In the days following the NFL draft, a controversy over the value of NFL draft consensus boards resurfaced with more energy. Are they good? Are they bad? How are they compiled? Are they accurate? Are they useful? Are reaches in the draft really bad? Are steals really good? Is there really wisdom in this groupthink?
Let’s take a closer look at these questions and the conundrum teams face when evaluating their own boards and draft decisions against the consensus board.
How Are Consensus Boards Compiled?
The first thing if you’re not familiar
with consensus boards is there isn’t just one. There are a handful. Each is compiled in similar ways but with some significant differences. All are essentially composites of anywhere from a few dozen or so to over a hundred NFL draft big boards and/or mock drafts put together by (usually) well known and/or competent sources, from Mel Kiper and Dane Brugler to less well-known sources that may also have more of a focus on particular teams or some other relevant specialization. But beyond that the parameters are different among consensus big boards. The DN’s own Arif Hasan has been doing consensus boards for many years now at different outlets and is at the forefront in their development. He did a piece here that explains some of the differences and parameters that can go into a consensus board. Different methods of averaging rankings, different weightings, different big board or mock draft sources, and how far back are big boards or mock drafts included in the weightings are a few ways they are different. And there are differences in how many players are ranked.
But comparing three – The Athletic’s, Arif’s and the NFL MockDraftDatabase consensus big board- produces substantially different results in just the top 30 rankings. Here are their 2023 consensus big boards:
As you can see, there are substantially different rankings on about half of the prospects. The problem with the different rankings arrives, for example, when the Chargers pick Quentin Johnston with the 21st pick. Depending on which consensus board you choose, that pick is either a major steal, a major reach, or about fair value. Which is right? Is one consensus board more legit than the others? The short answer is no.
Are They Accurate?
The reason there are more than one consensus board, apart from different outlets competing for traffic, is that they are compiled with different goals in mind. Some are better than others at various goals such as getting closer to actual draft results or ranking how good the players actually turn out to be based on some performance metric(s) over a period of time. Some blend the two goals. But in terms of predicting at what pick number a prospect will be selected, there is quite a bit of variance, as even the average first-round pick variance between a consensus board and actual result is +/- nine draft picks and the average pick variance over the entire draft is over 50 picks, more or less, from the period from 2016-2024 according to Arif Hasan’s analysis.
And looking at those three top 30 consensus boards after three seasons, none of them have done a tremendous job predicting the 30 best players in this draft so far, let alone the top 250, which is typical of consensus boards from other years too. Of course a big reason for that inaccuracy is because there are various other factors that impact player performance beyond their consensus board ranking, not least of which is how well they are coached and developed after they’re drafted, how they are utilized, how durable they prove to be, among other things that are impossible to predict with any reasonable degree of accuracy.
Another reason for the inaccuracy of consensus boards is that, for the most part, they lack inside information. There may be some big board creators that have some limited access to inside information from one or more teams, but that tends to get diluted among the dozens of others who do not. There is constant chatter between team sources over the course of the pre-draft process, but teams also guard both their intentions and the intelligence they collect about players and other teams, so much of that never makes it to the public and consensus boards. Indeed, it may be widely known among teams that certain prospects are ranked significantly higher or lower than they see on consensus boards, based on non-public information. Sometimes, but not always, that may come out in the days just before the draft. Either way, it isn’t reflected on consensus boards. Teams can also put out misinformation to smokescreen their intentions, which can also find its way into consensus board rankings.
Overall, consensus boards are not nearly accurate enough for a team to base any draft tactics on, or a gambler to place a wager, or a fan or pundit to cast judgment, but consensus boards may still bring some value as more general background data in some cases.
The Changing Consensus
It’s also important to note how consensus boards change over time. You’d think that rankings based strictly on traditional scouting evaluations would be relatively static by, say, mid-March or the beginning of April as by that time there have been months for film evaluations which make up most of the evaluation and the Combine results have also been available for a couple weeks as well. There’s also been plenty of time to distill whatever public information is available about a prospect’s character and so forth as well. Perhaps some minor adjustments for pro day results that might move the needle, or perhaps more significant movement in a select few prospects based on newly available information on injury or off-field issues, but for the most part by the beginning of April most of the traditional data big board producers use has been out there for quite a while.
