Which recent NFL draft year provided the most value?

Former Rams coach Jeff Fisher and current GM Les Snead were responsible for drafting high-value Aaron Donald. Jeff Curry/USA TODAY Sports

You can't evaluate a draft class for a few years, they say.

Of course, general managers are afforded no such courtesy, but that's another issue for another time. Today we will use that benefit of hindsight to rank the draft classes of the past 10 years (ignoring 2016 and 2017 due to the aforementioned aphorism) because, heck, it's draft season and that's what we do. We'll also pick out the best and worst value picks -- commenters beware: value picks -- of those drafts.

How do we objectively do this? I won't bore you with the details unless you really want them, but to make a long story short(er) we're using Pro-Football-Reference.com's approximate value from every player's first three seasons (3AV) to evaluate each class from 2008 to 2015 (more information on the methodology here). In addition, we'll identify the best and worst value picks from those drafts based on every player's 3AV relative to the expected 3AV for that player's draft slot.

But enough with all that -- let's get ranking. From worst to first, here we go.

8. 2009

3AV per player: 6.7
Best value: LB Clay Matthews, Packers, pick 26
Worst value: LB Aaron Maybin, Bills, pick 11

Looking back, the 2009 NFL draft was rocky from the start. The only player in the top 12 who exceeded his expected AV total by more than 2 over his first three seasons was Michael Crabtree, who did it after holding out until October of his rookie season. Matthew Stafford took time to develop and failed to produce at the level of a No. 1 pick over his first three seasons, while other early picks, such as Darrius Heyward-Bey and Maybin, failed to live up to their draft-day hype.

The second-worst value in that year's class was Jason Smith, the offensive tackle the Rams selected with the second overall pick. A few years later, he and Maybin had lockers just a few feet apart with the Jets.

In addition to a rough first round, the 2009 draft also provided far less value in the back half of the draft than any of the other drafts since 2008.

7. 2015

3AV per player: 7.0
Best value: WR Stefon Diggs, Vikings, pick 146
Worst value: DL Arik Armstead, 49ers, pick 17

There were stars in this first round, for sure. Jameis Winston and Marcus Mariota, despite their faults, have outperformed their respective top draft slots, and Leonard Williams, Todd Gurley and Marcus Peters all produced at least 10 AV over expectation given where they were drafted.

But the second round -- especially after first two picks Landon Collins and Donovan Smith -- fell well short of expectation. Even though picks Nos. 120-150 or so proved to be a relative gold mine (that's where Kwon Alexander, Shaq Mason, Grady Jarrett, Diggs and Jay Ajayi were found) this was still a relatively weak draft overall.

6. 2013

3AV per player: 7.1
Best value: RB Eddie Lacy, Packers, pick 61
Worst value: DL Dion Jordan, Dolphins, pick 3

The Packers actually had the two best values in this draft, with OT David Bakhtiari, the 109th pick, coming in right behind Lacy. No question that Bakhtiari is the better of the two now, but don't forget how productive Lacy was early in his career -- in real life and in fantasy. Ironically, Jordan might have a better career outlook from this point forward than Lacy despite very different starts in the NFL. And they were teammates last year.

Overall, the 2013 draft actually had the worst first round of the eight drafts we're considering today.

5. 2010

3AV per player: 7.2
Best value: TE Jimmy Graham, Saints, pick 95
Worst value: G Vlad Ducasse, Jets, pick 61

Not only did this class produce some incredibly talented tight ends but the best ones were taken outside the first round and were therefore incredible values. Graham led the pack, but Rob Gronkowski (pick 42) and Aaron Hernandez (pick 113) quickly follow on the list of most valuable picks that year.

Ducasse never had success with the Jets but has bounced around the league over the past few seasons and started 12 games for the Bills last season.

