Analysis
The NFL draft value chart explained, and where it lies
The NFL draft value chart explained: where the Jimmy Johnson points came from, what the newer charts measure instead, and why price and value diverge.
By CricketTaken EditorialPublished Analysis19 min read
Ninety seconds after a draft trade is announced, somebody has posted a total. So many points went one way, so many came back, and a verdict follows about which club won. The points come from a chart. Almost nobody quoting them can say where the chart came from, what it measures, or why a first overall pick is priced at exactly 3,000 of anything.
The NFL draft value chart explained accurately is a fascinating object: a piece of 1990s front-office arithmetic that became a league-wide currency by accident, kept working long after the assumptions underneath it stopped being true, and is now surrounded by better charts that nobody quotes. It is also the clearest example in sport of a measurement changing the thing it measures.
There are three charts worth knowing, they disagree with each other violently, and the disagreement is the interesting part.
The chart everyone means, and what its numbers say
The original is a single column of points, one per selection. It carries Jimmy Johnson's name because his Dallas staff used it at the start of the 1990s, and it spread across the league by imitation rather than by decree.
The first overall pick is worth 3,000 points. Pick 5 is 1,700. Pick 10 is 1,300. Pick 16, the middle of the first round, is 1,000. The last pick of the first round is 590, the last of the second is 270, pick 100 is exactly 100, and the final non-compensatory selection at 224 is worth 2.
Published point values at nine selections. The scale runs from 3,000 at the first pick to 2 at the last non-compensatory selection.
Show the numbers
| Item | Chart points |
|---|---|
| Pick 1 | 3000 |
| Pick 5 | 1700 |
| Pick 10 | 1300 |
| Pick 16 | 1000 |
| Pick 32 | 590 |
| Pick 64 | 270 |
| Pick 100 | 100 |
| Pick 150 | 31.4 |
| Pick 224 | 2 |
Look at the shape rather than the numbers. The line falls off a cliff and then flattens into nothing. By the middle of the first round a pick has lost two thirds of the top selection's value. By the end of the third round it has lost ninety-seven per cent of it. The chart says a first overall pick is worth fifteen hundred of the last pick in the draft.
The steepness has a consequence that is easier to feel as arithmetic than as a curve. On this chart, the first overall selection is priced above picks 10, 16 and 32 put together.
- Pick 101300
- Pick 161000
- Pick 32590
Published Jimmy Johnson point values. The three selections sum to 2,890 against 3,000 for the first overall pick.
Show the numbers
| Item | Value |
|---|---|
| Pick 10 | 1300 |
| Pick 16 | 1000 |
| Pick 32 | 590 |
A club that believes that chart is being told, in points, that three chances at a starter are worth slightly less than one chance at a better one. Nobody who has watched a draft class develop believes that literally. Clubs used it anyway, for years, because the chart was not doing the job people assume it was doing.
The chart was never a valuation. It was a price list.
This is the misunderstanding that produces almost all the bad commentary, and it is worth being precise about.
The original chart was built by looking at trades that had actually happened and finding a curve that described them. That is a descriptive exercise, not a predictive one. It answers the question "what has this cost recently", which is exactly what a general manager needs at nine o'clock on a Thursday evening with four minutes on the clock and a phone in each hand.
Seen that way it was a superb piece of work. It gave thirty-two organisations a shared currency, so a trade could be proposed, checked and agreed inside a few minutes instead of being argued from first principles. Transaction costs collapsed. More trades happened. The chart earned its place.
The flaw is structural and cannot be fixed by adjusting the numbers. A chart derived from completed trades encodes the market's beliefs, including the market's errors. If every club in the league was systematically paying too much to move up, the chart recorded the overpayment faithfully, published it as the going rate, and then every club used it to check that its next overpayment was fair. A consensus cannot audit itself. That is not a criticism of the arithmetic; it is what happens when a descriptive tool gets used as a normative one.
It also froze. The curve came from a market operating under rules that no longer exist. Drafted players negotiated their own contracts then, the top of the first round came with genuinely punishing money attached, and the trading behaviour the chart recorded was behaviour under those conditions. The rules changed in 2011. The chart did not.
The newer charts measure outcomes instead, and they are much flatter
Two serious alternatives are public, and they attack the problem from opposite ends.
The first, built by Chase Stuart in 2012, starts from what drafted players actually produced. It uses Approximate Value, Pro Football Reference's attempt at a single number for a player's contribution in a season, sums it across the seasons a club controls the player, and averages the result by draft slot across many years. Then it scales the whole thing so the numbers are comparable.
