People often say that one should “wait three years” to grade a draft, but it only makes sense that anyone willing to offer analysis of players before a draft should be prepared to offer analysis of how teams did selecting those players.

To that end, the 2018 Consensus Big Board has a ranking of players by which to judge the draft picks that were selected. It would be easy to tally up the amount of draft capital spent and subtract it from the ranks spent — a method we’ve tried in the past — but the method doesn’t account for need.

At some point, flaws in an automated methodology should be expected, but this year we went one better and incorporated the need matrix produced by the people at Drafttek, creating a multiplier that roughly closes the historical gap between positional value in the actual draft and the historical player rankings the Consensus Big Board produces.

Basically, the Consensus Big Board ranks players by their “true” talent level with only a small positional bias while the NFL will “overvalue” some positions, like quarterbacks and edge rushers, compared to their true talent. The best quarterback will always be drafted ahead of the best punter. Once we account for that as well as the fact that needs are different for every team, we can more appropriately grade the draft.

First, let’s take a look at how teams did relative to the amount of draft capital they used (table is sortable):

NFL team Capital Spent Capital Gained ROI
Tennessee Titans 3170.8 3740.8 118%
Washington Redskins 4896.4 5295.1 108.1%
New York Giants 6522 6800.7 104.3%
Dallas Cowboys 5273.9 5274.3 100%
Arizona Cardinals 4403.3 4358.4 99%
Jacksonville Jaguars 3914.5 3854.4 98.5%
New York Jets 4723.2 4556.1 96.5%
Philadelphia Eagles 2434.2 2249.5 92.4%
Los Angeles Chargers 4634.5 4275.3 92.2%
Pittsburgh Steelers 4556.1 4193.6 92%
Tampa Bay Buccaneers 6560.4 6011.5 91.6%
Chicago Bears 5526.5 5052.5 91.4%
Green Bay Packers 5633.1 5063.5 89.9%
Cincinnati Bengals 6055.8 5440.4 89.8%
Oakland Raiders 5861.4 5214.1 89.0%
Baltimore Ravens 6837.9 6028.5 88.2%
Denver Broncos 7416.7 6537.9 88.2%
Buffalo Bills 5951.4 5231.7 87.9%
Miami Dolphins 5028.4 4368.4 86.9%
Houston Texans 3826.9 3299.2 86.2%
Indianapolis Colts 8238.5 7009.1 85.1%
Carolina Panthers 5252.3 4191.5 79.8%
Minnesota Vikings 4151.1 3264.9 78.7%
Detroit Lions 4387 3401.3 77.5%
Atlanta Falcons 3962.7 3067.5 77.4%
Los Angeles Rams 3625.5 2721 75%
Cleveland Browns 10134.5 7579.9 74.8%
New England Patriots 5088.2 3719 73.1%
San Francisco 49ers 6014.6 4347.4 72.3%
Seattle Seahawks 4337.4 2969.7 68.5%
Kansas City Chiefs 3734.4 2277.3 61%
New Orleans Saints 3909.8 2353.2 60.2%

Generally speaking, it will be difficult for teams to return 100 percent of the capital they invested in a player of equal value, especially after accounting for the fact that it is easier to lose value with the early picks.

That said, the Vikings had been at the top of the class in the category — placing in the top three every year since 2014 and typically returned over 100 percent value. Though many of the picks that generated positive value on the board didn’t return it on the field — T.J. Clemmings, Mackensie Alexander and so on — most of the players did, including Stefon Diggs, Danielle Hunter and Eric Kendricks.

This year, the Vikings basically earned a “C” with the Consensus Big Board after three straight years of “A” grades.

Though the Giants’ pick of Saquon Barkley over a quarterback was controversial — and I personally disagree with their strategy — the method here is a slave to orthodoxy, rewarding teams for following the value model that teams have for the past decade. For the most part, that’s a pretty good guide, but in these instances can create some harsh disagreements.

That shouldn’t take away from what the Consensus Big Board considers a slew of high-value picks, including Will Hernandez, Kyle Lauletta and Lorenzo Carter.

Tennessee knocked it out of the park, earning big points for Harold Landry and Luke Falk and — importantly — not losing points in many places. The Saints and Chiefs, however, lost points — and that’s without accounting for the extra first-round pick the Saints gave up to move up for Marcus Davenport. The big negative value moves the Saints and Chiefs made include the Breeland Speaks, Rick Leonard, Cornell Armstrong and Kamrin Moore selections.

