Drafting Against Consensus: The 49ers Are Better at Drafting Players for their Team than the Media Pundits (2024)

The title of this article might strike some readers as odd. Surprisingly enough, it has been a topic of conversation on Hogs Haven. Let me explain.

Something happened in January this year that generated great confusion amongst the Hogs Haven readership. After 24 years of dysfunction and mismanagement, the new ownership of the Washington Commanders finally hired a real General Manager and empowered him to do his job, without interference.

For many, this development was the answer to a prayer. Others did not take it so well. In his first draft with the Commanders, the recently hired Adam Peters did things that were unfamiliar to many in the fanbase. While the Commanders’ draft selections met with high praise around the NFL, some within Hogs Haven were not happy.

Peters upset a small but vocal faction of the Hogs Haven commentariat by not drafting two starting OTs to replace the makeshift bookends left over from the previous regime. Generally speaking, one needs to pick around the top 15 to 20 to draft an OT who projects as a Day 1 starter. Peters had the number 2 pick, but it was earmarked for a QB. That left him to find two Day 1 starting OTs in the second round. He attempted to trade back into the first round to draft Washington OT Troy Fautanu, but the bid was unsuccessful.

In the end, he only managed to draft one OT in the third round. Clearly, if he had really valued the position he would have waved his magic wand and conjured two starting level OTs in the second round.

Peters’ critics have been coming up with increasingly creative arguments to raise doubts about his suitability as GM, only four months after his arrival. For example, we have heard that he wasn’t really involved in draft decisions, despite 49ers GM John Lynch crediting him for being integral in establishing the 49ers’ draft process. Then there are the ingenuous attempts to paint anyone who doesn’t agree with their criticisms as support for inferior offensive line play: “If you think Adam Peters did enough to address OT in the draft, then you must be happy for Wylie and Lucas to start in 2024”.

The Horror of Drafting Against Consensus

Photo by Michael Zagaris/San Francisco 49ers/Getty Images

The most contrived argument to be thrown at Peters to date is that he has a habit of reaching in the draft. We are told that in his first draft as the Commanders’ GM, Peters reached for 6 of his 9 draft picks. Furthermore, we are told that this appears to be a habit he learned during his time with the 49ers, who routinely reach in the draft with predictably bad results.

There are two very odd things about this argument. Since this trip through Wonderland included a thinly veiled personal attack on yours truly, I felt a need to set the record straight.

First off, we are led to believe that Adam Peters learned his bad drafting habits in San Francisco, where he was John Lynch’s second-in-command from 2017 through 2023. The strange thing about this line of attack is that, during that period, the 49ers were unquestionably one of the best drafting teams in the NFL, and arguably the best.

During Peters’ tenure, the 49ers drafted 5 first-team All Pro’s, tying with the Ravens for the most in the league. What’s more, they selected seven elite players and quality starters on Day 3 (TE George Kittle rd 5, DT DJ Jones rd 6, LB Dre Greenlaw rd 5, CB Deommodore Lenoir rd 5, S Talanoa Hufanga, G Spencer Buford rd 4, QB Brock Purdy – last pick in the draft). Few teams came even close to that record of success. Whatever habits Peters learned in San Francisco, we should count ourselves lucky if he brought them to Washington.

If that line of reasoning were not odd enough, things take an even stranger turn when we examine what the alleged “reaching” actually is.

The term “reach” is normally used to mean that a team intentionally picked a lower rated player on their board, bypassing superior talents. The main reasons that teams reach is out of desperation to fill a need, or because the decision makers and the scouting team are not on the same page. An example of the latter was Dan Snyder drafting QB Dwayne Haskins 15^th overall in 2019, despite his scouts having a third round grade on him.

However, that is not what Adam Peters and the 49ers are accused of. The allegations of draft malpractice have to do with Drafting Against Consensus. Hogs’ Haven’s Bobby Gould recently brought to our attention that the 49ers are the team most likely to pick players ahead of their ranking in the Consensus Draft Board.

