominating the sport, and even more the American League. That sport is still walking a tightrope away from disaster because of its financial inequity. I don't think the Cardinals or Giants or Packers making or winning the Super Bowl over some 15-1, 14-2, or 13-3 team is the end of the world. Even, no especially, over a 16-0 team.
It hurt not to see my Packers playing yesterday-- but they earned their fate.

99 Re: Conference Championship DVOA Ratings

"The Giants are better than in 2007."

Really? Same QB, yes he's improved but he was playing at a high level in late 2007. But sure, the 2011 version is better. Same RB's, but they were definitely better in 2007 (especially Jacobs). Better TE in 07. WR's are arguable, but with Plax hurt in the SB I'd say this years are better. MUCH better O line in 07, and this is probably the biggest difference between the two teams.

D lines are equivalent, both very good. Better LB's in 07. Better CB's in 07. Safeties are equivalent.

I think people still underestimate how good that 07 team actually was, and are starting to overestimate how good this years team is.

106 Re: Conference Championship DVOA Ratings

I think the 2011 WRs are clearly better - Nicks is approximately equivalent to Plax, and Cruz/Manningham is much better than Toomer and rookie Steve Smith. Toomer and Smith moved the chains, but all three of the current giants WRs can score on any given play. Shockey was obviously better than Ballard, but he was injured for the entire postseason. Rookie Kevin Boss didn't do much at all until his big catch-and-run in the 4th quarter of the SB, and I'd say Ballard is basically equivalent.

That aside, you make a good point. The 2007 giants were kinda a slightly worse version of this years Texans, IMO - dominant at the LOS on both sides of the ball, which allowed some decent skill players to look better than they probably were. That's not as sexy as Eli throwing big-play touchdowns to Cruz and Nicks, so we assume this year's version is better, but it was pretty effective in its own right.

146 Re: Conference Championship DVOA Ratings

The 2011 Texans' running backs and defensive back 8 were a lot better than the 2007 Giants'. On the other hand, the Giants had a considerably better defensive line. Watt and Smith are excellent, and Barwin and Reed pretty good, but Cody kinda sucks, and that Giants front 4 was awesome. Or to take a more macro view, the Texans had better run defense and coverage, but worse pass rush.

113 Re: Conference Championship DVOA Ratings

Eli Manning in the 2007 postseason was an efficient QB who helped you win games.

Eli Manning in the 2011 season is a fringe MVP candidate, one of the best quarterbacks in football, and a QB who was winning games his damn self.

We're talking about an upgrade on the magnitude of the fuel-efficient 2001-2003 Brady versus today's high-performance version.

A difference that great in your QB play can't be overstated in today's NFL.

195 Re: Conference Championship DVOA Ratings

I said fringe because we saw Drew Brees and Tom Brady break Marino's passing yards record yet both will finish behind Mr. Aaron Rodgers. That's pretty incredible.

If the above three players had great but not legendary seasons, given the circumstances Eli Manning had to deal with -- the absurd number of game-winning drives needed to even reach 9 wins, the utter mess of an offensive line, one of the worst rushing attacks in the NFL -- he would be the type of player receiving some MVP votes as a 4th place finisher.

In other words, fringe candidate.

Not outlandish whatsoever.

CAPTCHA: fact writ

142 Re: Conference Championship DVOA Ratings

So, as I said, WR's and QB are upgraded ( I don't think Eli has improved as much as you but he has improved ). But again, look at all of the areas that were better in 07 than 11. I can't see any way to argue that this year's Giants team is better than the 07 version. Better passing game, much weaker running game, much weaker D.

161 Re: Conference Championship DVOA Ratings

Special Teams:

Same kicker, better punter in 2011, better kick returns in 2007 (hixon > jernigan), equivalently inept punt returns by defensive backs, better kick coverage in 2007. Advantage 2007 (but not enough to impact the outcome).

Overall, 2007 > 2011.

100 Re: Conference Championship DVOA Ratings

Let me get this straight. If you cherry-pick the definition of 'recent games' to favor the Giants as much as possible, they almost come up to the Patriots' level? That's an analysis? What happens if you cherry-pick to favor the Patriots?