On that chart the first pick is worth 34.6, pick 10 is 19.9, pick 32 is 12.5, pick 100 is 5.3 and pick 224 is 0.1.
The second, built by John Fitzgerald and Bryce Spielberger for Over The Cap, measures the market's later verdict instead. Take the players drafted between 2011 and 2015, wait until their rookie contracts finish, and look at what somebody paid them next. Express each second contract as a percentage of the average of the top five annual salaries at that position in that year, so a guard is measured against guards. Smooth the results across neighbouring picks to stop one outlier distorting a slot. What comes out is a curve describing how often a given draft position produces a player the open market wants to pay.
That chart is deliberately scaled to 3,000 at the first pick so it can be read directly against the old one. Pick 10 is 1,833. Pick 32 is 1,244. Pick 100 is 666. The last pick of a full seven rounds is 190.
Put all three on the same scale and the argument stops being a matter of opinion.
The gap at pick 32 is the whole story. The old chart says the last pick of the first round is worth a fifth of the first pick. The outcome-based charts say it is worth between a third and two fifths. At pick 100 the divergence is enormous: three per cent against fifteen or twenty-two.
Every one of those gaps is a place where a club moving up is paying for a difference the evidence does not support. And the direction of the error is always the same, because the old curve is steeper than reality everywhere below the very top.
One small detail gives the original away. Pick 1 is worth exactly 3,000 and pick 100 is worth exactly 100. Those are not numbers that fall out of a regression. They are the fingerprints of a curve fitted by hand to land on round figures at memorable places, which is entirely reasonable for a tool meant to be used at speed by people holding two telephones, and entirely unreasonable as a claim about football.
A worked trade, priced three ways
Take a hypothetical that uses nothing but published values. A club wants the fifth pick and offers the tenth plus the first selection of the second round, pick 33.
On the old chart it is giving up 1,300 and 580, so 1,880 points, to receive 1,700. That is a premium of about a tenth, which any general manager would wave through as the ordinary cost of moving up.
On the Approximate Value chart the same three selections are 19.9, 12.3 and 24.3. The club is surrendering 32.2 to receive 24.3, a premium of roughly a third.
| Selection | Old chart | Approximate Value chart |
|---|---|---|
| Pick 5, received | 1,700 | 24.3 |
| Pick 10, given up | 1,300 | 19.9 |
| Pick 33, given up | 580 | 12.3 |
| Premium paid by the club moving up | about 10% | about 33% |
The trade did not change. The measuring instrument did, and the club's overpayment tripled.
That is the practical shape of the whole problem. The old chart does not tell a club it is getting a bargain; it tells the club its overpayment is small. Run the same deal through a curve built on outcomes and the small overpayment turns out to be a large one, and it is large in exactly the region of the board where clubs make most of their trades.
Approximate Value is the weakest link in the best chart
The outcome-based charts are better instruments, and it is worth being honest about what they are built on, because the case against the old chart is often made with more confidence than the replacement can carry.
Approximate Value is a single number per player per season, designed to be roughly comparable across positions in a sport where positions do almost nothing in common. It leans heavily on playing time and on team-level results allocated down to individuals. A player on a good offence collects more of it than an identical player on a bad one. Linemen are notoriously hard for it, because there is very little in the box score that belongs to them.
Those limitations do not sink the chart. Averaged across many players and many drafts, a noisy measure of production is still a measure of production, and the systematic errors partly wash out. But they do mean the resulting curve inherits every bias the metric has, including a preference for players who played early on established teams, which is not a neutral thing to reward in a study of draft position.
The second-contract chart sidesteps this by outsourcing the judgement to the market: rather than deciding how good a player was, it asks what somebody was willing to pay him once his rookie deal expired. That is a cleaner instrument in one respect and a lagging one in another, since it can only measure classes whose second contracts have already been signed, and it inherits whatever the market was mispricing at the time.
Anyone comparing player value seriously runs into the same wall in every sport, which is why the arguments about efficiency metrics in football and the arguments about draft charts are really one argument wearing two hats.
Why the price curve and the value curve cannot be the same curve
Here is the step that reframes the whole subject, and it has nothing to do with scouting.
What a pick is worth to a club is not what the player produces. It is what the player produces minus what he has to be paid. Everything else is a distraction. A pick is not a player, it is the right to sign one at a price set before the player is known, and the value of that right depends on the gap between the price and the production.
