2013 DVOA Projections
by Aaron Schatz
The time has come for our annual preseason DVOA projections, updated from the projections that gave us the season forecasts in Football Outsiders Almanac 2013. Our DVOA projections suggest a Packers-Patriots Super Bowl, the same matchup forecast by last year's DVOA projections. The mean win projections suggest a Seahawks-Broncos Super Bowl, the same matchup forecast by last year's actual in-season DVOA ratings.
We must start with the requisite link to an explanation of DVOA. For anyone new to our site, DVOA stands for Defense-adjusted Value Over Average and measures a team's performance on every play of the season compared to league average in the same situation, adjusted for opponent. I know a lot of people may be coming here from various message boards and this is just going to look like a jumble of pointless numbers. Trust me, there is a method to the madness, and over the past dozen seasons past DVOA ratings -- as well as these multivariable-based DVOA projections -- have been a far more accurate predictor of future performance than wins or points.
Offense, defense, and special teams DVOA are all projected separately using a system based on looking at trends for teams over the past decade. The equations include a number of variables based on performance over the past two seasons in different splits (by down, passing vs. rushing, red zone vs. whole field) plus variables based on recent draft history, injury history, offensive and defensive pace, coaching experience, quarterback experience, and even weather. Strength of schedule was then figured based on the average projected total DVOA of all 16 opponents for 2013, rather than based on last year's performance.
The major difference between this year's preseason forecast and the forecast produced three months ago for Football Outsiders Almanac 2013 is unfortunately caused by an error. As one step towards producing team offensive projections, I create a quarterback projection that does not take into account any team variables. (This allows us to account, to give one example, for the fact that Carson Palmer should be expected to improve the Arizona offense.) The variable I used for Russell Wilson in the Seattle projection was pasted incorrectly, leading to a Seattle projection that was a little bit lower than it should have been. Fixing this moves Seattle a bit higher, making the Seahawks one of the top projected teams in the league. Don't worry San Francisco fans, we're still projecting you into the wild card.
The quarterback variable is also an important one explaining why a couple of teams are lower than previously projected. Despite the recent success of rookie quarterbacks, our "quarterback projection without team variables" numbers projected E.J. Manuel a little bit below Kevin Kolb and Terrelle Pryor significantly below Matt Flynn. Geno Smith and Mark Sanchez are pretty much a wash.
If your response to that paragraph was "the idea that Kevin Kolb might have been better than E.J. Manuel is insane," then please remember the following: The numbers we are presenting here are exactly what the projection system spit out. As we say every year: "A few of them will look strange to you. A few of them look strange to us." As always, the offensive projections come out in a wider range than defensive projections because offense performance tends to be easier to predict (and more consistent from year to year) than defensive performance. If you are looking for subjective projections, tomorrow we will be running our usual staff predictions article where we all talk about where we think the numbers are wrong.
The first postseason odds report of the 2013 season is also online, and I've added the playoff odds and Super Bowl championship odds to the table below. The mean wins forecast produced by our season simulation seems very conservative. In fact, this forecast is even more conservative than the forecast you found in Football Outsiders Almanac 2013 because due to time constraints, I did not have the opportunity to adjust our win projections to better reflect the recent historical distribution of wins in the NFL (fewer teams between 7-9 and 9-7 than you would expect from a normal distribution, and more teams around 12-4 or 4-12). However, this kind of conservative forecast generally leads to smaller errors than a forecast that looks more like a real set of final standings with the best team around 14-2 and the worst team around 2-14. Obviously, the best team in the league will likely have more than 11 wins, and the worst team will have more than 10 losses.
Surprisingly, despite slightly more conservative projections for mean wins, the new forecast raises the playoff odds of the top teams; we now have four teams with over 80 percent chance of making the playoffs instead of just the two we listed in Football Outsiders Almanac 2013.
Projected division champions are colored in light yellow. Projected wild card teams are colored in light blue.