I think Giants fans would do better to put their faith in match ups. Or just to say that a 19% DVOA difference points to a close game where one good bounce could give the underdog a win. Cherry-picking is for losers.

102 Re: Conference Championship DVOA Ratings

It was almost a spoof analysis. I know Aaron is a Pats fan. He was just trying too hard to go the other way.
It does bring up two ideas.
1) A cherry-pick optimization chart, showing the two teams' DVOA as a function of the definition of 'recent'. Is 5 games the best Aaron could have done? This would be great service for Giants fans, I am sure.
2) Is it time to figure out the most predictive definition of 'recent'? If I recall, weighted DVOA was somewhat arbitrary originally.

(1) is not serious, naturally. But (2) would be interesting. Sadly, there's probably not enough time to do it right.

But if (2) and (1) came up with the same answer, that would create havoc!

165 Re: Conference Championship DVOA Ratings

Actually, to optimize the Giants' rating, the correct choice would have been the last four games for each team, because the Giants have close to 0% DVOA for the win over the Jets. I picked that one since it was their first win in the streak.

Anything over five games favors the Patriots. One or two games would also favor the Patriots.

167 Re: Conference Championship DVOA Ratings

I thought it might be like that.

I know you were using the winning streak cutoff as a way of explaining and quantifying the public perception, not because you are in the tank for the Giants. It wasn't a balanced analysis, but you weren't trying to do a balanced analysis in this case.

The interesting point, to me, is that even with the intentionally unbalanced analysis, you still came out with the Patriots slightly ahead. In other words, people who want to pick the Giants need to look somewhere other than straight DVOA for their reasons.

It actually looks quite close to me. 19% in weighted DVOA isn't much, and the specific match ups may favor the Giants. Unlike the 2007 season, a Giants win this time wouldn't be a shocker, just a run-of-the-mill upset by a slight underdog.

103 Re: Conference Championship DVOA Ratings

I think the most important factor is:

"The Patriots didn't start a former special-teamer waived by the Jets at SS, a one-and-done safety waived by the Raiders at CB, a spread-offense QB drafted as a WR/KR/PR as the nickelback/Nº1 CB, a former first-round, Pro Bowl sophomore CB at FS because of the most catastrophic sophomore slump in history for a defensive player, and a collection of ex-practice squad members and waiver-wire pickups rotating at LB, and their fifth best WR is the special teams captain who has only one reception in his curriculum SINCE HIGH SCHOOL... in 2007".

Seriously, after this season, can we all once and for all ditch the meme "Bill Belichick: Staff Genius"? This season showed what a disaster the late 2000s were for the Patriots' front office. Terrible, just... terrible.

I mean, their last decent pick was Mayo, right? Decent, at least. Their last great pick was Mankins. That was in 2005. Since then, this is what they got:

2006: Laurence Maroney and Chad Jackson in the first two rounds.
2007: Brandon Merriweather and nobody else.
2008: Mayo, pick Wheatley (suppossed to be a "steal") with their second-rounder and Wilhite in the fourth, pick a QB in the third (??????) no longer in the team. Pick Shawn Crable, who goes on to get his knee exploded in consecutive seasons before the regular season.
2009: Trade away from the first round. Pick Chung, Vollmer and botch the Darius Butler and Ron Brace picks. Botch the Tate pick (JUST before the Steelers pick Wallace).

All in all, along with the signings of past-their-prime free agents like Adalius Thomas, Joey Galloway, Fred Taylor, Torry Holt, Alge Crumpler, Albert Haynesworth and Chad OchoCinco, doesn't this rap sheet merit at least a few scouts being fired?

tl;dr: I realized that at some point during the AFCCG the Patriots were playing Edelman as a CB, Slater as a WR, Steerlin Cooper at CB and Ihedigbo as SS. This... shouldn't happen to a Super Bowl team

107 Re: Conference Championship DVOA Ratings

Belichick is not very good at picking the groceries, but you gotta admit that the man can cook! That ingrediets on defense are Wilfork and some cans of Spam with a little Mayo (heh), and while the meal ain't Le Bernardin it's actually edible. And on offense, he and his staff have turned some high quality, if unconventional, ingredients into a subtle masterpiece of molecular gastronomy. Bill Belichick is Ratatouille -- a rat who cooks with lightning and saffron.