The price side is not a mystery, an estimate or a negotiation. It is arithmetic written into the collective bargaining agreement.
Each draft is governed by a Year-One Rookie Compensation Pool, and every selection carries a Year-One Formula Allotment: a fixed fraction of that pool attached to that slot. The pool itself rolls forward year to year, growing either in step with the percentage rise in the salary cap or by a defined minimum increase, whichever is larger. The total pool covering all years of every rookie contract is calculated as the year-one pool multiplied by five and a half.
The agreement even shows its own working. In one worked example inside the text, the year-one pool was $232,741,000, the minimum salary owed to a player with no credited seasons was $495,000, and multiplying that minimum by the 224 non-compensatory picks produced $110,880,000 of unavoidable floor, leaving $121,861,000 as the part of the pool that actually varies by slot. Growth from one year to the next is applied to that variable part, not to the whole.
The contracts themselves are constrained just as tightly. A drafted player signs for four years. His signing bonus prorates over a maximum of four years rather than the five a veteran gets. Option bonuses, option exercise fees, non-exercise fees, buybacks, voidable years and any contract-within-a-contract are prohibited outright in a rookie deal. Annual increases are capped at a quarter of his year-one rookie salary unless the contract pays nothing but the minimum throughout. Guarantees in the third and fourth years cannot exist unless the preceding year is guaranteed on the same basis. First-round picks carry a fifth-year option that is explicitly non-negotiable and that the club cannot bargain away.
- 224Non-compensatory selections in a draft
- 4Years in every drafted rookie's contract
- 4Maximum proration years on a rookie signing bonus
- 5.5Total rookie pool as a multiple of the year-one pool
Set in Article 7 of the collective bargaining agreement rather than negotiated pick by pick.
Now put the two sides together. Production falls gradually as you move down the board, because scouting is not precise enough to make it fall sharply. Cost falls steeply, because the pool allotments were built to fall steeply. Subtract the second from the first and the surplus curve is far flatter than either, and in places it barely leans at all.
That is why the same pick can be simultaneously overpriced on the old chart and genuinely valuable. It is not a paradox. It is two different quantities being confused for one.
What Massey and Thaler actually argued
The best-known academic treatment came from Cade Massey and Richard Thaler, whose working paper on the subject appeared through the National Bureau of Economic Research in April 2005 under the title "Overconfidence vs. Market Efficiency in the National Football League". Its central finding is stated plainly: top draft picks are overvalued in a manner inconsistent with rational expectations.
The mechanism they proposed is a behavioural one rather than a football one. Clubs are too confident in their ability to rank players finely. That confidence is not stupidity, it is the natural output of a process where hundreds of hours of work produce a ranked list, and the list looks precise because it has numbers on it. But scouting cannot reliably separate the third best player in a class from the eleventh, and any pricing system that assumes it can will overpay for position at the top of the board.
The paper was interesting to economists for a reason that has nothing to do with football. The draft is close to an ideal test case: high stakes, expert decision-makers, repeated play, immediate feedback, strong financial incentives. If markets grind bias out of anybody, they should grind it out of these people. The finding was that they did not.
One important qualification is usually left out when the paper is quoted. It studied a market operating under the old rookie contract regime, when the cost attached to a top selection was large and negotiated. Cut that cost sharply, as the wage scale did in 2011, and surplus value at the top of the board rises. The critique of the old chart survives intact, because the chart's shape is still wrong. The stronger claim, that the first overall pick was a burden rather than a prize, does not survive the same way. A cheap, fixed, four-year quarterback contract is the single most valuable structural asset in the sport, and the arithmetic that makes it so is the wage scale itself.
How a chart is actually used in a trade room
The public version of this is a calculator. The private version is nothing like it.
- The board decides firstA club has one ranked list with tiers on it. The only question that matters is how many players it still wants are likely to be gone by its next selection.
- The offer arrives with a number attachedThe club calling has run the same chart and framed its proposal to look balanced on it. That framing is a negotiating position, not an appraisal.
- The chart gets run both waysAgainst the old points chart and against at least one outcome-based curve. The gap between the two answers is the size of the premium being asked for.
- Future picks are discountedA selection in next year's draft has no known slot, so it trades at a haircut, conventionally about a round for a year of waiting.
- The tier decides, not the totalIf several players remain in the same tier, moving up buys nothing and the club trades down or stands still. If one player is alone in a tier above the rest, the chart is overruled deliberately.
