Week 8 DVOA Ratings
by Aaron Schatz
Denver remains on top of this week's Football Outsiders DVOA ratings, and the Broncos still rank as one of the top ten teams in DVOA history despite a bit of a decline in recent weeks. Seattle remains in second place, although the Seahawks' overall rating drops after last night's narrow win over St. Louis. Indianapolis is third, with Carolina a surprising fourth and Cincinnati climbing from 11th to fifth thanks to their 49-9 drubbing of the New York Jets. Chicago climbs two spots to sixth, even on a bye week, although DVOA of course doesn't take into account the effect of injuries (or McCowns) on a team's future peformance. Close behind the Bears are three other NFC contenders, the Saints, 49ers, and Packers, and...
Wait a minute. Where the hell are the 8-0 Kansas City Chiefs?
The shocking answer is tenth place. The Chiefs drop six spots this week from fourth to tenth. Although many power rankings around the Internet will still have Kansas City in first place because they are slaves to ranking teams by win-loss record, it's pretty obvious that the Chiefs are not the best team in the league right now.
The biggest issue is schedule strength. We've pointed this out a few times this year, but the Chiefs' schedule has been phenomenally easy. The Chiefs have played only one opponent this year with an above-average DVOA, the Dallas Cowboys. If we take out the opponent adjustments and look at non-adjusted VOA, the Chiefs actually rank third in the league; the difference between the two ratings is more than 10 percentage points. Each week, as the opponent adjustments in our system become stronger and stronger, the Chiefs see their DVOA rating fall a little bit more. It doesn't help that the Chiefs have been declining a bit, with their three worst games by DVOA coming in the last three weeks. The game against Cleveland comes out as their worst of the year by DVOA, another reason for their big drop this week. In fact, DVOA has Cleveland outplaying the Chiefs this week even if we remove the opponent adjustments. The Browns gained 6.5 yards per play compared to just 4.7 yards per play for Kansas City, but were hurt by third-down efficiency (just 3-for-12) and the fact that they lost their one fumble while the Chiefs recovered theirs. Without opponent adjustments, that comes out to 19.5% DVOA for the Browns and -4.6% DVOA for the Chiefs. With opponent adjustments, it comes out to 28.1% DVOA for the Browns and -23.1% DVOA for the Chiefs.
(By the way, the Chiefs aren't the only team with huge opponent adjustments this season. Denver and San Diego have also had extremely easy schedules this season. Denver's gap between DVOA and VOA is almost as big as Kansas City's, but because the Broncos have been so good, the opponent adjustments don't knock them out of first place in our ratings.)
A couple weeks ago, we looked at where Kansas City stood among the "worst" 6-0 teams of all-time. At that point, the Chiefs weren't particularly low. However, only 14 teams since 1989 have made it to 8-0, and out of those teams, the Chiefs come out near the bottom. (Yes, one team that ranks below them ended up winning the Super Bowl, but remember that team got its best defensive player back from injury right before the playoffs.)
|8-0 TEAMS, 1989-2013|
(Ed. Note: This table originally left out the 2012 Falcons, but they have now been added.)
Despite the slight decline in recent weeks, the Chiefs have been one of the most consistent teams in the league, ranking second in variance. Which team comes out ahead of them? Why, it's the other team with a win-loss record much better than its DVOA rating, the 6-2 New England Patriots. The Patriots drop a slot to 13th this week despite their win over Miami, because their win over Miami has been like almost all their other games this year: slightly above average. Right now, every New England Patriots game falls in a tiny range between -3% DVOA and 15% DVOA except for the Week 5 loss to Cincinnati (-17.8%) and the Week 6 win over New Orleans (26.4%).
The idea of the Patriots as the most consistent team in the league seems ridiculous, because they seem so inconsistent from drive to drive. Nonetheless, the Patriots may not be the enigma that everyone thinks they are. The Patriots are not suddenly a bad team. They seem to be a slightly above-average team that plays most games slightly above-average. Maybe the reason the Patriots are so hard for people to judge is that their performance this year is so different from what we've come to expect from the Patriots in recent years. Their defense ranks sixth in DVOA and has been the most consistent in the league (variance of only 2.0%) despite all the injuries in recent weeks. The offense, on the other hand is just 20th in offensive DVOA (12th in variance). Even stranger, the offense is being carried in part by the running game. The Patriots rank 12th in rushing DVOA but 24th in passing DVOA.
