49ers, Bills Easily Pace Weighted DVOA Ratings
NFL Wild Card - After wild-card weekend, the San Francisco 49ers maintain their No. 1 spot in our weighted DVOA ratings (the hottest team right now) while the Buffalo Bills maintain their No. 1 spot in our total DVOA ratings (for the full season). Both teams went up after the wild-card games, and both San Francisco and Buffalo now have weighted DVOA over 40%. That's probably not a surprise when it comes to San Francisco, but it's more of a surprise when it comes to Buffalo.
The 49ers have won 11 games in a row and have positive DVOA in each one of those wins. In eight of those 11 wins, San Francisco had DVOA over 45%, including this week in the wild-card win over Seattle. Despite starting a seventh-round rookie quarterback, the 49ers are now No. 2 in weighted offense behind Kansas City. Surprisingly, they have dropped to fifth in weighted defense—it was not a great defensive performance against Seattle, as the Seahawks had a 48% success rate—but the 49ers are less than three percentage points behind No. 1 New England. The 49ers are even in the top 10 for weighted special teams. Because they are No. 1 in weighted DVOA with a sizeable gap between them and the rest of the NFC, the 49ers are now our Super Bowl favorites.
The Bills spent a lot of Sunday afternoon tripping over their own feet but that game with Miami featured a lot of non-predictive randomness. Despite a close victory, the Bills had a single-game DVOA of 59.4% and a Post-Game Win Expectancy of 98%. The Bills outgained the Dolphins 5.9 yards per play to 3.3 yards per play and ran more plays. Each quarterback threw two interceptions, but note that the two Miami interception returns gained 78 yards while the two Buffalo interception returns gained 10 yards. There is a lot of randomness to interception returns, and they aren't really a great indicator of how good an offense is going to play in the future. The Bills also fumbled four times and did recover three of those fumbles themselves, which is a little bit of luck, but the one fumble they lost was a touchdown for the Dolphins, which is definitely not good luck. A lot of the NFL commentariat came away from Sunday's game with the belief that the Bills are in real trouble, when DVOA thinks that the underlying play-by-play details show a team that outplayed the Dolphins in the way we expected when we found out Miami would be starting a third-string quarterback. The Cincinnati Bengals are a very good team, now fourth in weighted DVOA. They have a real chance to win next week but our numbers say that Buffalo is still the best team in the AFC.
If weighted DVOA is to believed, our Big Six from most of this season is now more of a Big Five. There's a good gap between Philadelphia at No. 5 and Dallas, the next playoff team, at No. 7. The other three playoff teams are much lower in weighted DVOA. The Jaguars drop to No. 14 while the Giants move up a spot to No. 17. Tampa Bay is all the way down at No. 26—much lower than their No. 18 rank for the entire season—but of course they can change that a bit with a nice healthy win over the Cowboys tonight.
As always, the following rules apply for postseason DVOA ratings:
- All 32 teams are ranked, whether they made the playoffs or not.
- Teams are ranked in order of weighted DVOA, not total season DVOA. Since weighted DVOA is meant to lower the strength of older games, these ratings do not include Weeks 1-4, and Weeks 5-12 are somewhat discounted.
- The ratings listed do not include the adjustments used in the ratings for our playoff odds report. Right now, Philadelphia is adjusted to account for the two Gardner Minshew games and Tampa Bay is slightly adjusted because it sat starters for part of Week 18.
- Only weighted DVOA is listed for offense, defense, and special teams. Total DVOA is also listed, but one game doesn't change much in an 18-game sample so these ratings will be similar to those from the end of the season.
- Teams which did not play in the wild-card round are treated as if they had a bye week. (That includes both the 18 non-playoff teams and the two teams with byes.)
* * * * *
To save people some time, we remind everyone to put their angry troll hatred into the official zlionsfan angry troll hatred Mad Libs form:
<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>
If you are new to our website, you can read the explanation of how DVOA is figured here. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.
We will also adjust these ratings with an article after Monday Night Football between Dallas and Tampa Bay.