So, which picks did the Consensus Big Board value the most?

Player NFL team Points Spent Points Gained Difference
Maurice Hurst Oakland Raiders 451.6 1343.0 891.4
Josh Rosen Arizona Cardinals 1813.1 2699.3 886.2
Derrius Guice Washington Redskins 897.4 1591.6 694.2
Harold Landry Tennessee Titans 1085.2 1693.1 608.0
Tyrell Crosby Detroit Lions 405.8 960.3 554.5
Ogbonnia Okoronkwo Los Angeles Rams 382.7 911.7 529.0
Josh Jackson Green Bay Packers 1037.1 1504.9 467.8
Quenton Nelson Indianapolis Colts 2076.6 2535.5 458.9
Duke Ejiofor Houston Texans 330.6 748.2 417.6
Equanimeous St. Brown Green Bay Packers 249.8 665.2 415.3
Josh Sweat Philadelphia Eagles 489.8 884.3 394.5
Tim Settle Washington Redskins 373.1 715.2 342.1
Ronnie Harrison Jacksonville Jaguars 662.6 999.0 336.4
Jamarco Jones Seattle Seahawks 357.5 672.4 314.9
Deon Cain Indianapolis Colts 307.8 601.2 293.4

Naturally, the board is not tuned into the medical issues that caused the falls for players like Josh Sweat or Maurice Hurst, but it is saying that these players have the best change of returning massive value for their franchises.

While the Hurst pick is the best “pick for value” after accounting for positional value and team need, I personally think the Rosen pick was a slam dunk. It was pretty close by the board’s definition of value anyway.

And who hurt teams the most?

Player NFL team Points Spent Points Gained Difference
Terrell Edmunds Pittsburgh Steelers 1281.9 553.1 -728.8
Tracy Walker Detroit Lions 727.6 81.6 -645.9
Breeland Speaks Kansas City Chiefs 1025.8 416.5 -609.2
P.J. Hall Oakland Raiders 915.2 338.1 -577
Baker Mayfield Cleveland Browns 3001 2435.8 -565.2
Rashaad Penny Seattle Seahawks 1300.7 744.2 -556.4
Frank Ragnow Detroit Lions 1455.5 923.4 -532.1
Hayden Hurst Baltimore Ravens 1340.4 821.3 -519.1
Rick Leonard New Orleans Saints 501.9 0.0 -501.9
Kolton Miller Oakland Raiders 1603.9 1127.2 -476.7
Darius Leonard Indianapolis Colts 1152.2 686.2 -466
Josh Allen Buffalo Bills 1997.1 1536 -461.1
Joel Iyiegbuniwe Chicago Bears 553.1 94 -459.1
Jordan Akins Houston Texans 635.6 203 -432.6
Denzel Ward Cleveland Browns 2285.8 1861.3 -424.5

This method can never be perfect. I strongly disagree with the Baker Mayfield value here, for example, but the board is ultimately implying that the Browns should have traded down to acquire Mayfield because he was the third-best quarterback on the Consensus Big Board. Many evaluators might give the Browns a pass for staying put to select their favorite quarterback.

While the board, in my opinion, is correct in docking the Browns for selecting Ward a few picks too high when they could have landed an additional fourth just by sliding a spot or two.

In essence, the Browns gave up Josh Sweat, Nyheim Hines, Daesean Hamilton, Da’Shawn Hand or Kyzir White just to stay put.

As for the Vikings, how did they do?

Pick Player School Cons. Rk Points Spent Points Gained Value
30 Mike Hughes UCF 33 1246.3 1113.4 -132.9
62 Brian O’Neill Pittsburgh 72 871.8 784.2 -87.6
102 Jalyn Holmes Ohio State 138 615 411.9 -203.1
157 Tyler Conklin Central Michigan 185 392.5 257.8 -134.7
167 Daniel Carlson Auburn 281 360.6 70.7 -289.9
213 Colby Gossett Appalachian State 170 235.1 313.6 78.6
218 Ade Aruna Tulane 178 223.1 313.2 90.1
225 Devante Downs California 402 206.8 0.0 -206.8

The board essentially thinks the Vikings only had two good picks.

For a full list of the values of each player after accounting for team need, positional value and rank on the consensus board, click here.


Catch Arif every week on The Zone Coverage Football Machine

How Well Do You Know The Twins?