| TEAM | TOTAL DVOA |
TOTAL RANK |
MEAN WINS |
OFFENSE DVOA |
OFF. RANK |
DEFENSE DVOA |
DEF. RANK |
SPECIAL DVOA |
S.T. RANK |
SCHED | SCHED RANK |
PLAYOFF ODDS |
S.B. WIN ODDS |
| NE | 27.4% | 1 | 10.4 | 24.1% | 2 | 1.2% | 20 | 4.6% | 2 | 0.4% | 17 | 86.4% | 16.4% |
| GB | 23.5% | 2 | 10.3 | 25.4% | 1 | 0.0% | 15 | -2.0% | 28 | 0.2% | 20 | 80.6% | 12.4% |
| SEA | 21.0% | 3 | 10.6 | 18.3% | 4 | -2.4% | 9 | 0.3% | 14 | -3.0% | 25 | 82.7% | 11.6% |
| DEN | 19.8% | 4 | 10.7 | 21.2% | 3 | 1.3% | 21 | -0.1% | 16 | -4.5% | 30 | 88.6% | 13.5% |
| WAS | 18.6% | 5 | 10.0 | 16.0% | 6 | -3.2% | 8 | -0.6% | 20 | -0.8% | 22 | 73.2% | 9.0% |
| CAR | 15.5% | 6 | 9.4 | 8.8% | 9 | -6.0% | 3 | 0.8% | 11 | 1.5% | 11 | 59.6% | 5.1% |
| SF | 14.9% | 7 | 9.4 | 10.6% | 8 | -3.5% | 7 | 0.8% | 10 | 1.1% | 14 | 58.9% | 5.5% |
| BAL | 12.4% | 8 | 9.3 | 11.2% | 7 | 0.0% | 16 | 1.2% | 9 | 1.5% | 12 | 63.5% | 5.5% |
| NO | 7.7% | 9 | 8.5 | 16.3% | 5 | 7.4% | 31 | -1.2% | 23 | 2.7% | 4 | 40.1% | 2.3% |
| HOU | 5.9% | 10 | 9.1 | 3.6% | 13 | -2.2% | 10 | 0.2% | 15 | -4.2% | 29 | 67.2% | 4.1% |
| CIN | 5.3% | 11 | 8.8 | 2.4% | 14 | -0.6% | 12 | 2.3% | 4 | -0.2% | 21 | 53.5% | 2.9% |
| TB | 4.3% | 12 | 8.2 | 1.7% | 15 | -5.5% | 4 | -3.0% | 30 | 3.3% | 3 | 34.0% | 1.5% |
| ATL | 4.1% | 13 | 7.7 | 6.6% | 11 | 2.4% | 24 | -0.2% | 17 | 6.3% | 1 | 25.3% | 1.0% |
| PIT | 3.0% | 14 | 8.6 | -4.7% | 19 | -9.3% | 2 | -1.5% | 26 | -1.1% | 23 | 48.7% | 2.1% |
| NYG | 2.4% | 15 | 8.1 | 6.7% | 10 | 1.1% | 19 | -3.2% | 31 | 2.3% | 6 | 32.4% | 1.5% |
| CHI | -1.1% | 16 | 7.9 | -15.6% | 30 | -9.6% | 1 | 5.0% | 1 | 1.5% | 10 | 29.5% | 1.2% |
| TEAM | TOTAL DVOA |
TOTAL RANK |
MEAN WINS |
OFFENSE DVOA |
OFF. RANK |
DEFENSE DVOA |
DEF. RANK |
SPECIAL DVOA |
S.T. RANK |
SCHED | SCHED RANK |
PLAYOFF ODDS |
S.B. WIN ODDS |
| DET | -2.6% | 17 | 7.5 | 5.1% | 12 | 3.6% | 27 | -4.1% | 32 | 1.9% | 8 | 22.8% | 0.5% |
| PHI | -3.5% | 18 | 7.8 | -3.9% | 18 | -0.1% | 14 | 0.4% | 13 | 0.3% | 18 | 27.2% | 0.7% |
| DAL | -6.8% | 19 | 7.0 | -1.3% | 16 | 4.6% | 29 | -0.9% | 22 | 0.6% | 16 | 14.9% | 0.3% |
| IND | -7.1% | 20 | 7.5 | -3.2% | 17 | 3.5% | 26 | -0.4% | 19 | -5.3% | 32 | 31.1% | 0.6% |
| NYJ | -7.1% | 21 | 7.3 | -13.9% | 28 | -5.4% | 5 | 1.4% | 7 | 0.9% | 15 | 26.2% | 0.5% |
| SD | -7.4% | 22 | 7.4 | -6.3% | 20 | 1.8% | 22 | 0.6% | 12 | -3.8% | 26 | 26.7% | 0.5% |
| MIA | -9.0% | 23 | 6.5 | -11.6% | 25 | -3.8% | 6 | -1.2% | 24 | 2.4% | 5 | 14.5% | 0.2% |
| CLE | -10.2% | 24 | 7.0 | -12.0% | 26 | -0.7% | 11 | 1.2% | 8 | 0.3% | 19 | 18.9% | 0.4% |
| ARI | -11.2% | 25 | 6.4 | -9.6% | 22 | 0.8% | 18 | -0.8% | 21 | 1.4% | 13 | 8.6% | 0.2% |
| MIN | -14.8% | 26 | 5.7 | -9.9% | 23 | 2.4% | 25 | -2.5% | 29 | 6.1% | 2 | 4.9% | 0.1% |
| BUF | -14.9% | 27 | 6.5 | -13.5% | 27 | -0.5% | 13 | -1.9% | 27 | 1.6% | 9 | 14.3% | 0.1% |
| OAK | -16.7% | 28 | 6.5 | -19.8% | 32 | 0.6% | 17 | 3.6% | 3 | -2.7% | 24 | 13.6% | 0.1% |
| KC | -16.9% | 29 | 6.8 | -9.3% | 21 | 9.2% | 32 | 1.6% | 5 | -4.0% | 27 | 16.6% | 0.2% |
| STL | -17.3% | 30 | 6.0 | -14.7% | 29 | 2.2% | 23 | -0.3% | 18 | 2.0% | 7 | 5.4% | 0.0% |
| TEN | -17.4% | 31 | 6.5 | -11.4% | 24 | 4.5% | 28 | -1.4% | 25 | -4.0% | 28 | 16.0% | 0.1% |
| JAC | -22.0% | 32 | 6.3 | -17.2% | 31 | 6.2% | 30 | 1.4% | 6 | -4.7% | 31 | 14.1% | 0.1% |
Comments
167 comments, Last at 20 Oct 2013, 12:05pm
#151 by Scott Crowder (not verified) // Sep 07, 2013 - 5:55pm
I'm sure as heck not hating on FO. Ever! I think you make very valid points. But there is a good point being made IMO by the other side too. The NE projections allow for a valid criticism. Teams go 11-5 to 13-3 over long periods and the projections shouldn't be afraid to, well, for lack of a better word, PROJECT that. There really IS an regression here and the fact that the consistently good teams never get a projection that reflects that exposes the problem.
