|ESPN.com: Draft Kit||[Print without images]|
It's a set of arguments I've been unable to kill:
"Here's another reason you should draft Player X! His schedule is going to be so easy this year!"
"Stay away from Player Y this season. He's got a brutal schedule against the run."
"It's OK to draft Player Z. Just be ready to trade him at midseason. His fantasy playoff schedule is terrible."
The idea that we can look at an NFL team's schedule and decide -- months in advance -- whether it is favorable or unfavorable for opposing skill players is pure hubris. But thousands upon thousands of people do it anyway. They ignore the fact that the NFL is pro sports' most unpredictable league, producing several "out of nowhere" contenders on an annual basis. They ignore significant defensive player movement (via free agency and/or trades), which can substantially alter an NFL defense's outlook. They ignore the possibility (nay, likelihood) of injuries, which can significantly affect how a defense performs and even how it approaches the game. And they ignore the pure randomness that is such a large part of a league in which the margin that separates "good" teams from "bad" ones is paper thin.
Do some defensive themes repeat from year to year? Of course they do. In each of the past five seasons, the Pittsburgh Steelers and Baltimore Ravens have ranked in the top five in rushing yards allowed. It's plainly a point of emphasis and an area of strength for these 3-4 goliaths. The Oakland Raiders have numbered among the five worst rush defenses (in terms of yards allowed) in each of the past four seasons. It's obviously a serious problem the team hasn't yet solved. But in a league in which teams make massive improvements and/or steep declines in the span of a single year, these examples are much more exception than rule.
But qualitative logic apparently isn't enough to rid the fantasy world of this insidious preseason crutch argument. So let's break it down. I'll look at three defensive categories over the past several years and focus on how repeatable or unrepeatable defensive performance is on a season-by-season basis.
If you could accurately predict which defenses would allow the most overall points in the 2011 NFL season, you'd be well on your way toward figuring out which fantasy QBs, RBs, WRs and TEs have the most favorable schedules. But how does one conjure up such a prediction? You could make qualitative arguments. ("Whichever team signs Nnamdi Asomugha is going to be tough to throw against.") You could come up with algorithms or run computer simulations. But most folks do what's easy: They look at last year's performance. The Steelers, Green Bay Packers, Ravens and Chicago Bears were the four stingiest defenses to score against last season, so (goes the logic) they're defenses to keep your fantasy studs away from this season.
Great. Except what would you say if I told you the Steelers, who allowed a league-low 232 points in '10, gave up 324 points in '09, good for only 12th best in the NFL? Or that the Dallas Cowboys' shamockery of a defense allowed the second-most points in the league in 2010 -- a whopping 436 -- a season after allowing only 250, which was second-fewest in the NFL? Last season alone, nine of the league's 32 defenses (a full 28 percent) saw their points allowed change by six or more points per game:
And this isn't an uncommon occurrence; it's happened 26 times in the past four seasons. Perhaps most unnerving is the fact that it isn't the same teams flip-flopping back and forth between high and low points-allowed totals. Eighteen franchises have seen their points allowed per game increase or decrease by six or more points in at least one of the past four seasons. Six points might sound a bit arbitrary (and it is, a bit), but when you consider that the average NFL defense allowed exactly 22 points per game last season, six points is a rather huge fluctuation. Yes, I think you can say looking at last season's points allowed is better than randomly guessing which defenses will be difficult opponents. But not leaps and bounds better.
"Fine," you say. "It's tough to predict which defenses will give up a bunch of points based on last season. But I don't care about points nearly as much as I do about yards. Scoring might come and go, but if I remember a defense getting marched up and down the field last season, it's sure to get marched up and down the field this season, too."
Let's start with run defense. As I mentioned earlier, there are some run defenses that have stayed steady (for good or bad) over the past five seasons. Five franchises have featured defenses that have ranked in either the top 10 or the bottom 10 in rush yards allowed in that span; the Steelers, Ravens and Minnesota Vikings have stayed in the top 10, while the Raiders and Cleveland Browns have remained in the bottom 10. Heck, if we want to get charitable and make our window the past three years, we can add the New York Jets on the strong side, and the Lions, Broncos and Indianapolis Colts on the weak side. How many of those nine squads do you think will stay in the "extremely tough to run against" or "extremely easy to run against" column in '11? All of them? The 2010 Chiefs and Philadelphia Eagles beg to differ. The Chiefs had put together three straight execrable years versus the run entering '10 and became a much tougher matchup last season. The Eagles had put together three straight strong seasons against the run and reversed course:
Yes, I agree that if I'm considering drafting an RB from the Browns or Bengals (say, Peyton Hillis or Cedric Benson), I have a smidgen of concern. Four games against the Steelers and Ravens probably will be tough. But should I significantly downgrade these players based on the possibility of tough sledding in one-quarter of their games? I tend to think not, especially not when I saw Hillis notch 180 total yards and a TD against Baltimore in a game last fall. No, to substantially downgrade such players, I'd need to be able to look at the rest of their schedules and decisively conclude that bad matchups consistently abound. I just don't think that's possible. Look at the historical rush yards allowed by several other of the best run defenses from last season:
For me, the huge variability here makes this stat untrustworthy as a predictor. Geez, look at that Bears roller coaster. It turns out I can't even look at the schedules of RBs who play the Steelers and Ravens twice per season and definitively find other off-putting opponents.