- The trade is filed and reportedThe points total appears within minutes, computed on a chart from the early 1990s, and the tier structure that actually drove the decision is never mentioned.
A chart never picks a player. It tells a general manager whether the price on the table sits inside or outside the market, and how far outside.
The fourth step deserves more attention than it gets. Discounting a future pick looks arbitrary and is not. The slot is unknown, the class around it is unknown, and money now is worth more than money later to an organisation whose head coach may not survive until later. The rough convention that a pick a year out trades at about the value of a pick one round lower today is crude, but it captures a real uncertainty, and it produces a useful side effect: a club willing to sell next year's second-round pick is telling you exactly how it sees its own timeline.
Positional value is the hole in every points chart
Every chart shares one blind spot, and it is a large one. A points chart prices a slot. Football does not pay slots, it pays positions, and the market's valuation of positions is wildly uneven and moves over time.
Two picks at the same number are not the same asset if one becomes a quarterback and the other becomes a running back, because the second contracts those two players will command differ by a multiple, and the surplus a club captures during the rookie years differs with them. A chart that assigns a single number to pick 12 is implicitly averaging across every position that has ever been taken at pick 12, which is a strange thing to do given that the club making the selection already knows which position it is taking.
The second-contract chart handles this better than the others, almost as a side effect. Because it measures each player's later salary against the top of the market at his own position, it captures how often a slot produces someone the market pays well relative to his peers. That is closer to the right question, though it still hands back one number per pick.
The blunt version of the problem is this: a club drafting a quarterback at pick 12 and a club drafting a guard at pick 12 are doing two economically different things, and the chart both of them are using cannot tell the difference. In practice front offices patch this with positional adjustments that never appear in public, which is one reason a trade that looks lopsided on the published chart can be perfectly sensible in the room.
What no chart can price
Four things sit outside every model, and between them they explain most of the trades that look inexplicable.
Medical information. A club's file on a prospect is not the public file. A player two clubs have flagged and a player nobody has flagged can have identical grades on every public board, and the examinations that produce that difference happen behind closed doors in February. Anyone modelling draft outcomes from public data is modelling a class with the injury risk stripped out of it.
Tier structure. The distance between the fifth and sixth players on a club's board is not fixed, and it is the only distance that matters at the moment of a trade. Some drafts have a cliff after four players; some run twenty deep at a position. A points chart is smooth by construction and cannot represent a cliff.
Certainty itself. Moving up is partly a purchase of certainty, and certainty has a value that varies with circumstance. A club whose head coach needs to win this season is rationally willing to pay more for a known outcome than a club three years from contention. The same trade can be right for one and wrong for the other, and no chart contains a field for that.
Knock-on effects. Trading a veteran or letting a free agent leave feeds the compensatory pick formula the following year. Signing one cuts the other way. Picks also cost roster spots, and roster spots are capped, so a club that hoards selections has to solve a different problem in August. None of that shows up in points.
Why trading down is the argument, and why it is not a law
The flat surplus curve makes an obvious recommendation: accumulate picks, move down the board, take more chances at a lower price. Read the charts side by side and the case is difficult to argue with in the abstract.
In the concrete it needs three qualifications.
The first is that surplus per pick is not the same as surplus per roster spot. A club can only field so many players and develop so many at once. Twelve selections in one draft is not twice as useful as six if the club can only coach half of them properly, and the marginal seventh-round pick displaces a veteran the club might have preferred.
The second is that the case against moving up is a case about average outcomes. If a club's evaluation of one specific player sits genuinely far above the consensus, paying above the model is correct, because the model is built from averages and the club is explicitly claiming this player is not average. The clubs that get burned are the ones that pay a premium for a player they have graded exactly where everybody else has, buying nothing but the certainty of getting him.
The third is that the charts themselves keep improving, and the market has partly absorbed the criticism. Public analysis of this kind changes behaviour, prices adjust towards the better model, and the edge available from simply knowing that the old chart is too steep is smaller than it was fifteen years ago. That is what happens to any published inefficiency, and it is a reason to be suspicious of anyone still selling the finding as a secret.
Where the money actually decides it
One structural fact reorganises everything above, and it belongs at the end rather than the start.
The rookie wage scale means a drafted player is cheap for four years at a price fixed before anyone knows whether he is any good. Every consequence flows from that: a miss costs almost nothing against the cap, which is why the surplus curve is flat, which is why the old chart is too steep, which is why trading down looks profitable in a model.