4. 2012

3AV per player: 7.6
Best value: QB Russell Wilson, Seahawks, pick 75
Worst value: WR A.J. Jenkins, 49ers, pick 30

It's no surprise that Wilson produced more AV over expectation than any other quarterback in these eight drafts. In retrospect, maybe his height wasn't such a big deal.

Especially given that the Seahawks also drafted Bobby Wagner with the 47th pick, no other team in 2012 came even remotely close to outperforming its draft slots to the degree the Seahawks did.

Jenkins failed to make a catch for the 49ers before he was traded to the Chiefs. He caught just 17 passes over the next two seasons for Kansas City and is now out of the league.

3. 2014

3AV per player: 7.6
Best value: DL Aaron Donald, Rams, pick 13
Worst value: CB Justin Gilbert, Browns, pick 8

The first 20 picks of the 2014 draft trailed only 2011 in terms of value, with superstars such as a Donald, Khalil Mack, Odell Beckham Jr. and Zack Martin emerging from the first half of the first round. However, the draft did start off with two players who failed to match expectations in their first three years: Jadeveon Clowney and Greg Robinson.

The 2014 draft also featured five first-round receivers to outplay their draft slot's expectation over their first three years in Beckham, Sammy Watkins (barely), Mike Evans, Brandin Cooks and Kelvin Benjamin.

2. 2008

3AV per player: 7.7
Best value: OL Carl Nicks, Saints, pick 164
Worst value: WR Dexter Jackson, Buccaneers, pick 58

In retrospect, 2008 didn't have the flashiest top of the draft, but by the end the talent was there. One reason: More players in the sixth and seventh rounds contributed something to their teams than in some of the other drafts this decade. Plus there were some late-round gems such as Peyton Hillis (remember him?), Pierre Garcon and John Sullivan. Also among that group was Nicks, who isn't the sexiest value pick discussed today but is someone who started 45 games in his first three seasons and was a second-team All-Pro in 2010 (and a first-team All Pro a year later). It's hard to find that in the fifth round.

Jackson was a complete bust who returned 20 punts and 14 kickoffs his rookie year but didn't catch a single pass and never played a game beyond the seven he participated in in 2008.

1. 2011

3AV per player: 8.0
Best value: CB Richard Sherman, Seahawks, pick 154
Worst value: OL Derek Sherrod, Packers, pick 32

The class of the decade. Despite that, the first-round quarterbacks after Cam Newton at No. 1 were a mess, and those QB mistakes were compounded by the fact that so many talented players were selected around them. In a year in which the draft was kicked off with the selections of Newton, Von Miller, Marcell Dareus, A.J. Green, Patrick Peterson and Julio Jones, only three of the first 18 picks fell short of their three-year AV expectation: Jake Locker, Blaine Gabbert and Nick Fairley (Christian Ponder just surpassed his).

Had each of these drafts ended after 150 picks or so, 2014 actually would have ended up as the best class because that year caught up to 2011 in the middle rounds. However, thanks to some valuable late-rounders such as Marshall Newhouse, Brandon Fusco and Jason Kelce, 2011 secured its spot as the best class in the past 10 years.


Ranking the draft classes was simple. Using approximate value, which incorporates basic statistics and awards, we merely added up each class' three-year production and divided that total 3AV number by the number of players drafted in that year's class.

Determining the best and worst values was slightly more complicated. Though AV does evaluate all positions, it doesn't evaluate them all on the same scale. So we wanted to adjust for that. We separated out drafted players by position* and then compared their actual 3AV output to expected 3AV output based on calculations made by ESPN's Brian Burke derived from smoothed historical results from each draft slot. That yielded a mean and standard deviation for each position, which allowed us to calculate a z-score for the players in a given draft. The best and worst z-scores determined which players were the best and worst values for that draft.

* Position information was missing for some players in our data set, meaning the mean and standard deviation were imperfect, though any error was likely minor. Position information was entered manually for any player with an extreme AV result to ensure that all candidates for best or worst value were considered.