Week 15 DVOA Ratings
by Aaron Schatz
After a dominating victory over their division rivals in Seattle, the Los Angeles Rams have now opened up a wide lead in the Football Outsiders DVOA ratings. The Rams are up to 38.0% DVOA, which ranks them among the top dozen teams we've ever tracked through Week 15. (You'll find that table in this week's Any Given Sunday column.) The Rams actually dropped one spot in offensive DVOA, so they no longer qualify as ranking in the top five for all three phases of the game, but their defense and special teams ratings both improved and more than make up for that small drop.
The Rams rate even better in weighted DVOA than they do in full-season DVOA, since seven of their last nine games (going back to Week 6) had a single-game rating over 50%. Since Week 6, there have only been 24 other games where a team recorded a single-game DVOA of 50% or more, or less than one per team. The Rams have seven. The only other teams with more than two of these games during the same period? The Philadelphia Eagles have three, and the Baltimore Ravens -- who narrowly climbed ahead of Pittsburgh this week to rank as the top AFC team in DVOA -- have a surprising five similarly dominant games since Week 6.
As has been covered in numerous other places, the Rams are currently undergoing one of the greatest turnaround seasons in NFL history. Barring some sort of unexpected collapse in the final two weeks, the Rams will set a new record for the biggest year-to-year improvement in total DVOA. That record is currently held by the 2013 Kansas City Chiefs, who added Andy Reid and Alex Smith and went from -40.1% DVOA (dead last) and a 2-14 record to 17.5% DVOA (sixth) and a playoff spot. Not bad for a team that didn't even get much help from their No. 1 pick, since they got to lead off one of the worst first rounds in NFL draft history.
|Biggest Year-to-Year Improvement in Total DVOA, 1986-2017|
The good news for the Rams is that a lot of the teams on this list went on to have several strong years of playoff contention: the recent Seahawks and Chiefs, the Peyton Manning Broncos, the Jon Gruden Raiders. The 1999 Rams won the Super Bowl, and the 2004 Steelers won the Super Bowl a year later. That Gruden team needs a little extra explanation, by the way. The 1999 Raiders went from 27th in DVOA to third in DVOA in Gruden's second year as head coach and managed to improve their record from 8-8 all the way to... 8-8. The 1998 Raiders were 5-1 in games decided by a field goal or less. The 1999 Raiders were 1-4 in games decided by a field goal or less, and didn't lose a game by more than a touchdown. The Raiders finally put luck and strong play together in 2000 and improved to 12-4.
The Rams are one of three teams this year that look ready to improve DVOA by over 30 percentage points from last year. The Saints were 21st at -1.9% last season, and are now over 30%, second behind the Rams. The Jaguars were 26th at -10.4% last season, and move up to seventh this week at 20.8%. In fact, the Jaguars move past the Patriots despite New England's victory over Pittsburgh, although the Patriots are still ahead in the weighted DVOA ratings. This is the first week where an early game completely falls out of the weighted DVOA formula, eliminating that big Week 1 Chiefs win over the Patriots.
On average, only 1.6 teams will improve by 30 percentage points in DVOA, so this has been a good year for turnarounds. But so was last year, which also had three teams that improved by 30 percentage points: Dallas, Atlanta, and at exactly 30.0% improvement, Tennessee.
Things are less extreme down at the bottom of the DVOA ratings. The Cleveland Browns now finish 0-16 in 51.4 percent of our simulations, but they are far from the worst team in DVOA history. They are essentially tied with the 3-11 Indianapolis Colts at the bottom of the ratings, and only two teams have been in last place after Week 15 with higher DVOA ratings than the 2017 Browns: Arizona in 1995 (-27.6%) and Minnesota in 2001 (-26.8%). This lack of an extremely bad team is also apparent in the offensive ratings. Despite all the gnashing of teeth about the bad quarterbacks around the NFL this season, only two teams have ranked in last place in offense after Week 15 with higher offensive DVOA than the 2017 Browns: San Francisco in 2015 (-18.0%) and Dallas in 1989 (-21.6%).
Defense shows the same trend, where no team is terrible. In fact, this is even stronger on defense than offense or overall DVOA. The Oakland Raiders are in last place in defensive DVOA, but they are the first last-place team ever after Week 15 to have a defensive DVOA better than 13.0%.
* * * * *
Once again this season, we have teamed up with EA Sports to bring Football Outsiders-branded player content to Madden 18. This year, our content for Madden Ultimate Team on consoles comes monthly, while our content for Madden Mobile comes weekly. Come back to each Tuesday's DVOA commentary article for a list of players who stood out during the previous weekend's games. Those players will get special Madden Mobile items branded as "Powerline, powered by Football Outsiders," beginning at 11am Eastern on Friday. Our stars for Week 15 are:
- DE Cameron Jordan, NO (HERO): 4 passes batted at the line plus 4 hurries.
- DT DeForest Buckner, SF: Led 49ers with 9 combined tackles, plus PD and 0.5 sack.
- OT Rob Haverstein, LARM: Helped lead Rams RB to 31 carries for 197 yards, with no sacks allowed.
- K Josh Lambo, JAC: 6-for-8 touchbacks. Other kickoffs forced returns that ended at the 18 and 20 yard lines for Houston. Also hit 38-yard FG, 6-for-6 XP.
- P Andy Lee, ARI: 50.2 gross yards per punt, with punts downed at the 6 and 12 yard lines plus a 61-yard punt returned for only 9 yards.
