The DraftBot Directives

Guidelines for grading the Commanders draft

Since the 17^th century, humanity has been embroiled in conflict between, superstition and religion, on the one hand, and science and reason on the other. Even today, many look to faith in the supernatural for answers to life’s big questions. While others recognize the value of science and technology to deliver solutions that work reliably and repeatedly.

The tension between science and religion can appear in unexpected places. It is even with us as we contemplate the decisions that Adam Peters will make in his first draft as Washington’s GM. Former Hogs Haven contributor, KyleSmithForGM implored Peters’ predecessors to look to received wisdom from on high, in the form of Draft Commandments (Vol. 1, Vol. 2, Vol. 3, Vol. 4, Vol. 5), to guide their draft selections.

The BrisVegas Systems DraftBot v. 4.0 is a triumph of modern technology. It’s prototype was originally conceived in a thought experiment to test the hypothesis that finding a franchise QB was simpler than the Commanders’ previous owner had made it appear. That successful demo attracted funding from a shadowy Australian intelligence agency.

Several generations and a few unfortunate incidents later, the current DraftBot version is powered by the most advanced artificial intelligence engine ever seen. DraftBot 4.0’s bio-cybernetic processor has been engaged in a deep data mining exercise to identify and validate empirical principles to optimize an NFL team’s ability to achieve success in the draft.

Interestingly enough, it’s scientific approach has validated many of the core beliefs professed by Hogs Haven’s keeper of the faith. Following its successful Redraft of the Rivera Era, BrisVegas Systems was approached by KS4GM’s successor, Bobby Gould, with a proposal to enlist the DraftBot to grade the Commander’s selections in the 2024 draft. Seeing an historic opportunity to reconcile the centuries old conflict between faith and reason, and perhaps a final chance to beta test a new technology that is about to be deployed to an undisclosed middle eastern location, the BrisVegas Systems executive board eagerly accepted the proposal.

The DraftBot has issued the following directives for NFL front offices to optimize their decision making in the draft. The DraftBot Directives will provide the basis for the DraftBot’s snap grades after Round 1 and at the end of the draft.

---

The DraftBot Directives

I. Prime Directive: Prioritize Drafting Quarterbacks

This is the core algorithm upon which all successive versions of the DraftBot’s firmware and network processing have been overlaid. It is hardwired into the fabric of the DraftBot’s neural network.

The reason that Dan Snyder failed to find a franchise quarterback is that he tried every approach but the most obvious, tried-and-true method. If your team is not set at quarterback, you have a high first round pick, and there is a QB with a first round grade available, take him.

Whatever you think your biggest need is, if you don’t have a quality starting QB, that’s it. Don’t pick a CB who needs glasses, don’t pick a DE, even if he’s the highest rated player in the draft class. Don’t even pick a future HOF safety. Pick the quarterback.

There are two caveats to the Prime Directive:

1. Squandering draft capital to acquire a QB undermines his chance of success (see the Corrollary to the Second Directive, below).
2. Reaching for a QB who does not have a first round grade is counterproductive (a.k.a. “The Owner’s Yacht Rule”).

Failure to comply with the Prime Directive may result in corrective action.

II. Second Directive: Draft Picks Are the Future – Do Not Squander Draft Capital

The DraftBot views the draft as the primary tool to build and maintain an NFL team’s competitive edge. Draft picks represent the team’s future and should not be squandered.

The DraftBot’s analyses have revealed that the best drafting NFL teams jealously protect their draft capital and seek to acquire more to maintain their ability to continually restock their talent pipelines. Its analysis of draft trades over the past two decades has revealed that the reason that teams give away excess draft capital in trades is a confluence of two frailties of the human brain: living in the moment (impulsiveness) and the capacity for self-deception.

Teams give away more value than it’s worth to “go get their guy” because they prioritize immediate desires over their team’s long term future, and because they convince themselves that their scouting evaluations are more predictive than is warranted by the facts. These phenomena reach their apex when teams trade up to draft highly-rated QB prospects at the top of Round 1. This failing is the subject of an important corollary:

Corollary of the Second Directive: Never Trade into the Top 5 for a Quarterback

Since the current rules on rookie contracts came into effect, seven teams have attempted to trade up into the top 5 picks for a highly-rated QB. Jared Goff is the only prospect aquired in those trades who has gone on to become a successful franchise QB. However, even he could not meet the outsized expectations generated when the Rams surrendered a future first and four Day 2 picks to move up from 15^th to 1^st to go get him. No other QB targeted in these trades stuck long term with the drafting team. A few were among the most colossal busts in recent draft history.