The Consensus Draft Board was originated by Arif Hasan, writing for a Minnesota Vikings fan site. The original Consensus Board was constructed by averaging rankings in 40 draft boards published by media draft experts like Mel Kiper, Todd McShay and Lance Zierlein. It gained steam, and was adopted by The Athletic in 2019, before going solo again in 2024. It has spawned many imitators. But when people cite the Consensus Board, it is generally what they mean.

The Consensus Draft Board is a fine piece of work, and has many uses. But it is no substitute for an NFL team’s own board. There are many reasons why it is a dubious guide for evaluating an NFL team’s draft decisions. Here are my top four:

1. The Consensus Board combines rankings from two kinds of experts. Evaluators, like Lance Zierlein, rank players in their boards based on how good an NFL player they are expected to be. Whereas, Forecasters, like Walterfootball, rank them by where they are expected to be picked. There is often wide disagreement between the two types of rankings, introducing huge amounts of jitter into any individual player’s consensus rank. Hasan’s version lists both rankings separately, as well as the composite rank, which is usually what people call the Consensus Rank.
2. No matter how anyone does their rankings, premium positions like QB and OT tend to get picked earlier in the first round than their rankings. For example, in this year’s edition, Jayden Daniels had a Consensus Rank of 8.
3. A few of the experts whose rankings contribute to the Consensus Board have actual NFL scouting experience. Most do not. Sports journalists and website hosts are no match for the experts working for the best NFL teams.
4. NFL teams painstakingly construct their draft boards to suit their team’s unique set of schemes and preferences. There are bound to be huge discrepancies between different teams’ boards, including exclusion of players for non-fit. The Consensus Board is a one-size-fits-all ranking. It is not made with any particular team in mind and will be full of players not listed on an individual team’s board.

As preposterous as it might seem to anyone who is familiar with the San Francisco 49ers’ recent draft history, this is where we have got to. One of the best drafting teams in the league is being held up as an example of poor drafting practice; and I have been labelled as a “fawning admirer” who “aggressively challenges any questioning of or negative comment about the GM” for my attempts to challenge those making blatantly unfair criticisms and propagating outright misinformation, as I am once again about to do.

Clearing the Record on Adam Peters’ Alleged Reaches Against Consensus

Before I get to the 49ers’ draft operations’ good name, I have to start with the object of my alleged fawning admiration. Here are the Commanders’ 2024 draft picks graded as reaches and steals according to the generally accepted Consensus Draft Board:

Using a 25% deviation from Consensus Rank as the criterion for a reach or steal, I only count four reaches and one steal.

The Hogs Haven author making dubious claims about Adam Peters’ draft technique used the Mock Draft Database Consensus Big Board to grade his picks. That board is primarily based on users’ mock drafts (211 internet big boards vs 1,507 first round mock drafts + 1371 team based mock drafts) and updates whenever someone does a new one.

At the risk of being accused of aggressively challenging anyone who dares to question the GM, I have to say that random people on the internet doing mock drafts is a particularly odd choice to hold up as the source of truth for grading an NFL team’s draft selections.

The 49ers’ Had a Better Feel for Prospects’ Draft Position than the Pundits

Photo by Michael Owens/Getty Images

Note: Definitions of capitalized terms and additional methodological details can be found in the Tools and Methods section at the end of the article.

Adam Peters was hired by John Lynch for his scouting expertise in a rebuild of the 2-14 San Francisco 49ers in 2017. In the third season of the rebuild, the 49ers were back in the Super Bowl.

During Peters’ tenure, the 49ers made 58 draft picks. Twenty-six of those (44.8%) qualified as Reaches relative to the media consensus rankings, using a 25% Reach threshold. As far as the media pundits are concerned, the 49ers were the reachiest team in the NFL draft. At first glance, that would seem pretty difficult to reconcile with his claims to drafting the best players available. But perhaps, just maybe, the 49ers’ board was different from Mel Kiper’s.

Call me a fawning sycophant if you like, but I don’t see how it matters if the 49ers draft picks differed from the Consensus Board. What matters is, when there was disagreement, who was right?

To answer that question, I graded all of the 49ers’ draft picks from 2017 through 2023 as Reaches, Steals, or on-consensus picks and compared the outcomes of their Reach picks to those of the other players picked in the same Draft Range.