It also means the value of a pick is not fully realised inside the draft. It is realised across the four or five years afterwards, in what the club does with the space the cheap contract creates. A club that drafts well and then spends the surplus badly has captured nothing. A club that understands what it has bought will use those years to pay market rate everywhere else at once, which is where the cap accounting and the contract structuring stop being a separate subject and become the same one.
That is the honest place to leave the chart. It is a price list, it is thirty years old, it is too steep, and it is still the number you will see quoted an hour from now, because it is the only one that fits in a graphic. The way it distorts coverage of the selection meeting itself is worth understanding precisely because the distortion is so consistent.
Its second life: it stopped setting prices and started setting stories
The most interesting thing about the old chart now is not what clubs do with it. It is what everybody else does with it.
Inside a building, it has been demoted. It sits alongside two or three other curves, a positional adjustment nobody publishes and a tier structure that overrides all of them. It is a reference point in a negotiation, useful mainly because the club on the other end of the phone is looking at the same one.
Outside the building it has been promoted, to the status of a verdict. Every reported trade acquires a points total within minutes, and that total becomes the frame through which the deal is discussed for years. A club that paid a premium on the old chart is described as having lost the trade, whatever its board said. A club that collected points is described as having fleeced somebody, even where the outcome charts say it collected very little.
This has a real effect, and it runs in an unhelpful direction. The chart is steepest exactly where public disagreement is loudest, at the top of the first round, so the trades most likely to be judged harshly in public are the ones where the public instrument is most wrong. A general manager who understands the flat surplus curve and trades down accordingly gets to explain himself twice: once to the people who wanted the player, and once to the people holding a points total that says he collected far more than he did.
The lag is the ordinary fate of a good tool that outlived its assumptions. The chart solved a real problem in 1991, it solved it well enough that nobody replaced it, and it is now doing a job it was never designed for, in front of an audience that mistakes it for a law of nature.
How to read a reported trade in about a minute
Four questions, in order, and none of them is the points total.
Which direction was the premium going? Somebody moved up and somebody moved down. The club that moved up paid above the outcome-based curves, essentially always. The only question is by how much, and whether it had a reason.
Was a future pick involved, and whose? A club selling next year to buy this year is making a statement about its own timeline that is usually more informative than the trade itself.
What position did the mover take? A steep price for a quarterback and a steep price for a specialist are different transactions, and the chart cannot tell them apart even though everybody watching can.
What did the club that moved down actually do with the extra picks? This is the only part that gets settled by evidence, and it takes three years. The verdict on a trade down is not the points it collected. It is whether the additional selections turned into players anyone else wanted to pay.
Answer those and you are already ahead of every points total posted in the first ninety seconds. More on the wider machinery, from how prospects are graded to how the money behind the picks is accounted for, sits in the American football archive.
Common questions
What is the Jimmy Johnson draft value chart?
It is a single column of points, one per draft slot, that came out of the Dallas front office at the start of the 1990s and spread across the league. It prices the first overall selection at 3,000 points and falls away steeply from there, reaching 590 at pick 32 and 2 at pick 224. Clubs used it to check whether a proposed pick trade balanced, and its numbers still frame almost every trade reported in public.
Are NFL draft value charts accurate?
They are accurate descriptions of what trades have cost, which is not the same as being accurate about what picks are worth. The original chart was built by observing completed deals, so it recorded the market's prices rather than the players' production, and if the whole league was overpaying to move up the chart recorded the overpayment and handed it back as a standard. The newer charts measure outcomes instead and produce much flatter curves.
What is surplus value in the NFL draft?
Surplus value is what a drafted player produces minus what he has to be paid. It matters because the cost side is fixed in advance: every selection carries a predetermined share of a league-wide rookie pool, so a club knows what a pick will cost before it knows who it is taking. Since the highest picks come with the largest rookie contracts, surplus is spread far more evenly across the draft than price is.
Why do teams discount future draft picks?
Because a pick a year away has no known slot and no known class around it, and because a club trading for help now is usually short of players now. The market applies a rough haircut of a round for a year of waiting, which sounds crude and is defensible: the uncertainty is real, and a club willing to sell next year's pick is telling you something about its own timeline.
Which draft value chart do NFL teams actually use?
Most clubs run several, and the honest answer is that no chart decides anything. A chart is a negotiating reference and a sanity check, not a valuation. The board decides, the chart tells the general manager whether the price he is about to pay is inside or outside the market, and the interesting decisions happen when a club deliberately ignores it.
Filed under American Football·nfl · draft · contracts · roster building · analytics