Teams that make these trades are making massive bets against the odds. Projecting college players to the NFL is hit and miss. Success rates for drafting starting QBs top out at around 50% for players drafted first overall. The hit rate is much lower for drafting premium starters, which is what teams making these trades expect to get. Trading away multiple early round draft picks for one 50% chance to hit on a starting QB defies logic. The reason these trades fail is that the decision makers have switched off the rational parts of their brains and are running on gut feel.

Photo by Jonathan Newton/The Washington Post via Getty Images

III. Third Directive: Use the Draft to Raise the Talent Ceiling

Constraints imposed by the salary cap dictate that the most effective way to build and maintain a competitive roster is through the draft. Teams that have attempted to become competitive by signing high end free agents, such as the Snyder/Cerrato Redskins, invariably fail. Diverting draft picks away from the primary mission to address immediate roster needs is counter to the team’s long term interest. Roster holes should be patched with inexpensive players in free agency to bridge any gaps until better long-term solutions become available.

Corollary to the Third Directive: Pick the Best Players Available

The DraftBot has observed that fans and most media pundits tend to have different ideas about draft strategy than GMs of the better NFL teams. While fans and pundits tend to focus on addressing immediate needs, the GMs preach picking the best players available, with need as a secondary consideration. In the words of Ravens’ GM, Eric DeCosta:

“I think it depends on best player, and also, team need. Both…If there’s a player that’s there that we think is too good to pass up on, we’re gonna take that player. We’ve proven that. We’ve shown that. That started with Ozzie Newsome and that’s gonna continue as far as I’m concerned as long as I’m here. We’re going to take the best available player.

“Now that being said, if you’re picking in the first round and you have an MVP quarterback and there’s a quarterback that’s the best available player, chances are we’re not going to draft the quarterback. But all things being equal, we’re going to draft the best available player when we’re on the clock.”

Conversely, the DraftBot has observed that noncompetitive teams often do prioritize need with their early round picks. If it becomes a regular habit, this tends to seek and maintain a level of mediocrity over time. Perhaps counterintuitively, these teams also tend to develop more and deeper roster holes than teams which draft proactively to continually restock their rosters with the best talent available.

The DraftBot frequently cites the example of Ron Rivera’s 2021 draft, in which he passed over left tackle Christian Darrisaw, who played the same position as recent FA acquisition Charles Leno Jr, in favor of linebacker Jamin Davis. Davis was rated as a second-round prospect by the majority of analysts, but played a position that Rivera and many analysts considered a greater need. Darrisaw, was near the top of most analysts boards when Rivera was on the clock. A few years later, the Commanders developed a massive need at OT, when Leno’s play declined as injuries took hold.

First Corollary to the Corrollary: Never Let an OK Player Stop You from Drafting a Great Player

The Davis-Darrisaw decision illustrates how a myopic focus on immediate need tends to perpetuate mediocrity. By opting not to draft Darrisaw, because he had a passable starter in place, Rivera passed on the opportunity to acquire a top-rated player at a premium position. The following year he repeated the mistake by passing on elite safety prospect, Kyle Hamilton, who had unexpectedly slid out of the top-10, to trade back and pick WR Jahan Dotson. That pick addressed a perceived need to give his new QB Carson Wentz more weapons. Dotson was projected to go about half a round later than where Rivera picked him.

By reaching for need with his first round pick two years in a row, Rivera pass on opportunities to add two elite players to his roster.

Teams that regularly contend for championships tend to have rosters that are stacked with high-end and elite players. They don’t get that way by directing their high-value draft capital toward players where their rosters are the weakest. They get there by making a habit of drafting the best talent when it is available to them.

The DraftBot sees drafting for need as form of complacency. A GM should never tolerate mediocrity at starting positions and should never pass on opportunities to upgrade starting positions with elite talent.

Second Corollary to the Corollary: Drafting for Need is a Myth

Even if it were advisable, the DrafBot questions the practical reality of addressing needs through the draft. The DraftBot suspects that drafting for need is a hoax, propagated by less informed segments of the sports media. There are two reasons:

1. It is nearly impossible to predict what a team’s biggest needs will be by the start or middle of the season in April, when the draft is held. In the words of Packers GM, Brian Gutekunst:

A team’s need profile can change instantly with an injury to a starter or key contributor on any snap in a game or practice.

Of course, some needs are predictable, such as the Commanders’ chronic lack of OL talent. However, that is mainly a problem on bad teams which neglect roster hygeine. Regularly restocking the depth pipeline is the best way to guard against holes developing and to prepare for the unexpected.