My analysis corrects one crucial flaw, which is likely to have influenced the results of Bobby’s recent look at the 49ers’ Reaches against consensus. Bobby used wAV (defined in Methods) to measure how well the draft picks turned out. Distributions of wAV data at any point in the draft tend to be skewed, with a relatively small number of high values distributed amongst a larger number of lower values. As the draft progresses, the distribution of wAV values becomes more and more “sparse”, with increasingly rare high values surrounded by growing numbers of 0s and 1s.

Bobby compared the wAV values of the 49ers’ reaches to the mean (average) of the next 16 players drafted, and concluded that most of their Reach picks were below average for the draft range. Sparseness creates particular issues with the use of the mean for this type of comparison. In a truly sparse data set, such as wAV values in the 7^th round, the mean becomes dominated by a few data points with high values. Most points in the distribution fall below the mean. As a result, unless the 49ers hit on a rare late round gem, their reaches on Day 3 would be expected to be graded as below average, along with most of the other picks in the round. This issue becomes less of a problem earlier in the draft, but is likely to have deflated the grading of the 49ers’ reach picks across the board.

To cope with this issue, I used what statisticians call a non-parametric approach. Player outcomes were measured using wAV Percentile, as detailed in the Tools and Methods section. Briefly, the wAV Percentile is the percentage of players in the Draft Range with a wAV value equal to or lower than that of the Reach pick in questions. The wAV percentile tells exactly where a player sits within the comparison cohort, regardless of the shape of the distribution, and is unaffected by outliers. The 50% wAV Percentile is also known as the median of the distribution.

I used wAV Percentile to measure how the 49ers’ Reach picks turned out in the NFL, compared to the choices of other teams’ GMs in the same Draft Range. A few of the 49ers’ own picks were included in a few of the comparisons, but not enough to matter.

The 49ers’ 28 Reaches during Adam Peters’ tenure are listed in the following table, color coded according to how they turned out:

Reach picks are color coded as follows:

Blue – Very good to exceptional value, wAV Percentile 80% to 100%

Red – Above average value, wAV Percentile 60% to 79.9%

Gray – Around average value, wAV Percentile 40% to 59.9%

Green – Dud pick, wAV Percentile < 40%

Players who did not appear in the top 300 Consensus Board were assigned Consensus Ranks of 301.

Blue and Red Picks – Beating the Market

Ten of the 49ers’ Reach picks, accounting for 38.5% of the total, were classified as either Red or Blue picks. These picks represented significantly better value than most teams achieved in the same draft range.

These picks didn’t all result in drafting All Pros and premium starters, because most of them (7/10) were made on Day 3, where simply finding a long-term starter or major contributor is beating the market. Two of the Reach picks landed premium talents, who have been key pieces for the Niners: WR Brandon Aiyuk, 2019 1^st round (2^nd team All Pro, 3,931 receiving yds, 25 TDs in 4 seasons), LB Dre Greenlaw (4 year starter).

Greenlaw was the best value relative to Draft Range among the Reach picks. Selected with the 148^th pick in the fifth round of the 2019 draft, Greenlaw became the 49ers’ starting middle linebacker as a rookie, and has held the job since, only missing significant time to injury in 2021. Greenlaw has been highly productive and grades as a high-end starter.

Greenlaw is one of only five players drafted by any team in the fifth round that season, excluding specialists, to stick in the league as a quality starter or major contributor, along with LB Andrew van Ginkel (wAV 24), WR Hunter Renfrow (wAV 21), WR Darius Slayton (wAV 25) and Redskins’ pick LB Cole Holcomb (wAV 25). Eleven out of 35 players (31%) drafted in the fifth round have wAV of 0 or 1.

Not all Blue and Red Reach picks were players that the average fan of another team will necessarily recognize. Many of them are players picked on Day 3, and are a mixture of backups, special teamers and recently drafted depth players who are working their way up a deep depth chart. Simply having earned playing time and logging recordable stats puts them ahead of most players picked in the same rounds.

Gray Picks – Fair Market Value

Seven out of 26 Reach picks (26.9%) have turned out to be around median value for their Draft Range. There are no real famous names in this group. The most prominent of the fair value Reaches is OT Mike McGlinchey, whom a few Hogs Haven regulars thought would be a good addition for the Commanders last offseason.