But consensus board rankings based on traditional scouting evaluations do change over the month leading up to the draft, despite 95%+ of primary data having been out there for weeks and months. And sometimes quite a bit. The reason most often is that big board creators get more data on what teams are thinking about prospects and that influences their rankings, and in turn the consensus boards as well.
For example, Treydan Stukes went from around pick 100 on consensus boards in late March to around pick 60 just before the draft. He ended up being drafted #38 by the Raiders. There was no new primary information about Stukes, on or off the field during his rise on draft boards. But there was secondary information suggesting teams may be rating him higher. There are plenty of such examples, particularly in the middle and late rounds.
Another situation that can lead to significant variance between actual draft position and the consensus board is last minute information that isn’t reflected in consensus board rankings. For example, this year news that Jermod McCoy may have a degenerative knee condition and may require additional surgery on his knee came just a week before the draft. That meant that teams may have taken him off their boards completely or docked him a few rounds or more to account for the risk. And yet McCoy dropped just a few spots on consensus boards after that news was made public. The reason is that consensus boards include big boards and mock drafts going back a couple months and so more dramatic changes in recent big boards or mock drafts are muted.
Are They Useful?
Despite not being tremendously accurate and with some known shortcomings, consensus boards can provide some value. For teams trying to determine how “the market” views prospects they’re interested in, consensus boards can fill in gaps in their own information to some degree. Consensus boards can also serve as a check on team’s prospect assessments. If a team has a substantially different ranking on a prospect than the consensus board, that may warrant a second look and discussion on whether the team’s assessment is accurate or not.
Cincinnati Bengals’ GM Mike Potts talked how the Bengals used consensus boards in their process along these lines- as a comparative tool to double check their board against. And whether a GM acknowledges it or not, every team has somebody that performs this function. What a team does in response to the comparison is another thing, but every team has one or more people that will compare the team’s board to the consensus board and point out notable differences. In some or most cases those differences exist for a reason that the team is comfortable with, but in a few it may prompt a review of the player evaluation.
The Baltimore Ravens’ GM Eric DeCosta discussed another phenomenon related to consensus boards. He said that even as there has been a convergence in team’s boards around the consensus board, which he attributes to analytics (most likely teams using similar models to rank players which filters through to consensus boards), there remained differences with what he termed outlier players. He said that in cases where the Ravens had ranked a player considerably different from the consensus board, he was finding that those players were actually selected closer to where the Ravens had them ranked rather than where the consensus boards had them ranked.
That phenomenon reduces the utility of consensus boards as a roadmap to navigate the draft to the extent there are outlier players. The more outlier players there are, or the more interested a team is in outlier players, the less useful consensus boards are to them in navigating the draft.
49ers head coach Kyle Shanahan talked more about the realities of these issues, both in terms of draft tactics and outside perceptions based on consensus board rankings:
This gets to the heart of the debate about consensus boards as a means of judging team’s draft picks.
Judging by Consensus Boards
Immediately following the draft come the knee-jerk reactions and grades. A lot of that is based on where the prospect was selected relative to their consensus board ranking. Whether the pick fulfills a perceived team need and fit are also factors. Relatively to the consensus board rankings, reaches are bad and steals are good and that tends to be the main focus. The Vikings’ selection of Caleb Banks went over like a lead balloon at draft parties and with pundits largely because he was seen as a big reach at #18. Differing perceptions of injury risk was the main reason for the difference between the Vikings’ ranking of Banks and the consensus board, but the widely held view by consensus boards that the Vikings would draft Dillon Thieneman as a perfect replacement for Harrison Smith contributed to that reaction.
We don’t know if any other teams had the same view of Banks as the Vikings, but there remains speculation that other teams picking later in the first round may have selected Banks had he been available. A trade down by the Vikings when Thieneman was available at #18 may also have been a signal to other teams that the Vikings, who had a big need at defensive tackle, were going to draft a defensive tackle and prompted them to trade ahead of the Vikings to get Banks. So, while Banks was ranked #37 on consensus boards, insider information may have suggested that one or more teams picking in the late first round may be targeting him, so is it worthwhile to trade down and hope he’s still available or is that trying to be too cute as Kyle Shanahan termed it?