It's hard to determine how to divy up the blame pie for Tom Brady's massive decline this year. Is his lack of accuracy related to the fact that he doesn't trust his new receivers? Is it related to age? An injury to his hand? How much do we blame the new receivers and how much do we blame Brady himself? What we do know is that Brady's year-to-year decline is one of the biggest for any quarterback in history.
For now, let's just look at quarterbacks since 1989 using DVOA ratings. After Week 8, Brady's DVOA rating stands at -16.6%. That's actually lower than replacement level, which is astonishing for a player who has led the league in passing DVOA three times and passing DYAR four times. As of right now, Brady's drop of over 50 percentage points in DVOA trails only Brett Favre's final season in Minnesota as the biggest year-to-year drop in DVOA for a quarterback with at least 300 passes in two straight seasons.
|Biggest Passing DVOA Decline, 1990-2013 (min. 300 passes)|
One interesting thing about this list is that very few of the quarterbacks on the list are particularly old. A huge decline like this doesn't necessarily represent the end of a player's career. It's also worth noting that most of these players did not play a full season in the year listed in our table. There's a small sample size effect going on here, even given our minimum of 300 passes. That should give Patriots fans hope that while Brady may never be the Brady of 2007-2012 again, there's at least a pretty good chance he's going to be better than this over the last half of the 2013 season.
One last aside: Since I went to look if the Chiefs were the worst 8-0 team ever, I figured I should also look to see if the Patriots are the worst 6-2 team ever. No, they are not. They aren't even in the bottom 20. But the actual list of the worst 6-2 teams ever is really quirky. The three worst 6-2 teams in DVOA history were all the Detroit Lions. They had -17.1% DVOA in 2007 (finished 7-9), -10.9% DVOA in 1993 (finished 10-6), and -8.0% DVOA in 1991 (finished 12-4). Two other teams had negative DVOA at 6-2, and both were in 1998: Atlanta at -8.0% DVOA and Oakland at -5.4% DVOA. Those teams went two completely different directions after midseason. Atlanta went on a crazy hot streak, winning its final eight games while raising its DVOA all the way up to 18.8% for the season. They finished with 30.3% weighted DVOA. Oakland went 2-6 in the second half of the season, finishing 27th in the league with -18.3% DVOA.
BEST AND WORST DVOA EVER WATCH
| BEST TOTAL DVOA
THROUGH WEEK 8
|x|| BEST OFFENSIVE DVOA
THROUGH WEEK 8
|x|| WORST TOTAL DVOA
THROUGH WEEK 8
|x|| WORST OFFENSIVE DVOA
THROUGH WEEK 8
|x|| WORST SPECIAL TEAMS DVOA
THROUGH WEEK 8
I've kept these lists to 12 teams each week, but if we extended the lists by two teams each, the 2013 New York Giants would show up as 14th on the "Worst Special Teams Ever" list.
* * * * *
During the 2013 season, we'll be partnering with EA Sports to bring special Football Outsiders-branded items to Madden 25 Ultimate Team. Each week, we'll be picking out a handful of players who starred in that week's games. Some of them will be well-known players who stood out in standard stats. Others will be under-the-radar players who only stood out with advanced stats, including DYAR, Defeats, and our game charting coverage stats for cornerbacks. We'll announce the players each Tuesday in the DVOA commentary article, and the players will be available in Madden Ultimate Team packs the following weekend, beginning Friday night.
The Football Outsiders stars for Week 8 are:
- LT David Bakhtiari, GB: Limited Jared Allen to three hurries with no sacks or QB hits.
- C Alex Mack, CLE: No sacks or hurries allowed vs. NFL's best pass rush.
- LE Robert Quinn, STL: Three sacks, three QB hits, and TFL. (We also assume there were some hurries, but we haven't had a chance to chart that yet.)
- ROLB Marcus Benard, ARI: Sack, three QB hits, and two hurries.
- SS Troy Polamalu, PIT (Limited Edition): Sack, INT, PD, and TFL.