* * * * *
All stats pages should now be updated through Week 15, including snap counts, playoff odds, and the FO Premium DVOA database.
An important note: For some reason, our data parser skipped over the fourth quarter of the Tennessee-San Francisco game, and I didn't realize it until this afternoon. Therefore, the DVOA numbers on the website (both team and individual) are not complete for those two teams. We'll get that all fixed over the course of the next few days, as we're also assimilating a series of league stat changes.
* * * * *
These are the Football Outsiders team efficiency ratings through 15 weeks of 2017, measured by our proprietary Defense-adjusted Value Over Average (DVOA) system that breaks down every single play and compares a team's performance to the league average based on situation in order to determine value over average. (Explained further here.)
OFFENSE and DEFENSE DVOA are adjusted to consider all fumbles, kept or lost, as equal value. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season.
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.
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).
87 comments, Last at 26 Dec 2017, 3:54pm
#1 by DezBailey // Dec 19, 2017 - 8:23pm
Aaron, I'm intrigued by weighted DVOA for the Chargers and Chiefs, mainly because I have them ranked almost the same in my latest report: http://besreport.com/week-15-bes-rankings-2017/
It's seemed like an odd thing for the Chargers to be among the top-10 despite a record that hovered at or below .500. Seems like Weighted DVOA and the BES believe the Chargers are the hotter team despite losing to the Chiefs 30-13 and being swept by the Chiefs on the season.
As for the Chiefs, under "new" offensive coordinator Matt Nagy, they've been on fire!...almost having to lean on Kareem Hunt more just to keep from scoring too fast and gassing the defense. Im not surprised to see them ranked 5th by DVOA in Offense. The BES arguably has them too low at 14th.
However, their defense has been phenomenal in the last two weeks. DVOA seems reluctant to give them props with a 22nd ranking while the BES has moved them up to 13th from 21st a week ago. But that's probably due to their improvement in pass rush and takeaways, two areas the BES stresses.
Equally interesting are the bottom-three in both systems. Like DVOA, the BES also has the Colts and Browns essentially tied. I agree that the Browns, though likely to go winless based on your prediction data, aren't THAT bad lol. they've been more competitive than their record would indicate, especially on defense. I think new GM John Dorsey could have them in playoff contention within the next three or four years if not sooner. He's that good.
#2 by Will Allen // Dec 20, 2017 - 12:31am
Zimmer has won two division titles in 4 years, while usually having his 1st choice for starting qb either injured or a rookie, and 3 of 4 years with awful to below average injury/suspension luck. Placekicker chokes likely cost the team a playoff spot in one of the two years they didn't win the division, and definitely cost the team a playoff win. He coached one year with a detached retina and multiple eye surgeries. The competition in the division includes a team generally considered to have the best or 2nd best qb in the league, and another one has a qb who usually finishes in the top 10 by DYAR.
It is nuts that he didn't get a head coaching job until he was 58 years old.
#5 by Aaron Brooks G… // Dec 20, 2017 - 7:55am
Caldwell is 8 games over .500 and without controversy since coming to Detroit.
That doesn't sound like much, but he's the first coach since Joe Schmidt to do that in the Ford era. His 11 wins in 2014 were the second most in franchise history*.
* - it's a sad history. The 1991 team which went 13-5 had the most wins in team history. Damn Redskins.
#8 by Will Allen // Dec 20, 2017 - 9:27am
I admit I was pretty hard on him when Rodgers Hail Maryed the Lions, but Caldwell is better than what he is given credit for, and some of it is for the same reason that Dungy is said by some to not be worthy of the HOF. If you aren't demonstrative on the sidelines or press conferences, and don't win 3 Super Bowls, then meathead fans say your players would be better if you screamed more. It's pretty dumb.
Oh, and if your team with the most wins ever is gonna get crushed in the playoffs, it may as well be by one of the best teams ever. Outside of DVOA land, that '91 D.C. team gets overlooked when the best champion ever debates are engaged in.
#33 by JoeyHarringtonsPiano // Dec 20, 2017 - 3:02pm
I’ve been a Caldwell defender (not that I think he’s great....just better than his reputation), but I kind of soured on him when the team pulled a no-show in the biggest game of the year (Thanksgiving), where they didn’t start playing until already down three scores. Then they did the same thing following week against Baltimore. Those two games will probably result in them missing the playoffs. Thing is, I don’t know how the GM can justify canning him if they win the next two games (likely) and finish 10-6. Also, I’m not 100% sure a new coach would do better than he has.
#34 by JoeyHarringtonsPiano // Dec 20, 2017 - 3:05pm
Any coach that loses his top two QBs to injury, then his top running back, and other assorted starters...yet still has his team in contention for the #1 seed late in the year, deserves coach of the year consideration for sure.
#6 by RBroPF // Dec 20, 2017 - 9:23am
The Pats defense has real and glaring flaws, no doubt. But I think it's well understood that DVOA hates the bend-don't-break approach that is Belichick's central defensive philosophy.
Imagine a game where the opposing offense has 10 drives that all start at the 15 yard line. Each drive lasts 10 plays and gains 40 to 60 yards, and 4 of them result in field goals. Belichick thinks that's a nearly perfect game, while DVOA sees a 90% success rate for the opposing offense and 500 yards surrendered and thinks that's an historically bad defense.
#15 by Eddo // Dec 20, 2017 - 10:30am
This is a myth you keep peddling, and it's not true. FO's numbers consistently have Sanders as a top RB.