It's not about how bad they might whiff on Seattle, etc. These are not foolproof projections. It's about making them finer tuned.
#80 by Chemdrj (not verified) // Sep 05, 2013 - 3:25pm
I see a lot of negativity or criticism of FO writers for the projections that the formula makes. I understand that they designed the formula but it was not with the intent to one day predict team a over b. (Unless team a is the patriots) I am sure they wish the predictions were 100% accurate they would be rich and could tell everyone to screw off. I think it would be interesting if readers would make a list or post a list of 6 teams that were locks to make the playoffs and the 12 locks to not make the playoffs. The first six locks to miss are not so hard but the last 6 get tougher especially with luck and injuries. I wonder how many of your 12 locks to miss make it.
#100 by Craig (Braintr… (not verified) // Sep 05, 2013 - 7:18pm
If the Vikings can get the top pick with 6 wins, I'll take it. If Ponder improves, they probably take Clowney. If he doesn't they should be able to replace him with a franchise QB. It's unlikely that they get the top pick with 6 wins.
#116 by KK Probs (not verified) // Sep 06, 2013 - 10:45am
I didn't read all of the comments up to this point, but just a note on these predictions (and predictions in general). The way the Football Outsiders system makes a prediction is:
1)Events happen -> 2)Mostly objective, partially descriptive data is recorded -> 3)Data translated into objective DVOA-based, backward-looking analysis system -> 4)DVOA-based backward-looking analysis translated into mostly objective, DVOA-based forward-looking predictions -> 5)Mostly objective translation from DVOA-based predictions to wins -> 6)Find way to state prediction that sounds like it makes sense in the moment and attracts readers and publish -> 7)Review for deficiencies in the system -> 8)Update for next year
The way most other football "analysts" made predictions, and the way they all did before Football Outsiders:
1)Events happen -> 2)Subjective, possibly descriptive data in -> 3)Skipped -> 4)Skipped -> 5)Subjective win predictions created -> 6)Find way to state prediction that sounds like it makes sense in the moment and attracts readers and publish ->7) Mostly or entirely skipped -> 8)Mostly or entirely skipped
When I say mostly objective vs. subjective: :mostly objective: refers to, for example, a difference in the reason why a FO game charter might consider a ball to be incomplete; one might say an overthrow by a QB and another person might say it's a misread by the receiver. "Subjective" refers to Rick Reilly saying what some star player wants him to say on television just so the star player will let Reilly be the first to tweet out the star player's next piece of news. Or, they'll say an offensive tackle is "big," or the team was "Top 5 in offense" by overall yardage, and these things are subjectively incorporated into predictions however they feel like. (Football Outsiders has long since researched the second of those into semi-relevance, and doesn't claim to exactly know the impact of the first.)
I don't look at DVOA as prophecy by any means. But it is a good, mostly objective baseline to look at how the parts of the game that are available to be quantified stack up going into a game or a season.
I agree that asking it to do anymore than that is asking too much. The one criticism I would have is that there is a focus on DVOA and less relevant discussion of "unknown unknowns" than I would like. Bitching about Norv Turner lack of people skills is justified and does affect the game, but as far as I know, Football Outsiders' analysis of how factors like this impact the game is no further ahead than that of any other analyst's, and even possibly behind some of the top local beat writers for a team and a few top national writers.