Most of you are probably asking: Where is Calvin Johnson? Well, in order to provide some variety in the game, the Madden 25 folks have asked us not to choose players who have been featured as either special elite Football Outsiders stars or Madden 25 Team of the Week stars in recent weeks. This week, that prevented us from choosing a lot of players that we were considering otherwise, including not just Megatron but also Andy Dalton and Sean Lee. And of course we can't choose players who made this week's Madden Team of the Week, and this week those players include players with high DYAR like Andre Ellington, Marvin Jones, Jordy Nelson, and Kenny Stills. Other players we considered this week who didn't make the cut: Mike Tolbert, Joe Thomas, and Chris Long.
* * * * *
All 2013 stat pages are now updated or will be updated in the next few minutes, including snap counts, playoff odds, and the FO Premium database.
* * * * *
These are the Football Outsiders team efficiency ratings through eight weeks of 2013, 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.)
[ad placeholder 3]
This week, we say goodbye to DAVE, our method which combines 2013 performance with our preseason projections. All numbers now represent 2013 only.
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. WEIGHTED DVOA represents an attempt to figure out how a team is playing right now, as opposed to over the season as a whole, by making recent games more important than earlier games.
Because it is early in the season, opponent strength is at only 80 percent; it will increase 10 percent every week through Week 10. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.
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>
- NON-ADJUSTED TOTAL DVOA does not include the adjustments for opponent strength or the adjustments for weather and altitude in special teams, and only penalizes offenses for lost fumbles rather than all fumbles.
- 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).
195 comments, Last at 17 Jul 2019, 8:39pm
#105 by Jon G (not verified) // Oct 30, 2013 - 7:55am
@ and vs. Houston and vs. Buffalo seem most likely. Perhaps week 17 if the Colts play their backups.
I could also see them narrowly beating Cleveland or Arizona.
They wouldn't be favored in any of these games, but I think their best shot is probably vs. Houston at home.
#125 by RickD // Oct 30, 2013 - 12:56pm
Aside from the Colts, none of those opponents are exactly lighting things up. And the Colts game is Week 17, when they might have nothing to play for.
Even if we, say, give them only a 10% chance of winning any of these 8 games in particular, the odds of winning at least one would be 1 - .9^8, i.e. about 57%.
#106 by bucko (not verified) // Oct 30, 2013 - 8:26am
I found the DVOA rankings to be their usual interesting reading and the griping Packer fans can please go to the Yahoo/FoxSports message boards to complain.
The injuries on defense have led to the special teams all but collapsing as guys are now starting who were previously playing on special teams. Not good
#112 by Todd S. // Oct 30, 2013 - 10:11am
Internet nit-picking time!
Yes, the one team that ranks below them ended up winning the Super Bowl, but remember that team got its best defensive player back from injury right before the playoffs
I don't agree with Sanders being their best defensive player that year. You still had Freeney and Mathis on the DL; one or both of those 2 has to rate better. Getting Sanders back really shored up the run defense, which was very helpful for that slate of playoff opponents.
#181 by Aaron Brooks G… // Oct 31, 2013 - 1:23pm
Bob Sanders was probably their most important defensive player. Without him, the secondary couldn't hold long enough for Freeney/Mathis to get to the QB, and their run defense was a crime against humanity.
Shame he was basically Polamalu without the health
#113 by 3Monkies (not verified) // Oct 30, 2013 - 10:39am
It's early, but Denver and Seattle look lot be on a collision course for a SB matchup. Each matches up well against the others offense, Seattle with great CBs and pass rush while Denver's defense is very strong against the run.
Perhaps similar to 2006 Indy-Chi, but with Seattle having better QB and Denver a slightly better defense, factoring in Von Millers return.
#118 by Nathan of Brai… (not verified) // Oct 30, 2013 - 11:56am
You may be correct that the Chiefs are the #10 team in the league.
But what if they defeat Denver twice but the Football Outsiders system still has Denver ranked higher?
Wouldn't that indicate your system has flaws?
Three things I think your ranking system may not account for:
- Strong defensive teams win with smaller margins of victory than strong offensive teams (lots of reason for that: if your O is good, you score fast, so your D is on the field more and more prone to collapses, so you need a 2-3 score lead to ensure a victory; if your D is good, you can feel fairly confident with a 1-point lead and your O just needs to take time off the clock and maybe score a FG if within safe range).