Teams in yellow are still alive in the playoffs. Teams in gray lost this past weekend.
| RK | TEAM | WEI. DVOA |
LAST WEEK |
W-L | WEI OFF DVOA |
RK | WEI DEF DVOA |
RK | WEI S.T. DVOA |
RK | TOTAL DVOA |
RK |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | SF | 41.4% | 1 | 14-4 | 25.4% | 2 | -11.7% | 5 | 4.3% | 7 | 29.2% | 2 |
| 2 | BUF | 40.5% | 2 | 14-3 | 19.8% | 4 | -12.8% | 4 | 7.9% | 1 | 36.5% | 1 |
| 3 | KC | 29.7% | 3 | 14-3 | 27.5% | 1 | -2.9% | 14 | -0.7% | 24 | 23.0% | 4 |
| 4 | CIN | 25.3% | 4 | 13-4 | 20.0% | 3 | -5.4% | 11 | -0.1% | 22 | 18.4% | 5 |
| 5 | PHI | 21.9% | 5 | 14-3 | 11.6% | 7 | -5.9% | 10 | 4.5% | 6 | 25.3% | 3 |
| 6 | DET | 19.8% | 6 | 9-8 | 17.1% | 5 | 2.0% | 20 | 4.7% | 4 | 7.7% | 8 |
| 7 | DAL | 13.4% | 7 | 12-5 | 3.4% | 11 | -8.5% | 7 | 1.4% | 19 | 18.0% | 6 |
| 8 | BAL | 10.6% | 8 | 10-8 | -3.7% | 19 | -10.4% | 6 | 3.9% | 11 | 17.5% | 7 |
| 9 | GB | 9.8% | 9 | 8-9 | 5.6% | 9 | -0.9% | 17 | 3.3% | 13 | 3.6% | 12 |
| 10 | CLE | 9.2% | 10 | 7-10 | 2.7% | 13 | -2.5% | 16 | 4.1% | 9 | 5.2% | 10 |
| 11 | NO | 8.2% | 11 | 7-10 | -6.9% | 23 | -13.2% | 2 | 1.9% | 17 | -1.7% | 19 |
| 12 | PIT | 7.0% | 14 | 9-8 | 2.9% | 12 | -7.4% | 9 | -3.3% | 29 | 0.9% | 14 |
| 13 | MIA | 5.4% | 12 | 9-9 | -1.9% | 17 | -4.9% | 12 | 2.4% | 16 | 7.1% | 9 |
| 14 | JAX | 5.4% | 13 | 10-8 | 7.1% | 8 | 6.3% | 26 | 4.5% | 5 | 2.2% | 13 |
| 15 | LAC | 3.9% | 16 | 10-8 | -2.6% | 18 | -2.5% | 15 | 4.0% | 10 | -0.6% | 17 |
| 16 | SEA | 3.0% | 15 | 9-9 | -0.9% | 14 | 3.8% | 23 | 7.8% | 2 | 4.4% | 11 |
| 17 | NYG | 2.6% | 18 | 10-7-1 | 13.1% | 6 | 10.2% | 29 | -0.3% | 23 | -3.0% | 21 |
| 18 | WAS | 1.4% | 17 | 8-8-1 | -10.0% | 26 | -8.5% | 8 | 2.9% | 14 | -4.9% | 22 |
| 19 | ATL | -3.0% | 21 | 7-10 | 3.6% | 10 | 10.5% | 30 | 3.8% | 12 | -2.8% | 20 |
| 20 | NYJ | -6.6% | 19 | 7-10 | -16.3% | 29 | -12.9% | 3 | -3.3% | 28 | -0.5% | 16 |
| 21 | LAR | -6.7% | 22 | 5-12 | -4.7% | 21 | 3.6% | 22 | 1.6% | 18 | -11.0% | 24 |
| 22 | CAR | -7.4% | 23 | 7-10 | -4.2% | 20 | 7.3% | 27 | 4.1% | 8 | -13.9% | 28 |
| 23 | NE | -8.3% | 20 | 8-9 | -10.6% | 27 | -14.4% | 1 | -12.1% | 32 | -0.3% | 15 |
| 24 | LV | -10.2% | 24 | 6-11 | -1.4% | 15 | 11.3% | 31 | 2.6% | 15 | -11.5% | 26 |
| 25 | DEN | -13.8% | 26 | 5-12 | -12.2% | 28 | -0.4% | 18 | -2.0% | 25 | -11.3% | 25 |
| 26 | TB | -14.8% | 25 | 8-9 | -1.7% | 16 | 5.5% | 25 | -7.6% | 31 | -0.7% | 18 |
| 27 | TEN | -15.8% | 27 | 7-10 | -9.2% | 25 | 3.5% | 21 | -3.1% | 27 | -9.4% | 23 |
| 28 | HOU | -19.5% | 30 | 3-13-1 | -30.2% | 31 | -3.9% | 13 | 6.8% | 3 | -26.7% | 31 |
| 29 | MIN | -19.6% | 28 | 13-5 | -5.0% | 22 | 9.0% | 28 | -5.7% | 30 | -13.8% | 27 |
| 30 | ARI | -21.8% | 29 | 4-13 | -17.8% | 30 | 4.8% | 24 | 0.8% | 20 | -22.4% | 29 |
| 31 | CHI | -31.9% | 31 | 3-14 | -8.0% | 24 | 24.0% | 32 | 0.1% | 21 | -26.6% | 30 |
| 32 | IND | -37.3% | 32 | 4-12-1 | -34.4% | 32 | 0.4% | 19 | -2.5% | 26 | -32.8% | 32 |
Click here for a look at full-season DVOA with offensive and defensive splits.
