Credits to Ben Baldwin and Timo Riske as I heavily reference work from both of them. They are among the absolute best in advancing thinking on player value and draft analytics. If you don’t follow them, you should be.
With the Combine complete and only pro days left, we are entering the home stretch ahead of the draft. And with that, the consensus 1st rounders won’t move around much more. There won’t be more Devon Witherspoon’s, Lukas Van Ness’s, and Drew Sanders moving up from the depths of the big boards into the 1st round.
What do we know about prospect rankings and improving success in the draft?
Big boards are as accurate as team evaluations
PFF’s Timo Riske did great work showing that the consensus big boards we all see are as accurate as NFL teams on who the top players are. Here is the full article:
“Can the Consensus Big Board really predict the 2021 NFL Draft?” – PFF.com
Couple of points from this research:
- For non-QBs, there is little difference between team rankings and consensus big boards through the top 3 rounds except for QBs
- As you get into day 3, big boards are much less accurate, probably for a couple of reasons including teams being better at identifying talent and player opportunity.
Reaching is usually a mistake
(…and the inverse, draft “steals” are rarely actually steals)
In related work, Timo Riske again analyzed players drafted at picks that differed greatly from their consensus ranks and found that players that went much higher than their consensus rank usually underperformed their draft location. Likewise, players drafted much later than their consensus performed closer to what is expected where they were drafted and were not “steals”.
Of course, people will point to the outliers that worked out like Tyreek Hill and Stefon Diggs as “steals” and Darius Leonard as a “reach”, but for every one of these, there are literally dozens and dozens of counter examples.
This does not mean It speaks to Sheil Kapadia’s great draft commandment “Don’t be overconfident in your ability to evaluate talent.”
An example of the Seattle Seahawks top-rated 2022 draft and the value of not reaching…
The Athletic has a good article pointing to this with Seattle’s 2022 draft, the top draft in 2022 in terms of actual player value and value over expected and the 6th best draft since 2010.
“How an apparent shift in strategy led to the Seahawks’ best draft in years” – The Athletic
Whether an intentional change in strategy or not, in 2022 Seattle departed from their recent history of reaching on players.
Each of their top four picks were within a couple slots of pre-draft consensus rankings and they generated 30% more value than the next best team, triple the value vs. expected, and had 3 of their 4 day one and day two picks hit. Add in Tariq Woolen in the 5th as one of the best value picks in several years and you see how good this class is.
First round miss rates are still high
Since 2010, an average of over 12 of the 1st rounders aren’t top 50 players. Even the very best years – 2018 and 2019 – had nine 1st rounders that missed.
Having a top pick doesn’t guarantee success – all but one year had at least one top-10 pick outside the top-50 players. Drafts average over two misses in the top 10 picks. The 2020 draft had Jeff Okudah and CJ Henderson. 2019 had Clelin Ferrell. And, of course, teams taking their shots at their QBs have missed plenty – Josh Rosen, Zach Wilson, and Sam Darnold.
Tackles are less risky than the other premium positions
Historically, and unsurprisingly, the skill positions have the highest miss rates. Below shows how many round 1 misses out of total round 1 picks by position:
Couple of points:
- Tight end (at 83% miss rate) and safety (63%) actually have the highest miss rates but neither is drafted often in round 1, both averaging fewer than 1 per year
- Of the heavily drafted positions, cornerback has the highest miss rate with half of round 1 picks missing between 2010-2020
- Wide receiver, EDGE, quarterback, and linebacker are next highest with miss rates between 36-41% of round 1 picks
- Tackles are historically the safest position to take
The most valuable picks are from the mid-1st to mid-2nd rounds
Last year, some amazing work from Timo Riske and Ben Baldwin aggregated a view on player value and financial cost that had some really interesting takeaways on which draft locations are most valuable and positional value in the draft.
“The surplus value of each position in the NFL draft” – PFF.com
First to explain their approach:
- Since the rookie wage scale was introduced, every pick has a set salary, regardless of position (orange line below)
- Player performance (blue line) is higher earlier in the draft, but it decreases at a slower rate than the rookie wage scale, especially in the first 10 picks
- The difference between a player’s value and their cost is the draft “surplus value” (green line)
Because the rookie wage scale drops faster than player value, and because it is still hard to pick good players, the value of a pick – the surplus value – is not highest at the top of the draft. Picks from the teens through the middle of the 2nd round are more valuable than picks 1-5.
An example of what surplus value says about drafting running backs
The above shows average value across all non-QB positions. But positions that are more expensive on second contracts (pass rushers, quarterbacks, receivers, and to a bit lesser degree, cornerback) have greater surplus value. Their value-to-market cost is just much higher and this is one reason why people say don’t take a RB high.