- Strong offensive teams w/ weak defenses tend to lose in the playoffs more often than strong defensive teams w/ weak offenses. Weather? Facing better game-planners?
- Consistency of performance: I could write 3 pages on this, but the point is KC has been extremely consistent at preventing 3rd down conversions, winning the field position battle, and holding onto the lead. Denver may score lots of points, and the defense may feast on teams passing excessively to try to keep up, but their ability to win due to facing weak defenses and turnover-prone QBs (as in the Dallas, NYG, and Redskins games) isn't predictive. They most likely wouldn't win those games in the playoffs.
Whereas the Chiefs have shut teams down in the 4th quarter, not by turnovers, but by having the offense execute a clock-killing drive and having the defense sack and prevent 3rd/4th down conversions. They have done it consistently enough to be well above 2 standard deviations, I would guess. It would be interesting to have you run the numbers.
Bonus point: field position battle.
You guys push the idea that field position is such an important, hidden part of the game that if you face 4th and short deep in your opponent's territory, you are better off going for it than kicking a FG. I don't have to run you through the math of that.
Well, apply that same logic to the Broncs and the Chiefs.
The Chiefs are winning games because their O can consistently convert 1-2 3rd downs. That is enough to allow Colquitt to pin the opponent deep inside their own territory, which increases the % chance to win. When the Chiefs get the ball back, they don't need to convert as many times before scoring. The O/D then work together to get the opponent back in that same position even after KC kicks off after a score.
Denver's O and D do not complement each other. While Denver's O can score plenty of points, it keeps the D on the field longer, risks more giveaways, and doesn't consistently keep the opponent pinned in their own territory.
The things they do contribute to higher FO ratings w/o setting up winning conditions as much as the Chiefs do.
I realize this is all just assertion. We'll have to see what happens in the 2 head-to-head games first.
I expect that the FO rankings will turn out to not be predictive in this case. It will be your job to figure out why, and improve the system to resolve the problem.
#124 by Karl Cuba // Oct 30, 2013 - 12:52pm
Smaller margins of victory mean that a team is more vulnerable to 'fluke' events ie. fumble returns, blocked kicks. A team that is regularly ahead by two scores will be more likely to survive such an incident.
3rd down conversions are by their nature a smaller sample size metric and so less reliable as a long term indicator.
#154 by Nathan of Brai… (not verified) // Oct 30, 2013 - 9:24pm
Interestingly, Chiefs haven't won on any fluke plays, like the Schottenheimer teams of 1995 and 1997 did.
In fact, the only reason some of the games have been as close as they were is the other team scored on some extremely-low-percentage plays:
Houston: Long bomb (admittedly from a QB who was extremely prolific in college), but he hit on the first try
Tennessee: .02 seconds from being sacked, Fitz flings the ball to C. Johnson, who was not expecting it and just kind of standing there. He then uses his speed and a good block to score a 40+ yd TD.
I've thought about it for several minutes, and I can't remember any corresponding fluke play for the Chiefs. No long low-percentage plays, and ST/Def scores all came when Chiefs were already ahead, or tied 0-0 at the very start of the game.
#157 by Karl Cuba // Oct 30, 2013 - 10:43pm
You've missed the point, I wasn't saying the Chiefs have scored oh any random plays.
The point is that there is variance in the result of every NFL game that has nothing to do with the strength of the teams: bad/missed calls, fumble luck, defensive TDs, weather, inopportune injury or even a player slipping at the wrong time. Every team is subject to these things but the larger a team's average margin of victory the more likely it is that the team will not lose as a result of one of them.
So a high octane offense that terms to create a high point differential will, over a down of time, be more likely to win more games than a defensive team that tends to win by only a few points.
For an example look at the Rams victory over the 49ers last year, the Rams couldn't move the ball all day but because the niners were a team that plays close defensive games they lost because of a dropped TD catch, a fumble near their own end zone and a missed field goal in overtime.
#165 by David // Oct 31, 2013 - 5:09am
Really? Really? Teams (sorry, *smart* teams) have worked out exactly what games are the "right" ones to win? Seriously? You're an idiot. Actually, to be fair, you may be very smart, but your proposition is idiotic, so going on the evidence I have available, you're an idiot.