Next, let's look at all the single-game ratings from wild-card weekend. Other than Buffalo-Miami, the results are probably close to what you are expecting. The Giants had a small but healthy advantage over the Vikings. The Jaguars-Chargers and Bengals-Ravens games were close. What's interesting with these games is how Post-Game Win Expectancy disagrees with DVOA. Post-Game Win Expectancy, for those aren't familiar, is a formula that asks: based on the play-by-play results of this game, how often would we expect each team to win? It accounts for VOA (without opponent adjustments), the total number of plays for each team, and penalties (including declined and offsetting).
As you might imagine, the 98-yard fumble return touchdown for Cincinnati makes the DVOA between the Bengals and Ravens very close. DVOA does not count the fumble return, because fumble recovery is random. It doesn't know the specifics of the fumble, the way that Tyler Huntley tried to go over the pile and had the ball slapped out of his hand, but remember there's a chance that the ball goes the other way into the end zone and either Baltimore recovers for a touchdown or Cincinnati recovers for a touchback and has to start its offense on the 20. That Sam Hubbard touchdown isn't necessarily predictive of how the Baltimore offense is going to play in the future (if it had a future). Despite this touchdown not getting credit in DVOA, Cincinnati came out of Sunday's game with a slightly higher DVOA than Baltimore. However, they only had PGWE of 33% because the Ravens ran more plays (66 to 53) while the Bengals had more penalties (7 to 5).
On the other hand, the Jaguars had a higher PGWE but lower DVOA in Saturday night's big comeback. Jacksonville ends up with 66% PGWE but a negative DVOA (and VOA before opponent adjustments). The main reason for this is again penalties, as the Chargers had nine penalties and the Jaguars only four. This overcame the fact that the Chargers had the better passing DVOA and the Jaguars the better rushing DVOA. Stronger passing usually indicates a win more than stronger rushing.
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Comments
31 comments, Last at 17 Jan 2023, 1:04pm
#1 by Tutenkharnage // Jan 16, 2023 - 2:02pm
I told people yesterday that the Bills would have a 90% win expectancy and that DVOA would love what they did yesterday, but a few people didn’t want to listen. Bills-Dolphins was a massacre disguised as a competitive game. Miami got destroyed everywhere except the scoreboard, and before anyone points out that the scoreboard is ultimately what counts, I’ll point out that contrary to all the gushing by Nantz and Romo yesterday, the Bills were up two scores in the fourth quarter and won the game.
P.S. I don’t like the way the Bills attacked the Dolphins on offense, just to be clear! I don’t like their carelessness with the ball. I don’t like their tendency to try to “punch a hole in the sun,” as Tanier often says. I don’t expect them to win the title if they keep playing like this. But they outplayed Miami by a wide margin yesterday, and anyone who thinks otherwise is just blinded by the scoreboard.
#6 by IlluminatusUIUC // Jan 16, 2023 - 2:40pm
Yeah, its a solid take filter to see who tries to blame Buffalo's defense for the final score. Miami's drives:
- 3 and out
- Interception on the 2nd play
- 6 plays, 10 yards
- Buffalo kicks out of bounds - 10 plays, 38 yards, FG
- Buffalo throws INT, gives up 50 yard return - 8 plays, 18 yards, FG
- Buffalo punts, returned 50 yards, 4 plays, 8 yards, FG
- Buffalo throws INT, Miami starts on Buffalo's 18 - 4 plays, 18 yards, TD+2
- Hail Mary, Halftime
- 3 and out (Buffalo then gives up a fumble 6)
- 3 and out
- Interception on the 3rd play
- 4 plays, 13 yards
- 11 plays, 75 yards, TD
- 3 and out
- 7 plays, turnover on downs
231 net yards and literally three drives that crossed midfield (the others scores started in Buffalo territory)
The blame (in as much as anyone requires blame in a winning effort) should land on Allen. The Dolphins defense consistently dialed up pressure and left our wideouts on islands up the field. Allen "took what the defense gave him" which was 50/50 balls, but they failed at a high enough rate that Buffalo couldn't boat race Miami.