Looking at rushing only, and limiting to Sanders's career (I don't want to add eight more years of stats for Bettis).
1989: 239 DYAR (3rd), 12.1% DVOA (5th)
1990: 330 DYAR (1st), 21.7% DVOA (3rd)
1991: 254 DYAR (5th), 8.7% DVOA (15th)
1992: 93 DYAR (19th), 0.3% DVOA (23rd)
1993: 70 DYAR (22nd), 0.7% DVOA (20th)
1994: 348 DYAR (2nd), 18.5% DVOA (2nd)
1995: 200 DYAR (7th), 6.8% DVOA (12th)
1996: 380 DYAR (2nd), 22.7% DVOA (2nd)
1997: 447 DYAR (2nd), 25.3% DVOA (2nd)
1998: 15 DYAR (29th), -6.4% DVOA (28th)
1989: -22 DYAR, -19.7% DVOA (not enough attempts to qualify)
1990: 202 DYAR (3rd), 25.7% DVOA (1st)
1991: 51 DYAR, 5.3% DVOA (not enough attempts to qualify)
1992: 45 DYAR (27th), 0.0% DVOA (24th)
1993: 25 DYAR, -3.8% DVOA (not enough attempts to qualify)
1994: 211 DYAR (5th), 17.1% DVOA (3rd)
1995: 211 DYAR (5th), 12.7% DVOA (6th)
1996: 86 DYAR, 20.1% DVOA (not enough attempts to qualify)
1997: -42 DYAR, -32.8% DVOA (not enough attempts to qualify)
1993: 371 DYAR (2nd), 21.2% DVOA (3rd)
1994: -101 DYAR (38th), -16.3% DVOA (34th)
1995: -53 DYAR (38th), -15.5% DVOA (39th)
1996: 384 DYAR (1st), 20.3% DVOA (4th)
1997: 280 DYAR (4th), 8.9% DVOA (9th)
1998: 46 DYAR (26th), -6.1% DVOA (27th)
Heyward's 1990 has a higher DVOA, and you know what? He averaged 4.6 yards per carry as a power back and a 60% success rate. That's pretty damn good! But DYAR is still saying that, because Heyward was situational, he didn't put up as much value as Sanders did over the course of the year.
Sanders's FO metrics dropped in 1992, but he had only a 43% success rate that year, and by yards per carry (4.3), that was his worst year until his very last (also 4.3). In 1993, he missed five games, which suppresses his DYAR a bit, but it also was a down year for Sanders (4.6 yards per carry, only 3 TD, 44% success rate).
Sanders's final season is interesting. I remember at the time thinking that he retired on top of his game, and some raw stats (1491 yards) do suggest he was still, at least, a very good runner; others are less impressive (4.3 yards per carry, 4 TD). But FO would have likely flagged his season as the beginning of the end, as he had the lowest DYAR and only negative DVOA of his career, on a meager 39% success rate.
#32 by JoeyHarringtonsPiano // Dec 20, 2017 - 2:55pm
Keep in mind that his ‘98 team’s offensive line had one of the worst adjusted line yards in DVOA history. It wouldn’t surprise me if him having to constantly juke guys in the backfield played a role in him deciding to hang up his cleats.
#36 by Eddo // Dec 20, 2017 - 3:13pm
Yeah, that's a really good point, too. I was just riffing off the stats and not really going in depth, but Sanders was pretty clearly fed up with a lack of supporting cast and figured continuing to get pounded wasn't worth it.
#69 by Aaron Brooks G… // Dec 21, 2017 - 8:55am
What myth? The numbers you cite put Sanders career DYAR at roughly 3x Heyward's, on 3x the touches. So back of the envelope, DVOA says they are interchangeable. I doubt even Ironhead thinks that's true. He himself told the joke that if they needed 2 yards, he could get 3; and if they needed 4 yards, he could get 3.
It's a little less keen on Bettis on a rate-basis (his usage is about 50% higher than Heyward's, and his DYAR about 20% higher), and both of them are boom-bust in the rankings, but when DVOA likes them, it really likes them. Heyward finished 3rd in DYAR in 1990 on 147 touches and 4 TDs.
You're arguing DVOA doesn't reward high floor-low ceiling guys like Bettis and Heyward (who are running jokes around here), and yet their DYAR/touch is on par with a 1st-ballot RB in the GOAT debate who has the 2nd most yards/rush and 3rd most rushing yards in NFL history -- and who was a boom-bust guy.
Bettis and Heyward are more useful comps than Smith, who spent most of his career behind a legendary offensive line. DYAR is going to love him, because it almost has to. It was nearly impossible to have a failed rush behind a line that moves the defense back three yards. The other guy it probably loves is Brown, but we're decades away from him.
#76 by Eddo // Dec 21, 2017 - 2:14pm
I'm arguing your point that "DYAR was hard on Barry Sanders". He was top ten in DYAR seven of ten years.
Do DYAR and DVOA rate Sanders as highly as the eye test? No. But you put forth the disingenuous argument that FO metrics "don't like Sanders" when they like him plenty, but can't account for the fact he had shitty teammates.
Also, you're confusing Heyward with Leroy Hoard on the "get you 3 yards" quote. And since when are Bettis and Heyward "running jokes"? Bettis is a Hall of Famer, and while I'm one of many FO readers who thinks he probably doesn't deserve that, I don't think very many will tell you he didn't at least deserve to be considered. And Heyward was a good back, in limited use, for a decent amount of time - hardly a joke.