Teams are absolutely trying to win as many games as possible. If that wasn't the case, then some "smart" teams would, by now, have not turned out for a game. I guess they didn't get your memo about only needing to win the "right" games.
If you, even for a second, believe that the goal of all teams is not to win as many games as possible, I don't know why you bother to visit this website. It's also not possible to have any form of meaningful communication with you about strategic or tactical thinking if our baselines for what would be 'successful' are so far apart.
By which I mean your baseline is stupid.
#169 by Buck B (not verified) // Oct 31, 2013 - 11:27am
I think "successful" is winning the SB, however you get there. Winning games and winning the SB are related, but different, and sometimes inconsistent. A GM can make decisions that decrease the chance of winning a game while increasing the chance of winning a SB. A coach can make decisions that would increase the chance of winning a game while decreasing the chance of winning the SB. A player can make plays that increase the chance of winning a game while decreasing the chance of winning the SB. How? Lots of ways. I will leave that as an exercise for you to try to stretch your mind a bit.
#170 by dmstorm22 // Oct 31, 2013 - 11:37am
Could you please give an example of a decision a coach can make that would 'increase the chance of winning a game while decreasing the chance of winning the SB'?
Honestly, this sounds awful lot like trolling. That makes no sense, even for people that believe in TEH RINGZZZZZ.
#175 by dmstorm22 // Oct 31, 2013 - 12:32pm
I guess, but is that really a concern in Weeks 1-8, which this DVOA article is based off of.
Maybe resting guys late in blowouts, but I don't believe the Packers (or most teams this year) have done that.
#176 by BaronFoobarstein // Oct 31, 2013 - 12:42pm
Player health is the most obvious, and the one that most applies to a contending team. But a team that's already eliminated or nearly eliminated if optimizing for super bowl wins across seasons would likely prefer to play backups in an effort to improve the future play of young players through experience, get better data for making decisions on who to keep in the off-season, and improve draft position.
#179 by dmstorm22 // Oct 31, 2013 - 1:13pm
Yeah, that's another idea, but I don't see how these personnel decisions effect how FO is wrongly judging these teams by DVOA. The only potential issue, which they've covered and given their reasons for, is late in teh season rest-o-ramas. Those haven't happened yet. If we're looking at the Chiefs, or even Packers, rarely have either team rested their top guys (I think the Packers may have late in the Redskins win). I agree that late season FO may not be accurately judging team performance due to resting of players, but they've shown that the model works better to count those the same.
#188 by Buck B (not verified) // Nov 01, 2013 - 10:34am
KC and DEN play each other at least two times, possibly three. Assume they play three times. The third game is the only game that matters. Is that going to affect how you play the first two games? Yes. Player health and intelligence management are the two obvious categories of factors that may at times be inconsistent with the goal to win each game, as if all games have the same meaning. They don't.
#184 by coremill // Oct 31, 2013 - 6:32pm
I can give an example of the opposite. The Niners pretty clearly held back a lot of their read-option packages with Kaepernick for the playoffs last year so that playoff opponents wouldn't be prepared for it (I think Greg Roman admitted as much later). And it probably cost them at least one game (at STL, where they lost in overtime 16-13).
#171 by Behemoth (not verified) // Oct 31, 2013 - 11:46am
Can we dial back the rhetoric here? Of course teams make strategic decisions all the time that could limit their chances of winning one particular game. In football, we've all known examples where certain teams rest players when a win would be relatively meaningless, often in weeks 16 and 17; whether or not it's a good practice or a poor one is open to debate - but it certainly has happened with very good organizations.
To borrow an example from baseball, didn't the Senators (sorry, the Nationals - change comes hard) sit down one of their uninjured stars a couple of years ago in an effort to try to preserve his health for many years, sacrificing short term wins - wins in the playoffs, no less) in a hope to achieve more in the long run? Again, the strategy may or may not have been wise, but it occurred.
Do you not think that, in retrospect, the Patriots and the Indigenous Persons regret their choices to put unhealthy Gronkowskis and Griffins back on the field seeking short term success?