To give an example, pick 10 (Garrett Wilson last year) makes an average of $5M a year on their rookie contract. Pick 20 is around $3.5M a year. If you take a QB or a receiver or a running back here, it doesn’t matter – they have the same rookie contract. But the top free agent RBs cost $6-7M a year and many are around the same cost as the rookie salary.
Compare that to premium positions which all approach or exceed $20M a year on 2nd contracts and you can see why it is just so much more valuable to replace a premium position with a rookie cap hit.
There’s a great line in the above link that the New York Giants would have been better off forfeiting their pick instead of taking Saquon Barkley at 2 because they committed more cap ($9M a year in today’s dollars) than the position was worth.
The case for trading down
We all love our players – I do it every year and am doing it again this year. If I’m Howie and Christian Gonzalez is there at 10, I turn my phone off and run the card up. You may want Tyree Wilson or Devon Witherspoon or Joey Porter Jr. there instead.
But what if, as we are waiting for Roger Goodell to announce our guy, we hear “the Philadelphia Eagles trade pick 10 to the…” and the Eagles move way down in the 1st? It would be a giant let-down but it could be the best move.
The Vikings 2022 trade down
We have an interesting real-life example from one of the newest, and most analytical, GMs last year. In 2022, the Vikings’ new GM Kwesi Adofo-Mensah surprised everybody by trading picks 12 and 46 to the Lions for 32, 34, and 66. It seems crazy, but look at the picks made (some of these picks were later traded again, I am just pulling the actual picks made):
| 1-12: WR Jameson Williams 2-46: DE Josh Paschal | 1-32: S Lewis Cine 2-34: WR Christian Watson 3-66; LB Brian Asamoah |
Jameson was well-regarded, but because of his injury, we didn’t get to see much of him in 2022. Unfortunately, the same with Lewis Cine who tore his ACL before the season. But would you take Cine and Watson over Jameson?
Time will tell but the Vikings raised their chance of success. Pick 12 has an expected value of 5.49 AV per year while picks 32 and 34 total 8.85 expected AV per year of value (32 is 4.47 AV and pick 34 is 4.38).
The Steelers 2019 trade up
Another example with not as big of a trade down was in 2019 when Pittsburgh traded up with Denver to take LB Devin Bush:
| 1-10: LB Devin Bush | 1-20: TE Noah Fant 2-52: TE Drew Sample 2020 3rd round pick |
The Steelers love their LBs and while Devin Bush is very good, he is exactly equal to the value you expect at pick 10 (5.97 AV per year). But Fant, an above average TE, and Sample combine to equal the value of Bush, not counting the future 3rd.
What this all means and options for the Eagles
All of the above does not say don’t use a high draft pick. Looking back at the Eagles 2021 trade up and pick of DeVonta Smith, it’s hard to second-guess it. Sure they could have stayed put and taken Micah Parsons in retrospect, but if they were set on getting a WR they would have walked away with Kadarius Toney or Rashod Bateman.
The way to think about it is, over time, there are things that raise a team’s chance of success. To recap:
- As we get closer to the draft, consensus big boards will be pretty accurate on actual draft locations.
- This means “accurate” on which players were taken, NOT who the best players are – around 40% of the 1st round is wrong every year
- The relative cost of players vs. their performance, and factoring in miss rates, actually points to mid 1st round through mis 2nd round as the highest value draft locations
If a team had a choice of a single top 10 pick or multiple picks in this best value range, over time they would be better off with multiple picks. And history shows this.
For an example with our Eagles, this year’s pick 10 is roughly worth picks 20 and 51 in a trade. If you look back at every draft since 2000, only five times has pick 10 outperformed the cumulative value of the picks at 20 and 51 (one of these outperformance years was the Eagles selection of DeVonta Smith):
Both pick 10 and the combination of picks 20 and 51 delivered at or above expected value at those positions nearly half of the time, which is expected. Pick 10’s variance in value is a bit higher (3.3 AV per year vs. 2.8 for picks 20+51) which is simply because two picks “diversify” some.
2023 Eagles draft options
This trade may not be possible – we are assuming somebody like Seattle or the Chargers would be willing to move up and give their top two picks for somebody they love. But running a mock simulator just to show the options available on staying at pick 10 vs. trading down for picks 20 and 51.
Here are names available at picks 10, 20, 30, and 51:
Pick 10: Popular picks available include Devon Witherspoon, Joey Porter Jr., Paris Johnson, or Lukas Van Ness
Pick 20: Calijah Kancey, Nolan Smith, Cam Smith
Pick 30: Emmanuel Forbes, Anton Harrison, Darnell Wright, Josh Downs, Felix Anudike-Uzomah
Pick 51: Derick Hall, Tuli Tuipulotu, Keeanu Benton, Matthew Bergeron, Garrett Williams, DJ Turner
Thanks for reading and if you aren’t already, please follow us at @PhillyCvrCorner, @PhillyEagleNews, and @GregHartPA