EDIT: Also, putting up the DYAR/carry on three times as many carries is a point in Sanders's favor.
#9 by Will Allen // Dec 20, 2017 - 9:34am
Look, Belichik is great, but no, he does not think bend but don't break is the ideal defensive approach. He'd be the first to tell you that, don't bend, don't break, crush, stomp, and force turnovers, like his '86 Giants team, is best. But you have to have the roster for it. He gets the most of out of what he has now, and it works because he has a great, highly efficient offense, with a HOF qb, and excellent special teams.
#16 by PatsFan // Dec 20, 2017 - 10:50am
As an example of how the NE offense can help mask the terrible defense, until the Miami debacle broke it, NE had a streak where opponents started something like 105 consecutive drives from the opponent's side of the 50.
#19 by Hoodie_Sleeves // Dec 20, 2017 - 11:13am
" He'd be the first to tell you that, don't bend, don't break, crush, stomp, and force turnovers, like his '86 Giants team, is best."
I think that BB would tell you that assembling a defense that can do that, and do it consistently year-to-year, isn't a wise use of salary cap space in the modern NFL, so that it isn't 'best'.
I think he'd tell you what is best is an offense that can exploit the post-2004 and 2010 passing changes to the maximum extent, special teams that can take advantage of the fact that special teams are an afterthought for a lot of teams, and a defense that can keep teams from easy scores is best.
The reason the Patriots defenses have vacillated between below average and above average, while maintaining that bend-but-don't-break philosophy since 2005 isn't because that's what BB has had - its because (aside from the OL/DL), the offense defines what matchups actually matter in a game, and good offenses can make a game about your worst defender. Defenses can't do the same thing without generational talents like Revis. Having a bad 2nd CB hurts you a lot more having an elite one helps you, and hurts more than having a bad 2nd WR - and conversely, having a good 2nd (and 3rd and 4th) WR helps you a lot more than having a good 2nd CB.
The last Patriots drive of this weeks game is a perfect example of this - Gronk and Brady are so frigging good that it really didn't matter what else was going on. Generally, if an offense executes well, plays succeed - which means that above a certain defensive threshold, improving your offense is almost always a better investment capwise.
(The Patriots defense was below that threshold the first 4 games this year, I think)
(this mostly applies to 'skill positions' - pass rushers can absolutely dictate games if they're good enough - and offenses don't adjust)
#23 by Will Allen // Dec 20, 2017 - 12:16pm
Well, yes, what is best is to draft one of the best qbs ever in the fifth round, and then have him have a nearly two decade career in which he never really pushes to be the highest paid player in the game.
If you consider that a viable repeatable strategy, you may wish to think again.
#24 by Hoodie_Sleeves // Dec 20, 2017 - 1:19pm
I don't think my point was clear - as your response doesn't seem to reference anything I actually said.
My point is that offense, defense, and special teams are not independent things in a salary cap constrained environment. Right now, the rules favor spending on offense, and the total range of offense is larger, which means that over a certain threshold, improving your defense is hurting overall team performance more than helping.
If you go back and look at the outlier defenses of the last decade - you'll find very few that have above average offenses (2013 SEA sticks out with the -25.9% defense, +10% offense) - most are like 2012 Chicago (-26.7 def, -10.9 off) or 2015 Denver (-25.8 def, -8.7 off). If you go looking for elite offenses, you won't find the opposite effect - there are plenty of teams with +25% offenses and above average defenses.
In addition - it seems to be significantly easier to maintain an elite offense than it does an elite defense - defenses seem to peak and fall apart faster - look at DEN over the last couple years:
Now, that 2015 defense was fantastic - and they won a superbowl on it (and getting lucky and acquiring Manning) and the '14,'16 defenses were great, but the cost of that defense has been $90M a year, which left them less than $30M to spend on their offense once Peyton Manning got paid.
If you go through and plot out defensive spending vs offensive spending in the NFL - you'll find that the correlation for wins to offensive spending is about 3 times the size as that of defensive (.35 to .13 - they're both positive because there are teams like the 49ers that are 60M under the cap)
Simply put - defense is a poor investment.
#26 by Pat // Dec 20, 2017 - 1:55pm
Your comment would make a lot more sense if that's the strategy that the Patriots actually *used*, which is partly what Will was saying. The Patriots haven't actively skimped on defense to prefer offense. They've just gotten ridiculously lucky in acquiring cheap offensive talent, either through a late-round QB (Brady), a disgruntled Hall of Fame receiver (Moss), a misused elite slot receiver (Welker), or a second-round TE (Gronkowski).
Why would you think they don't invest in defense? They signed Adalius Thomas for a deal worth $20 million guaranteed in 2007. They signed Rosevelt Colvin to a huge deal in 2003. More recently, they signed Revis for $12M for a year. From 2007-2013, they drafted a grand total of 5 offensive players in the first 3 rounds, as opposed to *16* defensive players.
"In addition - it seems to be significantly easier to maintain an elite offense than it does an elite defense"
Yes, exactly, which is why the Patriots, after having gotten an elite offense basically for free, have spent draft capital primarily on defense.
It's not that they don't *want* a defense. It's that it's not easy to *get* one.
#30 by Hoodie_Sleeves // Dec 20, 2017 - 2:41pm
". The Patriots haven't actively skimped on defense to prefer offense"
You should take a look at their salary cap distribution over the last decade. Even with Brady being a discount, they spend significantly more on Offense than Defense, EVERY YEAR.