If you want a sport where the most accurate predictions can be 70%-95% accurate on a regular basis, I encourage you to follow boxing, which probably has 1% or less of the variables of a football game. NFL Football is a relatively unpredictable sport.
#129 by Anonymousse (not verified) // Sep 06, 2013 - 2:55pm
My biggest issue with DVOA at this point is that it has become institutionalized. 5 years ago, if DVOA did something funny, they looked at it and said "Why is it doing that?"
Now, they say "You're wrong."
#119 by MaineRaider // Sep 06, 2013 - 11:40am
It seems some people are really ragging on FO's predictions without appreciation for what this group has accomplished. Who else was doing sabermetrics-type work on football ten or so years ago? I've been reading FO's print and online output since the early days, and I've seen them make astoundingly prescient predictions that went against the common wisdom. They make mistakes, too, but then make annual adjustments in their continuing efforts at improving accuracy. FOA 2012 helped me win both fantasy leagues I played last year, but, more importantly--and surprisingly--my statistical analysis skills in general have improved immensely since I started reading about DVOA, et al. Those skill improvements have helped me at least as much as economics classes have, and I'm finishing my law degree this semester with a focus on business and finance law. Natch.
#121 by PurpleBlood (not verified) // Sep 06, 2013 - 12:44pm
The sabermetrics is nothing but the trading method of finding undervalued players on the sport "market" based on some attributes and parameters. But to make predictions of their future "value" is something completely different and has its limits.
The football is not baseball, because the performance of each player is hugely affected by the performance of other 21 players on the field, including several coaches on sidelines and even referees, not talking about unpredictable factor of "luck"...
Anyway nice try and we will see how accurate it is at the end of the regular season.
#131 by Mostly Anonymous (not verified) // Sep 06, 2013 - 3:05pm
I can't comment on the playoff odds (address below), but they show the odds of which team will get the first pick of the 2011 draft, not 2013 draft.
http://www.footballoutsiders.com/stats/playoffodds
#134 by hi (not verified) // Sep 06, 2013 - 3:38pm
Surprised most the of top ranking systems, not Vegas though, have New England penciled in as the Super Bowl winner this year. I guess they have too much faith in Tom Tom. Suddenly SF and Seattle are getting the cold shoulder. Besides very few read option based teams have ever won the Super Bowl, much less any west coast offense teams in the past decade.
The Denver performance last night has me convinced they could go all the way. All the offensive weapons in New England had last year minus a second competent TE, now showing up in Denver. Not exactly earth shattering revelation by any means.
#137 by Alternator // Sep 07, 2013 - 12:17am
The peak of the AFC (New England and Denver, in some order) match up well to the top of the NFC (Seattle, San Francisco, Green Bay, NFC South team of choice) quite well, but look at the size of those groups between leagues. The top AFC teams are rated with a higher chance of reaching the Super Bowl at all, so of course they also have a higher chance of winning it.
#153 by Duff Soviet Union // Sep 07, 2013 - 6:52pm
The small variance in projected wins doesn't bother me at all. There is a TON of luck in football. A heap of 12-13 win teams are more accurately true talent 10 win teams who had good luck in close games, an easy schedule, a great run with injuries, a disproportionate number of players having career seasons etc. And the reverse is true as well. Over time, these things tend to even out which is why when you project the season 10,000 times records tend to regress to 8-8. Projecting which teams will win 13 games (or 3) usually means projecting which team will be both the best and the luckiest. Maybe you can project the first, but good luck with the second.
#156 by Subrata Sircar // Sep 08, 2013 - 5:39am
Any mathematical system that attempts to model real-world behavior has a tough time with regression to the mean, because it's a powerful force in the real world, but rarely dominates any given individual measurement.
Any objective projection system is going to spit out a compressed range of outcomes because it won't project outliers - that's kind of the point. Outliers occur when everything goes right/wrong and you often catch a break/get hit in the teeth completely out of left field to boot. No objective system will project that to happen, because its chances of being correct in doing so are zero on average.
So complaining that DVOA doesn't project anyone to win 14 games is a bit pointless. Yeah, someone will (and we can probably look at DVOA and the schedule to get reasonable guesses as to who), but damned if we can pick who with any real accuracy.
It's not being conservative - it's the nature of the system.
#167 by jhonh (not verified) // Oct 20, 2013 - 12:05pm
Thanks for a very interesting blog. What else may I get that kind of info written in such a perfect approach? I’ve a undertaking that I am simply now operating on, and I have been at the look out for such info.social bribe bonus|azon rocket 2.0 review