But the Banks selection, along with many others including a couple of the 49ers picks that Shanahan discussed, gets to the issue of team’s big boards versus consensus boards.
Critics of consensus boards (or judging by consensus board) point out that every team will deviate to some extent in their player evaluations based on their perception of how a player may fit with the team across many factors, as NFL dot com draft analyst Bucky Brooks points out:
FYI: It’s unfair to judge a team’s draft class based on Consensus Boards. Without a thorough understanding of the team’s plans for the team, players, or scheme, it’s impossible to suggest a pick is good or bad. Teams build their draft boards based on grades that project how a player will perform or contribute within their system, the Consensus Boards and media scouting reports lack that context. Without a clear understanding of the “why” behind each pick, critics are missing some key tidbits that are critical to a full assessment. In the NFL, it takes 2-3 years to determine whether players can play or not. That’s why we should reserve judgment until we see how the players perform and how teams utilize their personnel. That’s not fun, but none of us are privy to the 32 draft boards or the schematic or personnel discussions prior to the draft. We can debate who was selected and where they were picked, but until we understand the why and see how it plays out between the lines, we are still playing the guessing game.
Those fits and assessments can vary. For example, a 4-3 defensive team looking for a base defensive end may not view a 3-4 outside linebacker as high because he’s not a good scheme fit even though he is viewed as a great prospect otherwise. Similarly, teams that play a lot of man coverage may not value a cornerback whose strength is zone coverage. Another example might be a cornerback that consensus boards rank as a sixth-round pick at cornerback but one or more teams may view as a fourth-round safety. And teams may also vary in their willingness to deal with injury or character risk with a given prospect, among other examples.
But while critics of consensus board judgments point out (rightly) that the only way to judge draft picks is after 2-3 years of on-field performance, consensus board proponents point to some findings that are difficult to ignore as well.
Reaches and Steals
According to Arif’s research, significant reaches in the draft (pick is 15%+ lower number than consensus board ranking) in the draft underperform their draft pick value roughly 75% of the time among the top 100 picks of the draft. However, significant steals in the draft (pick is 15%+ higher number than consensus board ranking) still underperform roughly 25% of the time among the top 100 picks of the draft. Assuming that overall 50% of draft picks outperform and 50% underperform their draft pick value, there would appear to be some basis for being more critical of significant reaches in the draft. On the other hand, small reaches have outperformed, on average.
But while big reaches outperform in certain instances, the vast majority of them underperform.
And yet this is what every team in the league does.
If you visit this site, you can view how individual teams and the league overall have drafted relative to the consensus board and drill down to individual draft picks going back to 2016. Looking at the League View from 2016-2026, every team in the league is a net reacher in the draft.
While that may seem to explain why the league as a whole is so bad at drafting, there is some nuance as well. Not all reaches are bad, and not all steals are good, and sticking to the consensus rankings isn’t always the best approach either.
The Best Draft Strategy Based on Consensus Boards
Fans and pundits often talk about using a ‘best player available’ or BPA as the best draft strategy. But using consensus boards, that translates into selecting the biggest steal at every opportunity, which has not produced the best results. Prospects that fall significantly often do so for a reason that may not have been known or adequately reflected in consensus boards. But a strategy centered around reaches is unlikely to produce the best results either. Even simply following the consensus board and not reaching or stealing much isn’t likely to yield better results.
Instead, recent history from 2016-2023 suggests a varied approach at different points in the draft is optimal. Where those inflection points are depends on an assessment of the particular draft in terms of the number of top-tier prospects, depth of the draft, and where the overall prospect quality curve is steeper and where it flattens out.