Offensive Spending Minus Defensive.
They spend roughly 55-60% of the cap every year on offense.
"It's not that they don't *want* a defense. "
Nice strawman bro.
"It's that it's not easy to *get* one."
Yeah, no shit. That's what I said. Spending money on defense is less productive than spending it on offense - that's literally the whole point.
#37 by Eddo // Dec 20, 2017 - 3:19pm
Those differences don't really show anything without knowing what the league breakdown is, too. It could very well be that the rest of the league skews even further towards offensive spending, which means that the Patriots would be doing the opposite of what you suggest). I really don't know.
#41 by nat // Dec 20, 2017 - 3:29pm
Looking at just this year on overthecap.com, it looks like teams split about 50-50 for being offense or defense heavy on cap usage. The Patriots don't look like extreme outliers. But if he's right that they have spent the last decade on the offense heavy side every year, then he's probably onto something here.
It may not be as big a deal as he thinks. But it warrants a closer look, and deserves to be taken seriously.
#43 by aces4me // Dec 20, 2017 - 3:40pm
Especially as Will was quick to point out NE gets more value for their money on the offensive side of the ball than most teams. If in fact they wanted to put equal emphasis on both sides of the ball they would have to spend more on D than they do on O.
#38 by Eddo // Dec 20, 2017 - 3:25pm
Adding onto my last comment:
According to http://www.spotrac.com/nfl/positional/, the average salary of an offensive player in 2017 was $2.51M, while the average salary of a defensive player was $2.19M.
Some really simple math would be to assume teams have equal number of offensive and defensive players (not necessarily true, but this is a back-of-the-envelope thing). That comes out to 26.5 offensive, 26.5 defensive, which comes out to $66.6M spent on offense, $57.9M on defense. That comes out to a difference of a little less than $9M.
So if those assumptions are true, the Patriots have spent more on offense relative to the league since 2015. But not in 2013 or 2014, really.
I'm going to keep digging to see if I can find any more data on this.
#39 by Eddo // Dec 20, 2017 - 3:28pm
OK, I overlooked some interesting info on that same page (http://www.spotrac.com/nfl/positional/). The graph at the top shows total amount spent on offense and total amount on defense, along with the number of players.
In 2017, 928 offensive players made a combined $2.3B, while 994 defensive players made a combined $2.2B. If you average that out, it is definitely closer to 50% than the Patriots are, so your point does seem to be true.
#42 by Eddo // Dec 20, 2017 - 3:37pm
Now I'm going to look at some specific teams, for 2017 at least. All numbers are offensive spending minus defensive spending:
Spotrac shows the Patriots at +$9M for 2017; where are you getting your +$15M figure?
Looking at those numbers, I think you're attributing a rule to the Patriots that probably isn't true. What's more likely - as with most things that people attribute to the Patriots' "system" - is that the Patriots are really good at evaluating how much to pay individual players, and as such, have a strong roster year in and year out (and that having possibly the best QB ever is very advantageous).
#45 by Eddo // Dec 20, 2017 - 3:45pm
Oh, man, I could have done this at https://overthecap.com/positional-spending" so much more easily! (I used Spotrac's individual team pages.)
#78 by Pat // Dec 21, 2017 - 3:29pm
"You should take a look at their salary cap distribution over the last decade. Even with Brady being a discount, they spend significantly more on Offense than Defense, EVERY YEAR."
Ha! I love the fact that you use "the last decade" - as in, the decade when the Patriots defense has been significantly worse than it was the *first* decade of the Belichick/Brady Patriots. Nice cherrypicking.
In the earlier decade, the Patriots spent significantly more on defense than offense. And that's even including when Brady *wasn't* cheap. Brady was the highest cap-hit player on the roster in 2005, and the starting defense was still 10% more expensive than the starting offense.
And I think you might be stretching on "last decade," since that includes 2007, where they definitely spent more money on defense than offense. Note that you do have to be careful those years, since those were the years of "fake bonus that makes Kyle Eckel look like the 5th highest paid player on the team."
"Spending money on defense is less productive than spending it on offense - that's literally the whole point."
I think you're missing the distinction. It's not that spending money on defense is less productive. It's that there are *fewer defensive players worth spending it on*. So if you don't have one of those guys, you don't just throw equal amounts of money to someone who isn't worth it.
You're assuming the reason for the decreased spending on defense is demand. I'm saying it's supply. In the first decade of the Belichick/Brady Patriots, they had top-level defensive talent they needed to keep. So they spent tons of money on them. In the second decade, they had top-level *offensive* talent they needed to keep. So they spent tons of money on them. And they're not *acquiring* offensive talent more than defensive talent. They're still drafting pretty heavily defensively. They're just not *keeping* any of them, because they're not worth it.
#31 by Hoodie_Sleeves // Dec 20, 2017 - 2:44pm
I'm not talking about just the Patriots, and I've made that patently clear. You'd understand that if you'd stop just repeating your nonsense talking point, and actually read the post.
Across the league, spending money on offense is more productive than spending it on defense. You don't have to have Tom Brady for this to be true.
#66 by RobotBoy // Dec 21, 2017 - 4:56am
'They've just gotten ridiculously lucky in acquiring cheap offensive talent, either through a late-round QB (Brady), a disgruntled Hall of Fame receiver (Moss), a misused elite slot receiver (Welker), or a second-round TE (Gronkowski).'
So, about that...