2. Most drafted players take a season or more to be ready to contribute. Commanders’ GM, Adam Peters has commented that the supply of players who are ready to contribute right tends to dry up by the start of the third round:

“Usually the guys who are picked in the first and second round — the third round, sometimes — have starter grades on them. But it really just depends on how the board falls.”

After a team’s first two picks they are generally picking players who will be ready to contribute next season or later. It’s not logical to think of those picks in terms of needs for the coming season.

IV. Fourth Directive: Scout Prospects Against the Existing Roster

In practice, the DraftBot views the “Need vs BPA” debate as a false dichotomy.

Its study of Adam Peters’ draft process identified an approach to determining the Best Player Available which it has since discovered is practiced by many of the best teams. As described by Peters’ former Scouting director Tariq Ahmad:

What does it do for the scouting process when you have a roster that looks like there’s virtually a starter at every spot, the depth looks pretty good? Does that change how you approach it and where your eyes go when it comes to evaluating the talent?

TA: “Yeah, so it doesn’t change the process. We’re evaluating everyone through the fall as if we’re starting the roster from scratch to get the value exactly correct. But, as we talk about the players as we go in the winter and then, during the April meetings, we compare them to the guys on our roster and how they would fit in. So, initially it doesn’t change at all, but then we have to be able to clearly communicate what their exact value is a little bit later in the process.”

So, are you basically in essence scouting your own guys too to figure out?

TA: “Exactly, so everyone has a specialty that they focus on in those April meetings, and that’s one of the things that we ask them to do, evaluate our own roster and stack those guys amongst the guys on our roster.”

The same concept is reflected in the comments of DeCosta and Gutekunst quoted above.

The DraftBot advocates picking players with the greatest potential to improve the roster by comparing each prospect to players on the current depth chart. This involves evaluating the player’s chance of displacing a current player, the impact of that position on game outcomes, and the prospect’s eventual talent ceiling. A prospect at a position with a starting vacancy has a clear edge to earn a roster spot early. However, the team might achieve greater eventual impact by drafting an elite prospect at a premium position that’s currently occupied by an average starter.

Teams should always seek to draft the prospect with the greatest potential to elevate the roster, whether that results in filling a vacancy or upgrading an occupied position on the depth chart. Good teams should not go into the draft seeking to fill multiple roster holes in the first place.

V. Fifth Directive: Factor Position Value into Draft Decisions

Position Value is a critical consideration in identifying the Best Player Available. Position Value reflects a combination of two factors: Position Impact and Replacement Cost.

Position Impact is what people usually think of as Position Value. It is the relative impact that players at each position make to game outcomes. Quarterbacks obviously have disproportionate impact compared to everyone else. Edge rushers, wide receivers, cornerbacks and safeties also have high Position Impact. The Position Impact of offensive tackles is boosted because their contribution is so closely linked to that of the QBs playing behind them.

As a general rule, interior offensive line, linebacker, running back and TE have lower Position Impact.

However, the DraftBot does not treat position groups monolithically, when assessing Position Value. It is based on individual skillsets. For example, defensive tackles who can rush the passer are among the higher impact players, whereas space eaters are among the lowest. Similarly, coverage safeties and versatile safeties may have Position Impact approaching that of cornerbacks; whereas box-only safeties have similar Position Impact to linebackers.

Replacement Cost is the other part of Position Value. This measures how expensive it is to replace the production of player at a given position in the draft or via free agency. Replacement Cost is related to scarcity. Quarterbacks, again, dwarf all other positions in Replacement Cost. Starting-level QBs almost never hit free agency and are rarely found outside the top-15 picks in the draft.

Offensive tackle has the next highest Replacement Cost. Running backs and interior offensive line have the lowest Replacement Cost.

The DraftBot’s cost-benefit analyses have generated the following rules to help teams optimize their expenditure of draft capital:

1. Never draft wide receivers in the top-16 picks – The top half of the first round is a special range of the draft. Picks in this range provide access to the top-rated prospects, who have the highest chance of starting early and the highest projected ceilings. Picks in this range should be reserved for players with high Position Impact and high Replacement Cost. Wide receivers have high Position Impact, but low Replacement Cost, because they among two position groups (along with iOL) with no difference in outcomes between players drafted in the first and second rounds. Wide receivers begin to have acceptable value in the second half of the first round.

2. Never draft tight ends in the top-25 picks – Tight ends are similar to WRs, but the value inflection point falls a little later.

3. Never draft a running back in the top-50 picks – Running backs have the lowest Replacement Cost of any position group. The chance of drafting a RB is lower in the later rounds, but the difference in production from RBs selected in the early and later rounds is insufficient to justify spending high draft picks on them. Running backs are easily replaceable in the draft and free agency.