The Optimum First-Round Strategy
Looking back at the first-round picks from the 2016-2023 drafts, along with their value relative to consensus board, and sorting the picks based on Pro Football Reference’s Approximate Value (AV) metric for the team that drafted the player (DrAV) to measure performance and value provided to the team that drafted the prospect. This produced the following results:
I sorted the picks based on DrAV to give a better visual presentation of what Reach/Steal type produced the top performing players and which produced the worst performing players in each draft class. I categorized Big Reaches generally as picks 15% and at least three picks above their consensus board ranking, Big Steals as at least three picks 15% below their consensus board ranking, Equal value picks are within one pick of the consensus board ranking and sometimes 2-3 picks for later round picks, and Small Reach and Small Steal picks generally as greater than one or two pick variance from the consensus board but less percentage-wise than Big Reaches and Steals. The idea is not to overstate 1-2 pick variances as big reaches or steals even though percentagewise or from a draft pick slot value or rookie contact perspective they may be significant.
At first glance, you’ll notice a variance from year-to-year among the top performing players drafted in the first round in terms of whether they were reaches or steals or equal value picks. However, you’ll also notice that Big Reaches are generally well represented among the worst performing first-round picks in just about every draft.
Crunching the numbers, a few patterns emerge in terms of optimal strategy.
Picks 1-9: Follow the Consensus
Among the top picks in the draft, staying with the consensus has produced the highest average DrAV per pick. This is where the consensus is best at identifying top prospects and deviating from consensus results in more failures than successes. In terms of DrAV per pick, ‘Equal’ picks dominate in terms of both average and median DrAV in this range.
Picks 10-20: Follow Conviction
This is an area of the draft where truly elite talents begin to thin out and differences in skill level narrow. Scheme fit and filling a clear starting role matter more than strictly following consensus board ranking. This is the one area where Big Reaches have produced the highest average DrAV per pick, although they still have a high bust rate. But the hits have really delivered and boosted the average. Following the team board over the consensus board in this range can pay big dividends, but it’s not without risk.
Picks 20-32: Look for Value
This is an area of the draft where it is best to exploit market inefficiencies and look for prospects with top 15 ratings that have fallen for reasons outside of performance concerns. Sometimes prospects fall in the draft not due to concerns about their ability or character or injury, but perhaps because they play a non-premium position, scheme fit, or teams were prioritizing other positions earlier in the draft which caused this prospect to fall. This is the area of the draft where Big Steals outperform all others in terms of average DrAV. Big Reaches here have a high bust rate and the hits haven’t justified the risk.
It’s important to note that the landscape of each particular draft will influence both when and whether these strategies will produce the best results. The key is determining how many truly elite prospects there are. The more there are in a given draft, the longer following the consensus may produce the best results.
Best Strategies for the Rest of the Draft
After the first round, the accuracy of consensus boards falls significantly- both in terms of predicting the best players and when players are picked. A big reason for that is because the talent differential flattens considerably. How teams value a prospect may be more role/usage-based rather than talent-based given the flattening talent differential. For example, a team may value what they perceive as a starting guard higher than a wide receiver who may compete, but not necessarily win, the WR2 spot. Or a team may value what they perceive as a starting slot cornerback higher than an outside cornerback who may only be a backup, even though overall talent levels are similar.
As a practical matter, most teams target a group of prospects expected to be drafted around each of their draft pick slots. At a position of need, a team may research prospects at that position expected to go around several of their draft pick slots. That way, if they go another direction early in the draft or a prospect they targeted early was drafted by another team, they have backup options at that position they may be comfortable with later in the draft. Researching different levels of the same position may also help them gauge the talent curve and identify the best place in the draft to target the position.
However, all this planning isn’t a strategy to draft the best players once the draft begins. And after the first round, predicting when prospects will be drafted becomes more difficult as consensus boards become increasingly unreliable and uncovering other team’s draft intentions can be difficult. So what is the best way to navigate this chaos?
The Best Second Round Strategy
In the second round, being a little aggressive in getting your guy has yielded the best results. Big reaches are far-and-away the worst approach at every point in the draft except the mid-first round, and even there not without risk. But small reaches to help ensure a top target becomes a draft pick has yielded the best results in the second round and indeed throughout most parts of the draft after the first round. In the second round, Small Reaches have outperformed all other strategies in average DrAV per pick. This is an area to focus on immediate role players, whether starters or rotational role players- guys who suit up and hit the field on game day. It’s not the place for trait-based projects. It’s also not a place in the draft where steals outperform either.