I'm not sure that acquiring Wes Welker was 'ridiculously lucky' in the sense of finding the Hope Diamond stuck in your tire tread. The same goes for Moss, and Gronk. A lot of other teams had the same opportunity to be so 'lucky.' Teams didn't draft Gronk because of his injury history. Teams didn't go after Moss because of his attitude. Those are the sorts of gambles that Belichick makes all the time. They don't always work out, either, but the ones that don't still make the pattern more clear. Dominique Easley was one of those that failed but the risk assessment was similar.
Brady was different of course, but it was still BB who took a flyer on him, BB who kept him on the roster, BB who didn't give the All Pro QB Bledsoe his job back when he'd recovered. After a while, it isn't luck.
Let's not forget that since 2000, the Pats have not had the privilege of high draft picks so they've been forced to gamble in order to acquire elite talent.
Speaking of luck, I saw a great deal of moaning about the Steelers not getting to face the Pats fully healthy. However, according to Man-Games Lost, the Pats lead the league this season in quality players lost.
#46 by roguerouge // Dec 20, 2017 - 4:01pm
Like many teams that aren't the Redskins or Dolphins, the Patriots do try to target undervalued talent wherever they can find it rather than top talent at the peak of their value. Their modus operandi has always been to have a B+ everywhere so they can target the other teams' weaknesses, rather than a stars and scrubs budget. They're not going to turn up their nose at top talent, of course, or even avoid spending for it (as in the cases listed above) but getting Suh at mega-millions is never going to be something BB does.
#47 by Will Allen // Dec 20, 2017 - 4:33pm
I think it is a real mistake to think Belichik is devoted to any particular theory of priorities in roster building, other than getting good enough talent at a low enough price. He's really great at seeing how particular players, who aren't budget busters, can fulfill particular roles, and at effectively modifying player behavior so as to avoid mental mistakes that lose games. Where I think he separates himself even from other HOF coaches is being ruthlessly unsentimental with regard to veterans who have won a lot of games for him in the past.
#49 by aces4me // Dec 20, 2017 - 5:02pm
I disagree. I think the "special sauce" in the NE long run has been the "plan" behind how he builds teams. He has some plan of resource allocation that he uses as a guide (subject to modification due to outrageous deals). He also has the characteristics you mention which are also vital to make a plan work but I really do think he has an overarching resource allocation scheme.
#52 by Will Allen // Dec 20, 2017 - 5:24pm
I don't think that can be the case, because positional talent allocation among college prospects is entirely random, as is, to large extent, which players in the draft turn out to have their value approximate their draft status.
#51 by Anon Ymous // Dec 20, 2017 - 5:22pm
I think it is a real mistake to think Belichik is devoted to any particular theory of priorities in roster building
I'd say the complete opposite. It would be foolish - nay, downright absurd - to think Bill doesn't have overarching guidelines around roster building. In fact, it is these guidelines that are what enable him to ferret out bargains and underutilized skill sets.
As Bill says, "we're not collecting talent, we're building a team." :)
#53 by Will Allen // Dec 20, 2017 - 5:28pm
Yes, but because your priorities can't consistently match what talent is available in the draft, you really can't adhere to your priorities too closely. Yes, you are building a team, but no, you can't guarantee what parts will be available, so no, your plan better not be too set in stone.
#55 by aces4me // Dec 20, 2017 - 5:35pm
I suspect (obtaining North Korea's 5 year plan would be easier to get than Bill's plan) the resource allocation scheme is more akin to what percent of salary cap will be devoted to Oline,Dline,wide receiver group, etc than anything else. So the team has to mix rookie contracts with free agents to obtain the best unit they can while sill meeting the plan. There has to be some flexibility for unexpected bargains and injuries but there has to be some kind of blueprint.
#50 by Anon Ymous // Dec 20, 2017 - 5:14pm
Given that most of the players mentioned (Brady being the lone exception) were intentional arbitrage opportunities, I don't see how you could argue that there was no repeatable strategy involved. It's certainly been a repeated strategy, so I see no reason why we can't infer repeatability.
All that said, you have a reasonable point that ^this^ strategy may be more at play than the larger cap strategy offered by Hoodie.
#57 by Richie // Dec 20, 2017 - 5:55pm
Doesn't that usually end up happening when it's convenient for the Patriots? If they have extra cap space, they'll renegotiate Brady's contract to take advantage of the current cap space.
I have no source for my theory.
His Cap Hit rank among QB's for the past 7 seasons (per Spotrac): 19, 18, 14, 11, 5, 16, 7. His average cap value over those 7 years is about $15M/year, at a time when the top-paid QB's are averaging about $20M.
So Brady's generosity gives New England about $5M/year to spend elsewhere.
I wonder if Brady would still be willing to sacrifice that kind of money if his wife wasn't already worth twice as much as him. I have a feeling he might.
#58 by Will Allen // Dec 20, 2017 - 7:08pm
Over the last 12 years of Peyton Manning's career, he earned about 50 million more than Brady, in that period. Drew Brees, who was still on his 2nd round rookie draft pick contract at the beginning of that period, only earned about 8 million less than Brady.
#60 by Richie // Dec 20, 2017 - 9:17pm
The average NFL cap from 2003-2015 (excluding the uncapped 2010) was about $120M. So it's a difference of about 3.4% per year. But an extra $4M/year can be worth an additional 2+ players above replacement level on the roster.