179 Re: Conference Championship DVOA Ratings

Actually, I would take Moore, Arrington, and Edelman over Brown, Poteat, and Moreland.

Edelman has looked somewhat functional as a nickelback, more so than Troy Brown did. (Sorry, Troy...)

Arrington is an average, or possibly slightly above average, zone corner...clearly better than Poteat, who was pretty much the definition of replacement player.

And Moore, while not great, at least deserves to be on a football field at times. The same can not be said of Moreland.

108 Re: Conference Championship DVOA Ratings

The one thing the Patriots do very well, though, is find idiot franchises to mortgage their futures. As a Colts fan, I harbor bitter resentment towards all of the teams that, over the years, have given up high draft picks to the Patriots for some short-term nitwit move like trading for Deion Branch.

117 Re: Conference Championship DVOA Ratings

This subject has been tackled by various mediums pretty thoroughly, and I think your overall point, that Belichick clearly isn't a personnel wizard, is valid. But I think you went jussst a little bit over the top.

For one, you bizarrely stopped at 2009 when recounting drafts even though there's been two whole drafts since then. To say that Mayo was the last "decent" pick (and I'm assuming you're not exclusively talking about defensive picks, since you then referenced Logan Mankins) is to choose to ignore the 2010 draft, which yielded the league's best all-around tight end, another top five or six tight end, the team's best run-stuffing linebacker, the team's punter for the next fifteen years, a formidable DE, and a corner who despite his skills having considerably dropped off this year, started the Pro Bowl as a rookie. The 2010 draft was a grand slam by the front office. Hard to gauge 2011, since it was eight months ago, but Solder had a solid B+ of a rookie year, at least. Also, it should be noted that the team's 2nd, 4th, and 7th round picks during the 2007 draft were traded for Wes Welker and Randy Moss, pretty damn good front office decisions.

There have been some poor free agent signings, to be sure, but other than Thomas, none of the people you mentioned were relied upon immediately to have a massive impact on the team's fortunes. Besides, Alge Crumpler doesn't belong in that group, as he had a very solid 2010 as a run blocker and part-time pass catcher. There were certainly many fans sad to see him released.

Again, you're correct to a degree, but I feel like saying "Can we ditch the "Bill Belichick: Staff Genius" meme?" is the new "SNL sucks now." Anyone who bothers to comment on FO articles has likely heard or read a version of that sentence like 25 times over the last two years.

121 Re: Conference Championship DVOA Ratings

You might want to consider the TE position.

I mean, their last decent pick was Mayo, right? Decent, at least.


Their last great pick was Mankins.

Nope. Again, you might consider the TE position.

I always wonder at the people who criticize the Pats because Belichick isn't playing the guys that they expect him to play.

As for free agents, Andre Carter was named a starter in the Pro Bowl.

Look, the team really did go 13-3. I'm sorry they didn't do so using the players that you think they should have used, but really, a 13-3 team that makes the Super Bowl is a pretty good team. They didn't accidentally stumble into the Super Bowl with a posse of clowns.

If Sterling Moore is waived by the Raiders and then makes a crucial play to get the Pats into the Super Bowl, how is that supposed to reflect poorly on Belichick?

119 Re: Conference Championship DVOA Ratings

I really do not understand how DVOA did not rate the 49ers significantly worse on offense -- very large percentage of their total yards on 2 plays, and seemingly incredibly few DVOA-successful plays the rest of the time. Still count for 14 points, but I thougth DVOA murders that kind of all-or-nothing production. Perhaps their seemingly high special teams DVOA is the flip-side of my question -- 2 disastrous plays, but otherwise decent punt coverage and excellent kickoffs and kickoff returns?