In short, your premise is certainly open to debate. In light of that self-evident fact, wouldn't a less openly derogatory approach be more wise? Not to mention more civil ....
#126 by Perfundle // Oct 30, 2013 - 1:07pm
This was a decent if flawed post up until the last paragraph. It's their job to figure out why they couldn't predict the result of one or two games?
"Strong offensive teams w/ weak defenses tend to lose in the playoffs more often than strong defensive teams w/ weak offenses."
I would say that most strong defensive teams w/ weak offenses don't even make it to the playoffs, so there is survivor bias going on here. The strong defensive teams that make the playoffs typically have better offenses than the strong offensive teams have defenses.
#132 by theslothook // Oct 30, 2013 - 3:15pm
A few things:
You've stumbled on the tenuous nature of trying to link offense and defense together. Let me give you an example. Do the defenses of great offenses really suck as much as the numbers say? DO you remember when pittsburgh played the first 4 games without ben, their defense got better. Then suddenly when he came back in the same year, their defense declined somewhat. Coincidence? I think if your offense is great, opponents augment their playcalling - go for higher risk decisions, more throws, more 4th down gambles(which they should be doing anyways). These make the defense look worse, even by dvoa, than they probably are. On the other side, the reverse is also true of great defensive teams with weak offenses. The playcalling becomes less aggressive, more field position based, more runs, etc. Thus, its likely, at the extremes, the other side is being inflated/deflated.
#135 by Rick S (not verified) // Oct 30, 2013 - 4:43pm
Essentially, you are saying that the Chiefs are under-rated and should be getting more respect... I believe there is a format to submit those type of comments.
One could make the argument that teams that thrive/survive by having a positive turnover ratio and relying on a conservative offense (like KC in 2013, 1997 & Chi 2001, 05,06) fail more than many of the offensive minded teams, especially in today's NFL. KC's coach may be Andy Reid, but they are essentially playing "Marty-ball" as in Schottenheimer, which emphasizes a conservative offense that plays not to turn the ball over and relies on their defense to create turnovers. When the ball doesn't bounce their way - LOSE. When they need to score a lot to win a track meet - LOSE.
We have seen the Marty-ball script before and understand how it plays out in January.... "Men, I see a gleam..."
Perhaps Denver's offense is unsustainable or their defense brings them down like the Chargers of the 80s or Oilers of the 90s, but I feel much more confident in a team that can score and play situational defense than one that relies only on defense and turnover ratio. In regard to Denver's defense, they have had a ton of injuries and lost Von Miller for being a knucklehead. Getting Miller, who is one of the most impactful players in the league is a huge boost, and with him Denver's defense does compliment their offense in that they are good against the run which keeps teams from playing keep-away and they can rush the passer when ahead.
What makes Football Outsiders a great resource is that it is as objective as possible. Yes I have seen outliers, like the NYG winning a SB against the Patriots in 07 (with a miracle catch), or the the Eagles being rated on FO much better than their record indicates, but after mid-season, FO is usually spot on.
#151 by Cythammer (not verified) // Oct 30, 2013 - 6:44pm
Not saying you're wrong, but if you want to base your argument mainly on a bunch of anecdotal examples of teams with strong defenses and a lot of takeaways cratering in the playoffs, then you should also note how many teams with high-flying offenses have tanked in the playoffs.
#155 by Nathan of Brai… (not verified) // Oct 30, 2013 - 9:27pm
No, essentially I'm saying that perhaps offense and defense shouldn't be weighted equally, but it took a rather extreme set of circumstances to bring that imbalance to light.
The quest is to be as successfully predictive as possible. They have made other adjustments based on lessons learned.
Maybe even if it plays out as I stipulated, maybe it doesn't require an adjustment. But that's a hypothetical worth considering, isn't it?
#164 by David // Oct 31, 2013 - 5:05am
"The quest is to be as successfully predictive as possible"
That might be your quest, but FO (Aaron) have previously stated that they want DVOA to be both predictive and retrodictive, and that the current approach is a balance between the two.
Personally, my quest is to complete Jet Set Willy without POKEs, though I realise that this may be tricky...