#61 by Will Allen // Dec 20, 2017 - 9:51pm
The 2004 cap was at 80 million, the 2015 cap at 143 million, so say the average was at 112, that makes a little more than a 4 million dollar difference about 3.75%. A 3.75% advantage, in the hands of smart person, is not trivial.
#67 by Will Allen // Dec 21, 2017 - 7:18am
Eh, I will never, ever, criticize any player for putting the pursuit of cash as the primary objective in his football career, with everything else very distant in importance. At the same time,if a player decides other objectives can reduce the relative importance of making money, hey. It's their life.
#72 by Richie // Dec 21, 2017 - 12:56pm
I don't fault him for earning all the money he can.
But I always wonder, what is he going to do with $250M that he couldn't do with $200M? (Plus all the millions he's making as a pitch man.)
Just the interest on that money should provide over $10M/year in income.
I have no comprehension of how rich people live.
#71 by Aaron Brooks G… // Dec 21, 2017 - 9:20am
The 2005 Pats had a bad defense, but from 2006-2009, they were high-ranking in terms of points, yards, and scoring %. There as some vacillation based on turnover luck, but they didn't start hemorrhaging yards until 2010. They hemorrhaged everything from 2011-2013; it was just a bad defense.
They do tend to give up more yards than their points allowed would suggest. However, in years where they really give up yards, they give up points, too. It's not like they give up 80 yards per drive and block a ton of FGs.
I don't think Belichick has a rigid plan. His roster construction has varied too much. He started with an old-school Favreian offense based on power backs and bombs and a defense that was a poor-man's version of the Carroll Seahawks -- all LBs and slobberknocking DBs. He then moved to an offense that was based on having Dante Scarnecchia, Randy Moss and 16 tiny slot receivers lined up all over the place. This was replaced eventually by substituting two TEs for Moss and one of those slot guys.
His defenses are still DB-based (although this hasn't worked as well in some years, either because of talent or fit), but have started drifting away from the heavy LB focus it had in the earlier years. What's sort of interesting is how much his roster construction these days looks like Andy Reid's rosters -- it's heavy on lineman and receivers/DBs, is built on a turnover-averse QB, and considers LBs and RBs completely fungible. Reid valued RBs more and WRs less than Belichick (his Eagles and Chiefs teams had some terrible WRs and some really good RBs) and LBs less and DL more. But the trends are similar.
A lot of this is because Belichick has been able to count on the same QB for 16 of the last 17 seasons and all 34 playoff games. This vastly simplifies offensive roster construction, because you never need to radically alter philosophy to fit changing talent (and part of why it's so valuable to lock in a quality QB). That this works less well on defense is why that's probably changed more. That said, you can do it on defense. Pittsburgh and Baltimore are examples of that.
#73 by Richie // Dec 21, 2017 - 1:00pm
"I don't think Belichick has a rigid plan."
I think there is a moneyball aspect to his plan, where he tries to take advantage of rule changes and opponent strategies in an efficient way.
I have also wondered how many other people in the organization even know all of his strategies. It doesn't seem like former Patriots front office & coaches have really tried to emulate his roster-building or in-game strategies. Or maybe just haven't done it successfully.
#75 by aces4me // Dec 21, 2017 - 1:24pm
Because a plan is necessary but not sufficient. You still need to be able to evaluate talent, find roles for lesser talent players, and be able to ruthlessly implement the plan. Those things Bill does well that other might not be able to emulate.
#82 by RobotBoy // Dec 21, 2017 - 11:08pm
Exactly. I imagine there are lots of different 'plans' that Belichick implements giving an entire host of mitigating factors. Moneyball type price points where a certain type of of player is undervalued and therefore someone he'll find a place for. Wes Welker he specifically went out and acquired. It took a long time for the rest of the league to catch up on true value of great slot receivers, which is why Edelman remained such a bargain (the slot is neglected no more, however). Positional flexibility is extremely important, with jack-of-all-trades being more useful to him than players who are great at one thing; total football, so to speak. Opponents frequently say that the Pats adjust their approach according opponent's weaknesses and give very different looks based on that. Every team does all of these things to a certain extent but the Pats are certainly among the most, if not the most, dynamically flexible team in the league.
#84 by Mountain Time … // Dec 22, 2017 - 1:23am
He then moved to an offense that was based on having Dante Scarnecchia, Randy Moss and 16 tiny slot receivers lined up all over the place
AHA! So that's Belichick's secret! He discovered previously unknown slot positions all over the field!
#11 by Anon Ymous // Dec 20, 2017 - 10:01am
The bigger issue here is recency bias. NE was legitimately terrible in September, since then they've been close to average. If the defensive DVOA was weighted, it would probably align with RJ's expectations more.
#13 by dmstorm22 // Dec 20, 2017 - 10:04am
That said, they've shown in recent weeks that problems still remain (sure, somewhat due to injury).
Also, in that stretch when they went 6 or 7 games allowing less than 17 points, there was a lot of fortunate things that went their way to keep that streak alive:
- handful of missed field goals against them, not all 50+ yarders
- well-timed plays like ASJ fumble out of EZ
- multiple 4th down stops in the red zone or 4th-and-goal (again, good play by them)
They are nowhere near as bad as they were early in the season, but even by weighted D-DVOA they are 24th.
#25 by Hoodie_Sleeves // Dec 20, 2017 - 1:40pm
The problems that still remain are largely due to the fact that their linebackers are terrible.