#17 by Displaced Bill… // Jan 16, 2023 - 5:21pm
Really wish I knew what was going on on the inside, I'm curious if the constant deep throws is an Allen thing or a Dorsey thing? That whole game I was thinking, "can I see a drive that takes 7 or 8 plays that consistently net 6+ yards instead of chucking a bombs all day?" I like that they CAN do that but would also like to know that they CAN grind out drives that chew up clock and give their D a rest...
#29 by JacqueShellacque // Jan 17, 2023 - 12:33pm
I'm torn here: I'm with you in almost all of these points, and yet this might also come across as dissembling with statistical information that can give analytical approaches a bad name. Miami had the ball near midfield, with lots of time, down 3, giving them plenty of opportunity to tie or win. At that specific point, no matter how one might try to quantify such a thing, I don't think the Bills had a 90% chance of winning. There's a big risk of hindsight bias here. It should not have been 'obvious' on that last 4th down play that Miami would almost certainly lose, that defies common sense and seems to be missing the conditional aspect of that moment in the game. The point of analytics should be to remove fallacious thinking, not substitute one set of flaws for another. I don't have a problem with DVOA loving what the Bills did, but I do think there may be an issue with using DVOA and win expectancies to characterize the game as one-sided. There's a risk in being smart, but not wise, and that sort of criticism of analytical methods by coaches and players to me is well founded.
I think this game deserves thorough analysis, because it could very well be the template for what outmatched opponents try to do in the future: court disaster all the time in the hope of creating randomness and conditional probabilities (not expectancies) that are more favorable. For example when Allen lobbed those 2 bombs back to back to Davis late in the game on 2nd then 3rd and 10, what were the probabilities of completing one of those passes? Probably lower, and maybe even much lower, than win probability or expectancy would've given the Bills at that point in the game. So essentially the script was flipped: it was the heavy favorite who were baited or forced into a low-percentage situation.
#10 by Aaron Brooks G… // Jan 16, 2023 - 3:36pm
https://en.wikipedia.org/wiki/Observer_effect_(physics)
The problem with desiring that teams were judged based on a different set of outcomes is that teams would play differently if they knew the judging criteria were going to be different.
#3 by Aaron Brooks G… // Jan 16, 2023 - 2:09pm
On the other hand, the Jaguars had a higher PGWE but lower DVOA in Saturday night's big comeback. Jacksonville ends up with 66% PGWE but a negative DVOA (and VOA before opponent adjustments).
That's not really a surprise. DVOA doesn't really penalize you much if your failures occur on 3-and-outs, but you have some successful drives. Because it's per-play, it sees all those failed drives as relatively harmless so long as you score a TD on a 10-play drive to offset those quick failures.
Whereas it hates TOs, and Jacksonville had a ton of those.
Incidentally -- all year the refrain has been 'what's wrong with the PGWE?'. It seems like it's not been working well. Granted, a post-game win expectation system that completely ignores game state is already something of an oddity.
#8 by Aaron Schatz // Jan 16, 2023 - 3:10pm
I disagree. I think it's been working pretty well. For 2020-2021, the correlation of wins to PGWE was .816. For 2022 regular season, the correlation of wins to PGWE was .783. Pretty close. For the most part, games with around 20% PGWE will be won by that team about 20% of the time, and so on.
#16 by Aaron Schatz // Jan 16, 2023 - 4:00pm
It helps suggest when the "wrong team won," in other words, can we learn from this game going forward that the losing team actually played better than the score indicated and might be the better team going forward.
#24 by Eggolas // Jan 16, 2023 - 9:01pm
Interesting. By “going forward” do you mean during a season or from one season to the immediate next one? I ask for two reasons:
1. Your recent Tweet on Pythagorean results from one season to the next where there were results you found surprising; and
2. The level of player personnel changes between one NFL season and another, except for franchise quarterbacks (which, if a team has a legitimate one, changes everything).
#9 by big10freak // Jan 16, 2023 - 3:28pm
One of my grandsons received odds from one of those sites of 19-1 on the Niners to win the SB after Jimmy went down. I played a small part in that in that we exchanged text messages about Purdy where I gave him my assessment similar to what I did here. Now I did not expect Purdy to go all Fran Tarkenton but expected in a Shanahan scheme plus surrounded by so much talent he could look like a pretty capable player.
He's feeling pretty good. Wonder if I will get a cut if it pays off. Ha, ah
#18 by BigRichie // Jan 16, 2023 - 5:26pm
Which is my exact point. When you have a Monday postseason game you should run it on Tuesdays. Hell with what you've done for years.