MicGlinchey might not have lived up to the 49ers’ expectations, when they drafted him 9^th overall in 2017. He started his rookie season at RG and took over starting RT duties from 2019 through 2022. He signed a 5 year, $87.5 million UFA contract with the Broncos in 2023 and started 16 games for them last season. Based on wAV, he rates as good or better than five of the players picked in the top 12 of his draft class (DT Vita Vea wAV 36, CB Denzel Ward wAV 36, DE Bradley Chubb 35, QB Sam Darnold wAV 25, QB Josh Rosen wAV 3).

Adding the seven Gray Picks to the 10 Blues and Reds brings the number of the 49ers’ Reach Picks that achieved average value or better for their Draft Range to 17.

Green Picks – Duds

Nine out of 26 Reach picks (34.6%) have turned out to be below average for their Draft Ranges.

One of the two biggest busts during Adam Peters’ tenure with the 49ers was QB Trey Lance, who was acquired in the second biggest draft trade for a QB since the RG3 debacle. Lance was picked 9 places ahead of his Consensus Rank, which counts as 75% Reach in this scoring system. It is an equivalent level Reach to Adam Peters’ pick of Jayden Daniels in 2024.

Lance struggled in two starts during his rookie season, and lost his second season to an ankle injury. By the time he recovered, he had lost the starting job to 2022 seventh round pick, Brock Purdy, and was traded to the Dallas Cowboys, where he is currently third on the depth chart behind Dak Prescott. After having only attempted 102 passes in 8 games across 3 seasons, it might be early to pronounce final judgement on his career.

Other notable Reach picks that did not turn out well, are WR Dante Pettis (Round 2, 2018) and WR Jalen Hurd (Round 3, 2018). The others were either picked too recently or too late in the draft to qualify as genuine busts.

49ers’ Reaches - Overall Summary

Overall, out of 26 Reaches relative to consensus, 17 (65.4%) were in the 40^th wAV percentile of their Draft Ranges or better, indicating that the 49ers got around average value or better compared to other teams picking in the same range. Only 34.6% of the 49ers’ Reach picks were significantly worse than the median of players picked in the same draft range. This means that when the 49ers and the media experts disagreed on where a player should be picked in the draft, the 49ers were close to twice as likely as the media to be right.

Early Round Reaches

While the 49ers did well overall when they Reached relative to media consensus, what might be of greater importance is how they did when they Reached in the early rounds, as Adam Peters did in his first draft with the Commanders.

Out of 12 Reaches in Rounds 1 through 3, four were above average to exceptional value picks, grading in the top 67^th percentile or higher in their Draft Range. These include 2^nd team All-Pro WR Brandon Aiyuk, G Aaron Banks who became the full time starter in his second season, and 2023 draft pick K Jake Moody. Two more were fair value picks, bringing the total number or fair or better value Reach picks on Days 1-3 to 6/12 (50%). Two of the Dud picks were drafted in the past two seasons and may improve in value with more playing time.

In summary, in Rounds 1 through 3, when the 49ers and media experts disagreed on prospect projections, they were equally likely to be correct. While the 49ers weren’t exceptional at finding value by Reaching in the early rounds, as they were in the later rounds, there is no evidence that Reaching relative to consensus was associated with unusually poor outcomes.

Conclusions

Photo by Thearon W. Henderson/Getty Images

Adam Peters was hired a little over four months ago to lead a large scale rebuild of a Commanders’ roster, which had been left in a devastated state by the previous front office. He was recruited from one of the highest performing personnel teams in the NFL. His former GM, John Lynch, credits him with having been “integral” to building a draft operation which, in the seven years that he was with them, drafted five first team All Pros, second team All Pro Brandon Aiyuk, and Pro Bowl QB Brock Purdy.

The 49ers’ exceptional knack for drafting elite talents during Adam Peters’ tenure was matched by an equally impressive record of finding quality starters and major contributors in the late rounds of the draft.

In his first draft in Washington, four of the players selected were picked well ahead of their media Consensus Ranks. In proportional terms, his pick of Jayden Daniels 2^nd overall was a massive reach relative to media consensus. The three draft picks to close out Day 2 were also picked more than 31% ahead of their consensus ranks.