Which is largely because of injuries - the Hightower injury hurt, and the Mclellin one hurt, but frankly the injury to Kyle Van Noy has been much worse. Van Noy has actually been really good as a Patriot - which makes me wonder what the hell the Lions were trying to do with him.
At this point they've basically got the 4th and 5th and 6th guy on the depth chart playing LB - Flowers was someone they traded a 7th for in preseason, Roberts has his uses but isn't a 3 down LB'er by any means, and Harris is just too slow.
There should be significant improvement if/when Van Noy comes back - because the rest of them can be relegated to just doing what they're good at - but yeah, they're thin.
#17 by LyleNM // Dec 20, 2017 - 10:55am
Imagine a game where the opposing offense has 10 drives that all start at the 15 yard line. Each drive lasts 10 plays and gains 40 to 60 yards, and 4 of them result in field goals.
DVOA sees that appropriately as great special teams play to have the opposing offense start at the 15. If somehow those drives all started at the 30 instead of the 15, that same defensive performance would have allowed significantly more points. The credit is correctly apportioned in DVOA.
#83 by MJK // Dec 21, 2017 - 11:53pm
Except that if all those drives had started at the 30 instead of the 15, that same defensive performance probably wouldn't have happened.
The above discussion about talent notwithstanding, my own theory is that the Patriots defensive playcalling is likely responsible for the recurring disparity between their DVOA and their scoring defense.
I think the one thing that Patriots defenses have in common (except for early this season) is that they focus on not giving up the big play. Edge rushers are coached to stay in their lanes and maintain contain, rather than giving a freer hand to tee off on the CBs ang possibly get more sacks, but also give up big running plays to the QB. Interior linemen play a 2-gap style that allows offensive runners to get to or just past the LOS but not much further (hopefully), rather than shooting the gaps and getting tackles for losses but also giving up big runs. DB's play disciplined rather than gamble on turnovers and big plays. The defenses play this way because, when you are facing anything other than an elite offense starting deep in its own end, it is the right way to play, especially when you have an elite offense of your own on the other side of the ball (like the Manning-led Colts, starting deep in their own end probably matters to the Patriots offense less than for less offensively strong teams, so having a defense that gives up field position but not points is more acceptable).
But once the field gets shorter and the benefit of the big play drops, the Patriots start playing more aggressive. To my eye, it seems that they start playing very aggressive in the red zone.
#85 by Richie // Dec 22, 2017 - 12:25pm
I think the only flaw with the "don't mind starting deep in their own end" theory is that the Patriots are usually amongst the teams with the BEST average starting field position, year after year.
But that includes all field starts, not just following punts.
#62 by RobotBoy // Dec 20, 2017 - 11:40pm
Interesting. Is that true? I had assumed DVOA makes adjustments for red-zone and scoring defense, close game situations, etc., but are those seen as not predictive and the main emphasis is 'success' as measured by down and distance.
Losing Hightower hurt them more than people seem to realize. He wasn't really himself at the beginning of the season, either; my sense is that the reason they tried him on the outside is that they could save him some wear and tear. Then as soon as he went back inside, he was done.
That team has shed a great deal of LB and DE talent over the past couple of years. Although they didn't always play to their abilities, Jones and Collins were pretty remarkable players, so it's not surprising that they haven't been easily replaced. Trey Flowers is very, very good with an even higher ceiling but there's no player who really gives them a defensive identity in the way that Wilfork, Bruschi, Law, McGinness or even Ninkovich did. McCourty, I guess, might be the closest thing to it; cerebral player with a lot of range, underrated but hardly Hall of Fame. Belichick doing it with scheme and depth although there have been breakdowns that I've rarely seen before so late in the season.
I noticed that the Patriots have the biggest gap between estimated and actual wins - not quite two games. I seem to remember that from last year as well. Two 'Wins above Replacement' is a pretty good rough measure of what BB is worth solely as game-day coach, all else being equal.
#86 by ClavisRa // Dec 24, 2017 - 5:09am
You're getting to the heart of the matter. The Patriots play situational football. And here's the kicker: it's all situational football. A play is not just a play, and a yard is not just a yard. And certainly a score is not just a score.
The defensive calls depend clock, score, field position, opponent tendencies, personnel match ups, etc. The Pats practice low risk football as a rule: don't give up big plays, and don't turn over the ball. They give up higher median yardage to limit explosive and scoring plays.
They were #1 in scoring defense last year. They are currently #6 this year, despite a terrible start for a defense integrating a lot of new pieces. That's not luck, and it's not the offense. (Plenty of teams have successful offenses without an automatic boost to their D.)
#20 by Eddo // Dec 20, 2017 - 11:15am
It's talked about in this week's Any Given Sunday. The Rams' offensive DVOA was 3.9%, their offense was struggling in the red zone early (until Pharoh Cooper's punt return to the Seattle one). It also notes that this week's game was one of four on the season where the Rams' passing offense DVOA was less than 10%.
#29 by Eddo // Dec 20, 2017 - 2:25pm
Right. The Rams' high scoring offense (in this game) is related to the discussion of the Patriots' low points-allowed defense. The Rams' defense and special teams were outstanding against the Seahawks, and put the offense in very favorable situations. The Patriots' offense and special teams are generally very good, and put their defense in favorable situations.
#22 by NoraDaddy // Dec 20, 2017 - 11:35am
How are blocked kicks handled by DVOA? The Eagles blocked 3 separate kicks in their game and their ST DVOA when down. The rest of their ST play against the Giants was very average and I was surprised to see DVOA go down (albeit only by 0.1).