Take the Twins quiz and test your knowledge!

Take Quiz

11 COMMENTS

  1. That is some pretty charitable grading. I’d break it down by deciles of ROI:

    100 or more = A
    90 to 99 = B
    80 to 89 = C
    70 to 79 = D
    less than 70 = F

    Vikings grade: D+

    In the NFL, teams which fall into the bottom 10 usually put their GMs and head coaches at risk
    of losing their jobs, especially if they do so repeatedly. It doesn’t exactly sound like B- territory, even with grade inflation.

    • I understand, Mike. I did a quick-and-dirty school-style grading in Excel before writing the piece where I scaled the ROI values to a range of 100% to 55% and looked at what the grades would have looked like if I had turned in, say, a 75% to a teacher. Looked like a “C” to me. But your reaction makes a lot of sense because the deciles are fairly natural breaks.

  2. But the didn’t the Raiders choosing Hurst at 140 pretty much rescue their “score” given their, um, amazing reaches in earlier picks? Same with the Cardinals choosing Rosen skewing their more modest scores in the later rounds into a superior overall?

    • Yea, it did. I was going to write about that, actually – the Raiders had the highest variance in their picks of anyone else in the draft and their average overall score doesn’t go over how incredibly variant their draft is. As for the Cardinals … yes.

      • Ah, variance—the DNA that solves many a statistical crime. BTW: Your stuff is required reading at my house. Not liking the message doesn’t mean you have to hate the messenger. Keep doin what you’re doin.

  3. Great stuff Arif.
    If I understand correctly you are grading the draft solely on the player selections. What about the other component (Spielman’s favorite part) – trades? For example the Vikings started the draft with pick 92 (worth 657.1). With two decisions, they turned that into a total value of 733.8 by betting Jalyn Holmes (valued at 411.9) AND pick 180 (valued at 321.9). I realize that this makes it more difficult to evaluate individual decisions because there are a lot of interconnected ones. But to me, the total draft grade should incorporate the total of draft related decisions by adding up a team’s total value of the picks started with on Thursday and subtracting the total value of what they walked away with at the end of Saturday. (Trades for current roster players and future picks may add to the challenge a bit, but I am sure you could come up with a way to assign value to those also.)

    • Thanks, Adam. I’ve been thinking about doing that and almost did do that this year, but it is so difficult to find a reliable draft trade chart and the challenge of including player trades I think would be overwhelming. So for now, I just want to evaluate the players-to-picks and perhaps later if I’ve got time take a crack at creating a player valuation metric that takes into account talent level, contract size, remaining contract length and so on.

      • That would be a great addition for next year.
        For some reason I’ve spent an inordinate amount of time thinking about this. A few ideas (that may or may not be worthwhile):
        -For future picks you could assume the pick is in the same order as the current year and discount 32 spots (or 1 round) for each year in the future (pick #1 in 2020 is worth the same as pick 65 in 2018)
        -For trades involving a currently rostered player you could just call it a wash and assume the value of picks and players cancels out.
        -For pick trade value, AV value from Chase Stuart may more closely reflect the true value of each pick – http://www.footballperspective.com/draft-value-chart/
        Also, wondering if it would make more sense to use just your Consensus Evaluator player scores instead of also including Forecaster scores.

  4. You allude to incorporating the Need Matrix. Is that coming in a future post? I don’t see the table here. I’m a BPA-guy so I think the tables above tell the story (save for last minute injuries and character concerns that drop players), but it would be interesting to see how much the draft assessments change.

    Also, I’d use more of a grading on a curve, so something arbitrary like, 20% A’s, 30% B’s, 30% C’s, 12% D’s, 8% F’s and thus give the Vikings a C/C-. I also wonder how much Green Bay rises when you add in the 2019 first round and same for any other teams that added 2019 picks and how much Seattle falls (and perhaps other teams) when making bad trades like only receiving a 3rd and a 6th to move down 10 spots in the 1st round from their original pick.

    • Hi Shalesh, sorry I didn’t see this earlier. The needs matrix just comes from Drafttek.com – http://drafttek.com/teamneeds2018.asp and I didn’t do much other than multiply the value of a pick by anywhere between 0.7 to 1.4 times the player’s original value. If you’re a BPA guy I understand that it’s quite a bit but this worked for me because it better matches how teams actually seem to approach the draft.

LEAVE A REPLY

Please enter your comment!
Please enter your name here