If you read through and followed my calculations in this space a week ago, first of all, bless you. Not that I'm the best statistician out there or anything, but it was a lot of numbers. It was more than 1,000 words, and I never named a single player.
Mathias Kiwanuka. There.
All that said, I was (am) trying to find an objective, math-based method for ranking team defenses week-to-week, as it has historically been a crapshoot of a position, especially if you aren't the owner of one of the top handful of units. Also, I think the defense's opponent is far more important than it normally gets credit for being — for example, if you told me that the Seattle Seahawks were playing a random team, and I had to guess the D/ST's production, I would say 10 or 11 points. Meanwhile, if you said the Jaguars, I'd guess 2, maybe 3. But then you tell me either team is playing the Chiefs, and I'm guessing somewhere between 4 and 7 fantasy points. With no hesitation.
In 10 games this season — with opponents as varied as the Jaguars, Titans, Broncos, and Cowboys — every single defense that has faced the Chiefs has scored in that range. Facing a team that doesn't turn the ball over, like Kansas City, limits the upside of a D/ST. On the flip side, the Giants love to turn the ball over, meaning no team has given up more points to opposing defenses.
Back to my original point. If you read last week's piece, you might have noticed that my formula was missing a piece. As a refresher, my formula from last week ranked defenses based on their average fantasy scoring (weighted double) and their consistency rating, as described in my Week 5 rankings, creating a ranking of all the defenses based on a neutral opponent. I then did the same calculations for each team based on the defenses they faced, for a ranking of all the opponents facing a neutral defense. From there, it was simply a matter of scaling the two lists to have a specific ranking.
There was, though, a missing piece — recency. Defenses get better and worse all the time; while it's foolish to see a defense go from poor production to dominant in one game, sample sizes in football become meaningful fairly quickly. For example, the New York Giants' defense had -0.833 (yes, negative) points a game in their first six games. In their four wins since, the unit has averaged 14.75 points. The awful early scores matter, sure, but should the improvement matter significantly more?
My formula last week didn't do much to account for this. The Giants' average was improving, sure, but the fluctuation meant that their consistency rating was going the other way.
I conceded last week that, while I see a team's consistency as important, it can't be argued that the actual score is moreso, which is why I double-weighted the average. It was simple, but inelegant, like using a semi truck to deliver pizzas. To fix the inelegance, and account for recency, I ditched the double-weighting, and replaced it with a team's average over its last four outings. Four games are enough to establish a new trend without incorporating so many games as to make the whole exercise pointless.
That gives the formula three variables — overall scoring average, last-four scoring average, and consistency rating. This is still a work in progress, but that phrase includes the word "progress," so ... good stuff.
Anyway, the following table is a ranking of the defenses against neutral opponents. In other words, it is a functional ranking of the defenses down the stretch, if you're picking someone up now and riding with them over the season's final six weeks:
|Rank||Team||Scoring average rank||Average (Last 4) rank||Consistency rating rank|
At the end of this piece, I am again going to include both my subjective rankings and my formula's objective ones. I did this a week ago, and the formula's rankings were slightly closer to reality than my subjective ones. Hopefully, with enough tweaking and fine-tuning, I'll get it even closer. I'll never get it perfect, or if I do, it'll be my last piece on this site, as I will immediately blow this popsicle stand for Vegas.
Until then, feel free to chime in in the rankings (again, because I got some good feedback last week, and thank you). Here is a look at a few defenses this week:
Cleveland Browns (No. 2 in my rankings, No. 6 in the formula) — The Steelers have two primary weapons — Antonio Brown and Le'Veon Bell. I'm sorry, I'm not buying in on Jerricho Cotchery yet. Well, the Browns do two things really well on defense: they stop the run, and they make Joe Haden do things. Haden held the Bengals' A.J. Green to two catches and seven yards Sunday, limiting Andy Dalton to only 93 total passing yards. On a team that relies on a running back and one dominant receiver, the Browns could have a field day.
Arizona Cardinals (No. 6 in my rankings, No. 5 in the formula) — I don't actually care much about the Tyrann Mathieu "redemption story," or whatever you want to call it. Frankly, as a (wannabe) numbers guy, I'm rarely excited by some off-field/court narrative. What I do care about, especially as it regards Mathieu, is that he was a third-round pick in 2013 who has become a stout defensive player, already winning Defensive Rookie of the Month once. He and Patrick Peterson are a stout combination, and it bodes well for the whole unit.
Jacksonville Jaguars (No. 25/23) and Houston Texans (No. 13/2) — My formula loves the Texans this week, based primarily on the success every defense in the world has facing the Jaguars. I think this game will be a heck of a litmus test as to whether the Texans are done-done, with the potential of 2-14 or 3-13, or whether they will be respectable down the stretch, maybe getting to 5-11 or 6-10. Practically, subjectively, that means that there is a wide range of slots both defenses could be ranked this week. But the Texans have two of their next three games against Jacksonville. Now's the time, Texans.
Now, to the rankings:
|Rank||My ranking||Opponent||System ranking||Opponent|
|1||Carolina Panthers||at Miami||Carolina Panthers||at Miami|
|2||Cleveland Browns||Pittsburgh||Houston Texans||Jacksonville|
|3||New Orleans Saints||at Atlanta||New Orleans Saints||at Atlanta|
|4||Tennessee Titans||at Oakland||San Francisco 49ers||at Washington|
|5||San Francisco 49ers||at Washington||Arizona Cardinals||Indianapolis|
|6||Arizona Cardinals||Indianapolis||Cleveland Browns||Pittsburgh|
|7||Kansas City Chiefs||San Diego||Green Bay Packers||Minnesota|
|8||St. Louis Rams||Chicago||Baltimore Ravens||NY Jets|
|9||Green Bay Packers||Minnesota||Tennessee Titans||at Oakland|
|10||Baltimore Ravens||NY Jets||Dallas Cowboys||at NY Giants|
|11||Chicago Bears||at St. Louis||Chicago Bears||at St. Louis|
|12||Oakland Raiders||Tennessee||San Diego Chargers||at Kansas City|
|13||Houston Texans||Jacksonville||New York Jets||at Baltimore|
|14||Dallas Cowboys||at NY Giants||Oakland Raiders||Tennessee|
|15||Indianapolis Colts||at Arizona||Pittsburgh Steelers||at Cleveland|
|16||New York Giants||Dallas||Tampa Bay Buccaneers||at Detroit|
|17||New York Jets||at Baltimore||St. Louis Rams||Chicago|
|18||Miami Dolphins||Carolina||New England Patriots||Denver|
|19||Tampa Bay Buccaneers||at Detroit||Miami Dolphins||Carolina|
|20||Denver Broncos||at New England||Indianapolis Colts||at Arizona|
|21||Pittsburgh Steelers||at Cleveland||New York Giants||Dallas|
|22||San Diego Chargers||at Kansas City||Jacksonville Jaguars||at Houston|
|23||New England Patriots||Denver||Kansas City Chiefs||San Diego|
|24||Washington Redskins||San Francisco||Denver Broncos||at New England|
|25||Jacksonville Jaguars||at Houston||Detroit Lions||Tampa Bay|
|26||Minnesota Vikings||at Green Bay||Washington Redskins||San Francisco|
|27||Detroit Lions||Tampa Bay||Minnesota Vikings||at Green Bay|
|28||Atlanta Falcons||New Orleans||Atlanta Falcons||New Orleans|