The Best Third Round Strategy
The third round is a place to be opportunistic. Small Reaches are still relatively successful, but Big Steals produce the highest median DrAV here. It’s a point to balance value with team fit. It’s not a point to take a chance on flawed players or make big reaches, but there can be some quality players that slip through the cracks in the second round who can become meaningful contributors here.
The Best Fourth Round Strategy
This is a round to be practical and focus on prospects with a clear path to get some real game snaps. Rotational players and special teamers. It’s not a time for projects. Big Steals again have produced the highest median DrAV here, but they tend to become less successful later in the round.
The Best Strategy for the Rest of the Draft
Beyond the fourth round, Big Reaches amounts to throwing picks away most of the time. Big Steals also underperform- they fell this far for a reason. Small Reaches for players with some utility have performed best in the late rounds, although there are generally few picks of this type even though Small Reaches have resulted in the highest average DrAV in the last three rounds of the draft. Overall having greater conviction on late round picks tends to produce better results than gaining better value compared to consensus boards.
It should be noted that Big Reaches, followed by Big Steals, dominate Day Three of the draft, despite the poor relative success rates. The consensus board accuracy in terms of when prospects are picked is lucky to be within a round, more or less, and frequently is not. There may only be a handful of picks each round that are even roughly around equal value compared to consensus boards.
So Why Do Teams Reach Big So Much?
Looking at the 2023 draft as an example (which I included at the end of this article), nearly 50% of Day Three picks were Big Reaches, which produced the lowest average and median DrAV of any pick type. Small Reaches produced the highest average DrAV but few picks on Day Three were of this type. Big Steals had the highest median DrAV and were the second-most frequent pick type, so looking for value in fallen prospects has some basis for success.
But the question remains: why do teams reach big so much with Day Three picks in particular, when they produce the worst success rate?
We can only guess at the answers, and there are probably more than one, but here are a few possibilities:
- Teams aren’t meeting with every prospect expected to go on Day Three, and they may be focused just on prospects they’ve met or have done more research on, which may cause them to reach with the picks they have available to them.
- Teams are drafting almost exclusively to fill needs at this point in the draft and value is less important from a draft capital and rookie contract value standpoint.
- 90% of Day Three picks don’t amount to much and there isn’t much of a talent differential, so what difference does it make, really? The highest median DrAV for the best pick type on Day Three (Small Reach) is still only 4 in the 2023 draft after three seasons, which isn’t very significant. The rare 1% of Day Three picks that really hit big varied in terms of being reaches or steals, so no clear precedent for finding first-round talent on Day Three.
- Beyond pick 200 or so, more and more teams are drafting based on whether the targeted prospect can be signed as a UDFA or not. Those they think may sign with another team, they may decide to draft to get their guy. Draft value has little or no significance at that point.
- While trading is more frequent on Day Three, finding trade partners to optimize draft value at every instance isn’t always possible on decent terms and may not really be worth the trouble.
Bottom Line on Consensus Boards
Consensus boards, like all NFL analytical tools and metrics, offer no predictive value. They are not particularly accurate in predicting which prospects will have the best NFL careers nor in predicting the actual draft order. But despite those known shortcomings, they can provide some value directionally in terms of navigating the draft and crafting draft strategies.
But as a tool for judging draft picks before the players hit the field, they may be more useful in assessing the risk level of a first-round draft pick relative to others based on draft value, understanding that even relatively “safe” picks don’t always pan out well. But as a tool to predict player outcomes, results are too varied to offer much certainty whether a particular first-round pick, or any other round pick, will pan out or not. Football is not an exact science.
Addendum: The 2023 NFL Draft
As a somewhat representative example, below is the results of the 2023 NFL draft along with Pro Football Reference’s AV values for each player for the team that drafted him (DrAV), and the variance in pick number relative to Arif’s consensus board. Positive CBB +/- numbers indicate the prospect was selected higher than the consensus board and therefore represent reaches beyond a certain amount, and negative numbers represent steals.