#119 by Erock (not verified) // Oct 30, 2013 - 11:56am
I continue to be surprised that Green Bay's defense ranks so low. Other than some hiccups in the pass defense the first couple of weeks, they've looked good to me. They have good gap control, get good sack production, and are getting solid CB play from guys like Sam Shields and Davon House. How much of an impact do garbage-time touchdowns have on DVOA? The Packers shut down the Vikings until the game was out of reach. This is just one of the few instances where DVOA doesn't match what my eyes see.
#133 by Perfundle // Oct 30, 2013 - 3:40pm
One of GB's biggest issues is that, unlike past years, they are getting almost no turnovers at all. Also, they are the absolute worst red-zone defense without adjusting for opponent strength, giving up the most points and the highest TD percentage per opponent red-zone trip. They were very bad at this last year and it's only gotten worse.
#120 by hrudey (not verified) // Oct 30, 2013 - 12:24pm
Jacksonville is clearly ranked too high because the site fears losing exclusive access to Shad Khan's moustache. Mike Martz' unfathomable analysisis way better than this. lol OPEN DATE has MOAR PRO BOWLERS, wait, no, that was ur sister and them PBA bowlers, mybad brah.
#174 by QCIC (not verified) // Oct 31, 2013 - 12:30pm
"(Yes, one team that ranks below them ended up winning the Super Bowl, but remember that team got its best defensive player back from injury right before the playoffs.)"
You don;t need this caveat. You should entirely expect teams at 15.6% DVOA to occasionally win the Super Bowl. Have the courage of your convictions man and don't apologize for counterexamples that are not actually counterexamples, just because you are trying to head off criticism. Such over concern with criticism is how we ended up in the mess this website is partially helping to lead football out of.
#190 by Cythammer (not verified) // Nov 02, 2013 - 7:06pm
This is late, but if you think that DVOA is being harsh on the Chiefs, check out Massey-Peabody Analytics' rating: http://massey-peabody.com/nfl-2013-current-power-rankings/
They have the Chiefs at 20th!… and the Panthers all the way up at third. I don't know anything about their methodology, but I find it fascinating that any statistical ranking can find an 8-0 team to be below average.
#191 by Perfundle // Nov 02, 2013 - 10:26pm
Well, they provide a rough sketch of their methodology, but it's fairly similar to FO's. A think the biggest culprit is the influence their preseason rankings are providing. It seems like an average performance doesn't move the rankings much from the previous week.
#192 by DisplacedPackerFan // Nov 03, 2013 - 6:47am
They are not trying to be descriptive at all, they are simply trying to be predictive while weighting things in such a way as to avoid outcome bias, chance, and over or under confidence. While they use play by play they cook a few things into the numbers that DVOA doesn't, like home field, and recent performance carries more weight than past performance (DVOA doesn't start doing that until weight DVOA shows up). They also do opponent and situation adjustments like DVOA.
KC has been playing bad teams and been playing worse each week. So those ranking are saying that they are playing like the 20th ranked team in the league right now. While they are trying to be purely predictive, they do still use preseason and past performance data, so yes the preseason stuff might be pulling KC down a bit too, but really I think the way they use opponent and situation adjustments is simply a bit stronger than what DVOA does and really hurts KC. You'll note that they also have NYG much higher. My guess is because the play by play, even against crappy teams, has looked better recently (and looking at the weekly ratings the improvements have come recently). Their predictive indicators have gotten a lot better, even in losses.
I do wonder what DVOA would like like if they had separate descriptive and predictive numbers. DVOA tries to be both, some of that is to help avoid the same things that Massey and Peabody talk about avoiding, but it's been said that they know certain things are more predictive. M-P gives those much more weight than I think DVOA does.
#193 by Perfundle // Nov 03, 2013 - 2:38pm
Well, their predictions on Kansas City are awful, then. Even though they're ranked 10th here, it's a very strong 10th and far ahead of 11th place. How did they drop significantly after beating Dallas? Then, they barely moved up after dominating the Giants, and New York barely moved down. I'd be interested in seeing how their predictions compare to FO's.
#194 by Ibis fan (not verified) // Nov 04, 2013 - 10:38pm
How would you decipher a player like ed reed who doesn't get the stats of apolomalu because of nobody throws near him c,mon guys I nd an updated ed reed for my all miami hurricane team big fan of your guys work thank you