123 Re: Conference Championship DVOA Ratings

The Giants gave up a lot of sacks on offense. They were sacked six times! Each one of those counts as a negative play.

The discussions today make me wonder at the difference between how impressions are formed psychologically and how they are formed when data-driven. DVOA is data-driven, but since it conflicts with the impressions, people are baffled and adhere to their impressions over the statistics.

125 Re: Conference Championship DVOA Ratings

very large percentage of their total yards on 2 plays

No - a very large percentage of their total passing yards came on 2 plays. An even larger percentage of their total yards came on a dozen or so runs up and down the field.

162 Re: Conference Championship DVOA Ratings


The 49'ers had homefield advantage in favorable weather conditions for them. Although they have a superior defense to the Giants they were not the better team. If they played that game in New York I would bet it would not have been close.

The Ravens definitely played better that day than the Patriots but teh Pats are the better team with better talent.

166 Re: Conference Championship DVOA Ratings

"The Ravens definitely played better that day than the Patriots but teh Pats are the better team with better talent."

VOAf would tend to disagree. It tells me that the Patriots did, on average, about 10% better than league average on all their plays, while the Ravens did, on average, about 2% worse. In other words, the Patriots played better that day.

"Although they have a superior defense to the Giants they were not the better team. If they played that game in New York I would bet it would not have been close."

No, without the rain to slow down the Niners rush, Eli would have been absolutely killed in the first quarter.

173 Re: Conference Championship DVOA Ratings

It is also impossible to say how much rain and windy conditions hurt Smith's passing. Both QB's played in the same conditions. Why do you assume that better conditions would have favored only one side?

Besides, I am of the belief (and Tom Brady, among others, agrees) that rain actually HELPS the passing game. It slows down the pass rush, giving the QB more time. And though it makes it harder for the receivers to make cuts, that goes for DB's too. And if it's hard for both sides to make their cuts, the advantage goes to the offense, since they know where the play is supposed to go, whereas the defense has to try to react.

I know the Giants have a good pass rush. However, the Niners have a good pass rush AND fantastic ILB's to clog the quick throw and hot-read routes.

Without the sloppy field to slow down the Niners' rush, I don't think Eli has time to find Cruz abusing Rodgers. Granted, a more aggressive Giants rush might have made Smith's life even more unpleasant, but I think in better conditions, both passing games get worse (if possible). In which case the better defense may well prevail.

171 Re: Conference Championship DVOA Ratings

The VOAf to DVOA conversions for the Pats/Ravens confuses me. NE gets +21%, but baltimore gets +12%, even though NE had a much higher DVOA going in. The pats gain 16% on offense when baltimore's defense was around -12%, but the Raven's defense only gains 23% when the Pat's offense was around 40%?

188 Re: Conference Championship DVOA Ratings

Interesting that the Patriots had a better defensive game, relatively speaking, than the Giants did, which is the opposite of what I would have thought from watching. The Giants defense I thought looked very intimidating...was SF's offense generally weak this year?

Also interesting that the Patriots' offensive DVOA was so high. Most commentators and fans seem to think that was one of the worst offensive performances of the weekend, but apparently Baltimore's defense really is that good.

191 Re: Conference Championship DVOA Ratings

Well, the Patriots were able to move the ball pretty easily vs. the Ravens until they got to the red zone. When Lewis and Reed and Pollard don't have as much ground to cover, they're really good (at least anecdotally) against the pass. I don't claim to be a DVOA expert, but I think driving the field and repeatedly kicking field goals is probably good for DVOA.

194 Re: Conference Championship DVOA Ratings

Sounds about right...or at least, driving down the field against a team that doesn't generally let opponents drive down the field. Still, I would have expected the multiple red-zone failures to "count" a little more, since yards don't matter if they don't lead to points. Then again, Baltimore has an outstanding red-zone defense.

Overall, I'm pleased that the DVOA shows that the Pats measurably outplayed the Ravens. My impression from watching the game and the current media comments was that it was a "knife-edge" type game where the final outcome was due primarily to luck.