Folks, the story is less reassuring when it comes to pass yards allowed. In that category, wild season-to-season swings seem more like the norm than the exception. There is exactly one defense that has ranked exclusively in either the top 10 or bottom 10 of pass yards allowed in the past five years: the Raiders, who've finished second, seventh, 10th, eighth and first, and who incidentally are primed to lose Asomugha this summer. No other pass defense qualifies. And even if we limit our scope to the past three seasons, only the Tampa Bay Buccaneers join the Raiders as consistently good in pass defense, while the Jacksonville Jaguars and Seattle Seahawks have been three years' worth of consistently bad.
No matter how you slice the pass-defense numbers, they're alarming, and let's face it: Good yardage stats in this category often have less to do with the quality of your pass defense and more to do with how easy it is to run against you. Look at last season's complete top 10 and how those teams fared in the previous four seasons:
Up and down and up and down. Is there anything here that should make a fantasy owner say, "Clearly, I have to avoid Team X as an opponent for my QB or WR"? Maybe I'll grant you that the prospect of, say, Brandon Marshall facing Darrelle Revis a couple of times this season puts a slightly negative edge on Marshall's fantasy value. But can I really look at Marshall's entire schedule and say, "Oh, based on who he's playing, I shouldn't draft that guy"? Recent history shows that the Bills, Chargers and Giants -- all of whom Miami also plays this season -- could be average at best against the pass in '11.
"All right," you say. "Enough with these season-long counting stats that don't actually affect my fantasy bottom line. I'm going to go strictly by what really matters: how many fantasy points a defense allows to a particular position."
In general, my instinct tells me that fantasy points allowed actually would be a less reliably predictive stat, because as anyone who's ever lost his or her fantasy championship game to Aunt Hilda (who drafts her team based on player handsomeness) can tell you, fantasy points are sometimes pretty random. Touchdowns -- a primary source of fantasy points -- are extremely difficult to project on a week-to-week basis. For instance, if a defense allows Adrian Peterson to amass 98 yards on a particular drive, but then Percy Harvin catches a 1-yard TD, well, the fantasy points amassed are misleading.
But let's look at the repeatability of fantasy points allowed anyway. There are tons of ways I could slice this information, but I don't want to drown you in data. Let's look at just a few key samples. Here were last season's top and bottom five defenses in fantasy points allowed to QBs, and how they each fared in those categories the past five seasons:
The difference from '09 to '10 is stunning. I mean, based on '09 numbers, you might actually have looked at the Broncos, Redskins, Seahawks, Cowboys and Texans, and predicted that as a group they'd fare better allowing fantasy points to QBs than the Chargers, Saints, Panthers, Steelers and Bears would. And obviously, that didn't happen. Again, what's notable here is how up and down the numbers are. To say that I could have looked at the Broncos' defense last summer and predicted they'd be so hospitable against fantasy QBs when they'd been fourth-least-hospitable the season before? Sorry, those are fortune-telling skills I don't have.
I'll admit that the table displaying fantasy points allowed to RBs has more year-to-year consistency. At the very least, no defense that finished among the 10 best at preventing RBs from scoring fantasy points in '10 finished outside the top 20 in that category in '09. I'll give the same chart for RBs that I just gave for QBs:
Again, in the name of giving credit where credit is due, the Steelers and Ravens stand out here as being remarkable. The Jets have toughened up extensively under Rex Ryan. And in general, I guess you'd say year over year, we at least see fewer wild swings, although apparently there's no telling when a rush defense is suddenly going to go bad. You'll have to trust me on this: I stared at this data for a good long while, trying to find mathematically predictive factors that explain, for example, why the Redskins and Jaguars, both among the 10 most difficult defenses for RBs to score fantasy points against in '09, became No. 23 and 25, respectively, on that list last season. I can't find any.
Even if we accept that we've finally found at least a vaguely predictive worry-about-schedules model for fantasy -- look at RB fantasy points allowed for last season and draft RBs accordingly -- we still might be doing ourselves a disservice. Looking at 2010 fantasy points allowed to RBs, maybe the team with the most daunting schedule is the Patriots. They play the Steelers once, and the Jets and Dolphins twice, and that second Dolphins game comes during fantasy championship week. Stay away from New England RBs! Except guess what? Last season, the Pats played these five run-defending teams seven times. BenJarvus Green-Ellis still had a pretty good year, didn't he?
I'm not saying fantasy schedules don't matter. Of course they do, very much. I'm merely contending that more often than not, we're flat-out wrong about which defenses represent good matchups before the season begins, because last season's numbers often don't apply this season. There are a few carryovers, yes, especially examining fantasy points allowed to RBs. But are there enough on which to base your fantasy draft strategy? I think not.
Once we've seen the 2011 versions of NFL defenses, absolutely, I'll be all aboard the defensive-evaluation bandwagon, writing about what I actually observe. And in-season data is extraordinarily helpful. By the time November rolls around, looking at fantasy playoff schedules will be good for business.
But don't bother now. Seriously. Year after year, it's proven that we don't know what an easy schedule looks like. And this year? With the lockout, with rookies performing without organized team activities, with free agents arriving late and training camp rushed at best? Anything short of a crystal ball or a DeLorean-based visit from the future just isn't going to be compelling. Draft good players, and adjust to opponents once we're well into the 2011 season.
Christopher Harris is a fantasy analyst for ESPN.com. He is a six-time Fantasy Sports Writers Association award winner. You can ask him questions at www.facebook.com/writerboy and follow him on Twitter at @writerboyESPN.