Print and Go Back ESPN.com: 2010 [Print without images]

Monday, February 14, 2011
Value-Based Drafting, 2010 in Review

By Christopher Harris
ESPN.com

It's a fantasy football debate that will likely never be fully resolved: What's the best positional draft strategy? Should you still, as the old wisdom of fantasy in the 1990s would have it, always draft a running back in the first round? Or do these days of committee backs and anxiety over high carry totals compel you to cast your lot early with the NFL's more explosive passing games?

As I seem to write constantly leading up to fantasy drafts every year: there is no "correct" way to look at positions, only players. If you could've looked into the future last August and understood what was about to happen during the season, you'd have answered the "Should I draft Adrian Peterson or Chris Johnson first overall?" debate with a simple answer: Draft Arian Foster. If you could redo the debate which had you flummoxed deciding between Rashard Mendenhall and Randy Moss at the tail end of your 2010 draft's first round, well, I think we know by now that debate was less about which position to draft, and more about which player was going to get traded twice because his skills no longer make up for his attitude.

Nevertheless, I think there's something to be learned by taking a look at the positional trends of 2010. That's where Value-Based Drafting comes in. For frame of reference, here's a piece from last summer in which I broke down VBD in depth. The upshot is that VBD is a way to compare the relative values of quarterbacks, running backs, wide receivers, tight ends and defenses, so that we can get a sense of whether there's an advantage in drafting one position early above all others. After all, in absolute terms, you might look at 2010's fantasy point totals and conclude pretty clearly that quarterbacks are the most valuable players in fantasy:

Thinking in purely absolute terms, you'd be tempted to say, "Forget about drafting running backs in the first round! Obviously, it makes more sense to get a quarterback! They're a source of high, safe points each and every year!" But of course, even the quasi-savvy fantasy player knows that's unwise: Precisely because QBs are by their natures a consistent source of points, and because RBs and WRs fluctuate far more wildly, you can easily make an argument that in relative terms, it makes a good deal of sense to wait on a QB, because the one you wind up with might not be Aaron Rodgers, but should still be just fine.

Value-Based Drafting is one way to quantify positional scarcity to get a sense of which positions were more valuable in the '10 season. Conceptually, the idea is to compare the top point-earners at each position to some "baseline" player at the same position. If we can get a sense of how much better the superstar QBs are than the baseline QB, and compare it to how much better the superstar RBs are than the baseline RB, we can start to be able to compare positions.

My preferred method of establishing baseline players comes from Footballguys.com: In a 10-team league, I look at how many QBs go in the first 10 rounds, and that number represents the rank of my baseline QB. I look at how many RBs go in the first 10 rounds, and that's my baseline RB, etc. The logic is that most players who'll be fantasy starters get drafted in the first 10 rounds of the average draft, and therefore the last such player at each position represents the final guy you'd take to fill the spot you didn't use if you eschewed, for example, an elite quarterback in the first or second round. It's a simplified "value over replacement player" that puts us in the neighborhood of how difficult it would be to replace a given player at a given position.

Historical statistics teach us that in a 10-team draft, on average, 16 quarterbacks tend to go in the first 100 picks (each team drafts a starter, about half the teams draft a backup), 38 running backs tend to go in the top 100, as well as 33 wide receivers, nine tight ends and four defenses. That means for the 2010 season, the following men qualify as our "baseline" players:

Now, there's no question there are mitigating factors with these particular players, one that a purely statistical evaluation method is going to miss. Orton was on pace for a much better season before getting yanked in favor of Tim Tebow for three games. Floyd might've been higher on his positional list without his injury troubles. Nevertheless, if we could go back and redo a 10-team fantasy draft and fill each team in accordance with the positional numbers I listed above (i.e., 16 QBs, 38 RBs, 33 WRs, 9 TEs, 4 D/ST get drafted), these five players would roughly represent the final few picks of the 10th round. They are our baseline players.

Now our task is to compare each of these guys to the players who finished ahead of them at their respective positions. The goal? Establish how much better than Orton, for example, Michael Vick was. Sure, his season was spectacular. But was it so spectacular that it was worth picking Vick ahead of the elite running backs? After all, if the baseline QB got you 200 fantasy points, were Vick's 300 fantasy points as valuable as they look? Without further ado, here are the top 100 fantasy players in 2010, according to VBD:

Wow. Let's just say that running backs made something of a comeback, eh? Instead of seeing one RB and nine QBs in the top 10, as we did when we looked at absolute fantasy point totals only, now we see the first 12 spots held by rushers, with Mr. Vick clocking in at No. 13, in a tie with the first wideout off the board, Brandon Lloyd. And the clear dominant force in fantasy, at least from a VBD perspective, was not Vick, but rather Arian Foster. His 81-point gap over second-place Adrian Peterson (and his 228-point gap over "baseline" running back Ricky Williams) made him fantasy's clear MVP for 2010.