Photo by Kevin Sabitus/Getty Images

VI. Sixth Directive: Continually Restock the Depth Pipeline

The best insurance against future needs is solid roster depth. After around pick #80, where immediate starters become scarce, the DraftBot seeks to upgrade the depth ranks with players with potential to displace current backups. In particular, the DraftBot seeks to replace any career-backups on the roster with prospects with developmental upside.

The DraftBot seeks to achieve an ideal situation in which every starter is backed up by his eventual replacement. It has observed that the best teams, such as the 49ers, Ravens and Steelers, have deep rosters in which starters face competition from players beneath them on the depth chart. These teams are better able to withstand attrition throughout the season and loss of starters in free agency than teams with sparse developmental pipelines, like the Commanders.

The DraftBot also avoids drafting players with projected ceilings as backups. It prefers to gamble on traits-based developmental prospects in the later rounds, than players who are likely to peak with high floors and low ceilings. If a player occupying a roster spot does not have potential to develop into a starter or core special teamer, he needs to be replaced with one who is.

Depth at QB deserves particular attention:

Pick a developmental QB in the mid to later rounds every few drafts. A good backup QB is almost as valuable as starters at some positions. Continually refreshing the depth at QB gives the best chance to hit on a quality backup, and to find the rare players with starting potential. The DraftBot has a determined that a frequency of once every three drafts is about right.

VII. Seventh Directive: Seek to Accumulate Top 100 Draft Picks

NFL draft selection is highly uncertain. Scouting evaluations are not strongly predictive, resulting in high failure rates. The predictive value of player evaluations also declines from a maximum at the first overall pick to close to negligible near the start of Round 4. From around the fourth or fifth round, depending on position, the hit rate for draft selections approaches chance level and is largely dictated by the availability of quality players remaining in the prospect pool.

In a system governed by high levels of uncertainty, the best strategy for success is to make as many choices as possible. To maximize their chance of hitting on quality players, NFL teams should seek to accumulate as many draft picks as possible.

However, not all draft picks are equal. Picks earlier in the draft have the greatest chance of success and should be hoarded whenever possible.

The DraftBot has also identified two strategies to optimize teams’ accumulation of draft capital:

1. Trade top-10 draft picks for picks in the 11 to 50 range. The DraftBot has determined that trade vaule charts used by NFL teams overvalue early first round picks and undervalue picks in the late first and early second rounds. Trading top-10 picks for picks later in the top 50 returns maximum value.

2. Always seek future picks in trades. The DraftBot has found that future picks are usually discounted well below actual value in trades. Its results suggest that teams appear to value future picks in trades as the last pick in the round, or even discount them by a full round.

Discounting picks that heavily is illogical. It is equivalent to assuming that the trade partner will win the Super Bowl and end up picking last, or that the talent pool in the future draft will be drastically lower than that in the current one. Teams should take advantage of the opportunity to acquire future early round picks at unreasonable discounts whenever it is offered.

Photo by Scott Winters/Icon Sportswire via Getty Images

VII. Eighth Directive: Do Not Gamble with Early Round Picks

The DraftBot’s analyses have revealed that one of the most common causes of draft busts is gambling on traits-based projections in the first round. It considers early round draft picks too valuable to gamble on high risk prospects. It advises teams to avoid players in the top 50 who have elite traits not backed by quality film.

Prioritizing factors that are actually predictive of success, over elite traits with early round picks can also help to avoid busts. For example, quarterbacks with elite processing ability have the greatest chance of NFL success. Yet, for reasons it does not understand, NFL teams persist in prioritizing size and arm strength in QB evaluation, which are only marginally related to player outcomes. The DraftBot also puzzles at how little being able to catch the football appears to be weighted in selection of wide receivers in the first round and how much attention is given to 40 yard dash times.

In short, never pick a player in the top-50 who has not demonstrated an ability to play their position at a high level regardless of how enticing their physical traits may be.

IX. Ninth Directive: Use Day 3 Picks to Take Chances on Players with Latent Potential and Special Teams Value

From the fourth round onward, the value of draft picks become low enough to warrant taking risks on players with elite traits who are still developing at their positions. The DraftBot considers this to be a better use of late round picks than drafting players who are likely to have peaked or have projected ceilings as backups.

From the fifth round onward, players who project as core special teamers with developmental upside and specialists become good value. However, it prohibits trades to move up to select specialists, since they seldom work out.

---

Final Word

The DraftBot has spoken. The Commanders draft selections will be judged for compliance with optimal draft principles.

---