I was looking forward to Mike's little notes about each team. (understanding maybe he tossed those into his main article this time, what with the fewer games to cover)
#25 by BlueStarDude // Jan 17, 2023 - 7:50am
But this year's wasn't like any previous year what with the Monday Night Playoff game. You don't run DVOA for the week during the season on Monday, you run in it on Tuesday because that is when all of the week's games have finished and you have all the week's data. Should have done the same here.
#19 by rh1no // Jan 16, 2023 - 6:03pm
I'm pretty surprised at how much DVOA appreciated Cincinnati's performance, because I did not.
A lot of people have commented that the Bengals were lucky to get away with a win yesterday, and I disagree with that assessment. The Ravens did a great job at limiting Cincinnati's possessions amd taking advantage of Cincinnati's defensive miscues. Tyler Huntley had the game of his life, throwing a couple of ridiculous passes to beat perfect coverage, but the Bengals defense ultimately outmatched the back-up.
The Ravens executed their game plan about as well as they could have hoped but still came up short. The Bengals didn't play well, but still played well enough to win. An 8% offensive DVOA seems too high for the Cincinnati performance, but maybe it's that high because the Bengals had only 7 meaningful possessions in the game and three of them were scoring drives of 10+ plays ...
#20 by big10freak // Jan 16, 2023 - 6:45pm
Just glad Baltimore is out of the playoffs. Normally a run first offense and rugged D is my sweet spot. But the Ravens are always jawing, poking and prodding trying to elicit a retaliation, and basically being jerks that detracts from the appeal. Then add in a coach who has a constant aura of being aggrieved and I am checked out.
I always credit Baltimore for playing hard. That team leaves it all on the field. If they would just PLAY. But then maybe it’s the other that makes them do what they do. Just wish it wasn’t so
#22 by rh1no // Jan 16, 2023 - 8:55pm
Can't blame you for feeling that way. Baltimore was twisting ankles and throwing punches from the first drive ... and getting away with it, as usual. At least they got whistled for taunting.
But I still root for the Ravens when they're not playing the Bengals. The Ravens are usually good enough to take a game or two from the Stealers, who are the dirtiest team in the league. "The enemy of my enemy," and all that.
#23 by rh1no // Jan 16, 2023 - 8:55pm
Can't blame you for feeling that way. Baltimore was twisting ankles and throwing punches from the first drive ... and getting away with it, as usual. At least they got whistled for taunting.
But I still root for the Ravens when they're not playing the Bengals. The Ravens are usually good enough to take a game or two from the Stealers, who are the dirtiest team in the league. "The enemy of my enemy," and all that.
#28 by BlueStarDude // Jan 17, 2023 - 11:21am
Aaron, just saw that you're putting up a second DVOA commentary today with MNF included. Perfect. I thought you were going to do something like that, but then your reply above made me think you weren't going to.
#30 by DGL // Jan 17, 2023 - 12:38pm
Tampa Bay is all the way down at No. 26—much lower than their No. 18 rank for the entire season—but of course they can change that a bit with a nice healthy win over the Cowboys tonight.
And once we successfully moor at Lakehurst, it's only a short train ride into Manhattan.
#31 by Oncorhynchus // Jan 17, 2023 - 1:04pm
I find it a little unbelievable that the 49ers have such higher odds of reaching the Superbowl than Philadelphia. I'd think they'd be more evenly matched (each around 38%).
Vegas is still the best at predicting games. There is no model more predictive than the moneyline. If there were, then the person with the model could make a lot of money selling it to Vegas. Right now, the line for the Cowboys @ 49ers is -3.5 (or -4) and for the Giants @ Eagles is -7.5. That translates to a win probability of around 64% for the 49ers and around 74% for the Eagles. FO playoff odds for reaching the divisional round are 73.5% for the 49ers and 72% for the Eagles. That's shifting the line towards the 49ers by 3 points over Vegas's expectations and down about a half a point for the Eagles.
Same thing for the Bills too - though I think the model is little better calibrated here. 68% is in line with the -5 point spread. Their inflated SB odds probably has more to do with the JAX @ KC odds being a little too low for KC.
In both cases the #2 seed appear to have the harder path to the SB. The 49ers have to beat the 7th ranked in WEI DVOA at home and then (most likely) the 5th ranked team away (or less likely the 17th ranked team at home). The Bills have to beat the 4th ranked team at home and then (most likely) the 3rd ranked team at a neutral site (with some chance having to beat the 14th ranked team at home).