That might be a cause for concern, if not for the fact that it replicates a pattern of drafting behavior exhibited by one of the most highly successful drafting teams, where Peters’ developed his craft.

Washington’s roster would get a big boost by the addition of players like Dre Greenlaw, and Brandon Aiyuk who were big Reaches relative to media consensus. The depth ranks could also use an influx of talent from players like OT Colton McKivitz and G Aaron Banks.

Interestingly enough, many of the very same media experts whose consensus prospect rankings are being held up as a cause for concern have given Adam Peters’ first draft in Washington glowing grades.

In essence, what we are really being asked to worry about here is an elaborate version of instant draft grades. Everyone should know that it takes between one and three NFL seasons to reach any kind of meaningful conclusion about a draft class. Instant draft grades are fun, but they are purely for entertainment purposes, and are ultimately meaningless.

The suggestion that Adam Peters’ first draft in Washington is flawed because he reached for four players ahead of where Mel Kiper and company thought they should go is premature and is based on arguments with holes in them big enough to drive trucks through. If he achieves the same success on Reach picks in Washington that his team did in San Francisco, our future is looking bright.

My advice to Commanders fans is:

Don’t believe the hype

Apologies to anyone who was offended by my obsequious fawning.

Tools and Methods

Consensus Draft Board (Consensus Board) - the 300 prospect media consensus draft board developed and curated by Arif Hasan. Draft prospects are ranked according to the average of rankings from 101 “draft experts’” big boards. The Consensus Board includes separate rankings from Evaluators and Forecasters, as well as composite rankings generated by combining the two.

Consensus Rank - a prospect’s composite rank in the Consensus Board

wAV – is a variant of Pro Football Reference’s Approximate Value (AV) metric. AV is an attempt to put a single number on the seasonal value of a player at any position. wAV is a weighted sum of the player’s career total AV values, emphasizing peak seasons. wAV is a cumulative stat and therefore cannot be used to compare players in different draft years without some form of correction.

Draft Ranges – the career value of players at each of the 49ers’ Reach picks was determined by comparison with other players picked in their same Draft Range using the wAV Percentile metric, described below. The wAV values of players picked in the draft tends to decrease logarithmically as a function of pick number. However, there is so much variance between wAV numbers of players at nearby draft picks, that it is difficult to see reliable differences between players taken many picks apart. I have previously estimated the limit of draft resolution to be about a full round at the end of Round 1.

I took advantage of the resolution limit to generate large sample sizes of players for comparison with the 49ers’ draft picks. From the second round onward, the Draft Range used for comparison with the 49ers pick was the full round in which they were picked. The inflection point at which the wAV curve changes from steep to shallow occurs in the range from pick 12 to 16, making the early part of the first round fairly special. Therefore, the first round was split into two Draft Ranges: picks 1 to 12 and picks 13 to 32.

wAV Percentile – The percentage of players in the Draft Range with the same or lower wAV than a 49ers’ Reach pick. The use of wAV Percentile provides a precise indication of where each Reach pick sits within the Draft Range, and avoids grade deflation associated with use of parametric measures on skewed datasets.

Reach – A player picked 25% or more ahead of his Consensus Rank.

Steal – A player picked more than 25% later than his Consensus Rank.

Reach/Steal Coefficient – the Reach/Steal Coefficient quantifies how much of a Reach or a Steal each pick was. It is the difference between the pick number at which the player was taken and their Consensus Rank as a percentage of the Consensus rank. Negative values indicate Reaches, and positive values indicate Steals. A Reach/Steal Coefficient value between -25% and +25% indicates that the player was picked in the range where he was expected.

The Consensus Boards list the top 300 ranked players in each draft. Draft picks who were not listed on the Consensus Boards were treated as having Consensus Ranks of 301. This means that the Reach Coefficients for unranked players are, if anything, conservative underestimates of their reachiness. Unranked players can’t be Steals by the definition used in this article. In fact, some of the 49ers best draft picks during Peters’ tenure were unranked players who turned out to be massive steals by the more usual definition.