By way of comparison, this is what the top 10 VBD performers from 2009 looked like:

So you wouldn't instantly put Foster's '10 season in a league entirely of its own; in fact, he and Chris Johnson circa '09 were basically equally dominant. But as you can see, this year's best running backs were even more top-heavy than last year's, while this year's quarterbacks and wideouts had fewer top-end standouts.

Does this automatically mean you should take a running back in the first round of your 2011 fantasy draft (assuming, of course, that an NFL season will actually be played this fall)? Of course not. As I always write, players are what wins fantasy leagues, not positions. Draft the best players, and you'll be fine. But I do think these VBD numbers are sobering for those who'll hurry to say, "Because there are so many committee backs, you should shy away from taking rushers early." If you draft the correct running backs, you'll be operating from an advantage that few (if any) quarterbacks or receivers can supply.

Of course, there are several things that a system as (let's face it) mathematically simple-minded as VBD can't account for, which is why we should take none of these lists as utter gospel:

Week-to-week variability. It's all well and good that Darren McFadden is sixth on our VBD list. I mean, no question, he had a breakout season. But his week-to-week variability was a killer. Compare his weekly lines to the three men directly below him, LeSean McCoy, Michael Turner and Rashard Mendenhall, and you'll see one of these things is not like the other. Forget the games he sat due to injury; even when he played, Run-DMC had a weekly standard deviation of nearly 12 fantasy points. By contrast, McCoy (6.9), Turner (8.6) and Mendenhall (6.0) achieved their equally high point totals without big ups and downs. Yes, it was nice to own McFadden in Weeks 7 and 14, when he scored 43 and 38 fantasy points, respectively, but not as nice to own him from Weeks 9 through 12, when he managed a total of just 16 fantasy points. And all this is to say nothing of players like Austin Collie or Jahvid Best, who show up on our list despite crushing their fantasy teams with season-altering injuries in midstream, or a guy like Vick, who missed most of four midseason games, yet still was huge for fantasy teams come playoff time.

Unlikely studs. It's all well and good to proclaim running backs the way to go in the early rounds of your fantasy draft, but how many of you would've put Foster in the elite category before the season began, especially after the Houston Texans drafted Ben Tate last April? Or to take more extreme examples, how about Brandon Lloyd and BenJarvus Green-Ellis? Heck, by the letter of the law, Ray Rice had a better fantasy season than BJGE. But you won't find many Rice owners crowing about it. His expectations were much higher than finishing 10th among fantasy rushers, while Green-Ellis came from nowhere to dominate goal-line touches in New England. Automatically proclaiming you have a bias toward taking a rusher in the first round doesn't mean you'll pick the right rusher. (For this reason, one of my offseason projects is to try to develop a VBD "Reliability Index," to get a sense of the likelihood that a player tends to wind up in an acceptable VBD range on a season-to-season basis.)

Late bloomers. Someone like LeGarrette Blount suffers on a season-long VBD survey because he wasn't really given an opportunity to play until Week 7. From that point on, he was a monster, but his VBD rank still puts him tied with Thomas Jones for 49th overall (24th among rushers). Of course, anyone in contention down the stretch this November and December knows which of those running backs you'd rather have had toiling for your team. You could put players like Chris Ivory and Ryan Torain in that same category.

Bottom line, are these VBD results predictive of what will happen next year? On a player-by-player basis, in large part, no, and it's obvious why: at bottom, they're simply based on the previous year's statistics, and the NFL is a year-to-year league. When players like Ricky Williams, Brett Favre and Joseph Addai grade out very well in 2009's VBD, it's because they performed well in fantasy, because they performed well in real life. As soon as they started to stink it up for your fantasy team? Naturally their personal VBD numbers stunk, too. The most useful thing VBD does is allow us to look among positions, and figure out, for example, where a "spectacular" fantasy quarterback season (like Vick's) should rate among the best of other positions. And for me, the bottom line continues to be: no matter how amazing you expect a given signal-caller or receiver to be next season, he's not going to be worth as much as a running back you feel relatively comfortable will be a top-10 guy. And that's a notion worth keeping in the frontal lobe of your fantasy football brain.

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.