I haven’t watched a minute of the NFL Combine in probably ten years. I follow the league, watch most games through the season, catch the big headlines in April — but the combine? I checked out a long time ago. Forty times don’t move me. Bench press reps don’t move me. It’s four days of content engineered for people whose football fandom runs a few levels deeper than mine.
Which makes me exactly the kind of person the NFL has already lost for that weekend.
I was listening to the 3 and Out podcast with John Middlekauff and he was floating the idea of a draft lottery timed around the combine. Something about that framing clicked immediately — not just “yeah, good idea,” but the specific question of how you’d actually build it. What mechanism would make it genuinely appointment television? What would make a Browns fan and a Giants fan both clear their Sunday night?
I spent a few days turning the concept over, then worked with Claude to develop the full proposal. I want to be upfront about that: I didn’t hand Claude a prompt and copy-paste the output. The core ideas — the Fibonacci mechanism, the inversion, the re-seeding logic, the dispensing tower concept — all came from actual thinking about what I’d want to watch. Claude helped synthesize the pieces and pressure-test the math. But this is genuinely the lottery I’d tune in for.
Here’s the full proposal.
The Problem the Lottery Solves
The NFL owns the American sports calendar at a handful of moments: the September kickoff, Thanksgiving, the January playoffs, the Super Bowl, the April draft. Those are the peaks. The valley between the Super Bowl and the draft — late February through late March — is the quietest stretch on the NFL calendar. It’s when fandom goes into hibernation.
The combine sits right in the middle of that valley and does almost nothing to fill it for casual fans. It was built for scouts, analysts, and front offices. It’s become increasingly niche. The hardcore audience shows up; everyone else doesn’t. I haven’t tuned in for any of it in a decade and I’d bet a lot of people reading this are in the same boat.
A lottery anchored to the closing night of the NFL Combine — a Sunday primetime broadcast — changes the entire math. The hardcore audience that was already watching all week gets a dramatic live payoff at the end. The casual fan who ignored the whole event now has a reason to tune in for the finale: their team is in the drum and the top ten draft order is about to be set in real time.
For the fans of the ten worst teams specifically, this is the first genuine appointment TV moment since their season ended in disappointment months earlier. Hope is the product. The combine becomes must-watch television — not just for the enthusiasts.
Why Not Just Copy the NBA?
Every time I started sketching this out, I kept bumping into the same assumption: a draft lottery is a competitive balance tool. That’s the NBA model — teams get lottery odds based on record, the lottery exists to give bad teams a better shot at transformative picks, the envelope reveal is a boardroom ceremony on ESPN.
I didn’t want any of that.
The NBA lottery works in a league where a single player can change everything — where the difference between the 1st and 5th pick is potentially the difference between a dynasty and irrelevance. The NFL is a different sport. It’s a 53-man roster game with pervasive injury risk, salary cap mechanics, and scheme dependency baked into every decision. No single draft pick reliably transforms an NFL franchise the way LeBron or Kobe transformed theirs. The league already self-corrects with remarkable regularity. Teams cycle in and out of contention. The existing draft order isn’t broken.
So the lottery doesn’t need to fix parity. It needs to be television.
That reframe changed everything about how I approached the design.
The Fibonacci Lottery
The premise: the 10 lowest-ranked NFL teams enter a lottery to determine draft order for picks 1 through 10. The lottery runs as 9 successive draws, each assigning one pick — starting from pick 10 and working down. The last team standing gets the #1 pick.
The worst team should still win the #1 pick most of the time. That’s important. But “most of the time” leaves meaningful room for drama, and that room is the point.
Ball allocation
Each team gets lottery balls based on their finishing record using the Fibonacci sequence — but inverted in a specific way. In software I use Fibonacci numbers regularly for sizing work estimates: you pick from the sequence (1, 2, 3, 5, 8, 13, 21…) because each number roughly doubles the previous, capturing the non-linear nature of uncertainty. The same logic works here, but I needed to apply it carefully.
Because each draw assigns the least desirable pick first (we’re counting down from 10), we want to protect the worst team from early elimination. So the worst team starts with the fewest balls in round 1, and the best team (10th worst) starts with the most. Round 1 draws the 10th pick — the team most likely to be drawn is also the team that can most afford to land there.
| Rank | Balls (Round 1) |
|---|---|
| 1st worst | 1 |
| 2nd worst | 2 |
| 3rd worst | 3 |
| 4th worst | 5 |
| 5th worst | 8 |
| 6th worst | 13 |
| 7th worst | 21 |
| 8th worst | 34 |
| 9th worst | 55 |
| 10th worst | 89 |
| Total | 231 |
In round 1, the 10th-worst team has an 89/231 chance of being eliminated and assigned the 10th pick. The worst team has a 1/231 chance. That’s 0.43% — nearly impossible.
Re-seeding each round
After each draw, the eliminated team is removed and the remaining teams are re-seeded with fresh Fibonacci counts based on their updated relative ranking. If the 4th-worst team gets eliminated in round 1, the remaining 9 teams re-rank from 1 through 9 and the ball counts reset accordingly.
This keeps the math honest and proportional in every round regardless of prior results. It also means the ball counts for each round can only be dispensed after the previous draw is confirmed — which becomes a central part of the production design.
The dramatic arc
The escalating elimination risk for the worst team tells the show’s story in a single table:
| Round | Pick Assigned | Worst Team Elimination Risk | Narrative Tension |
|---|---|---|---|
| 1 | #10 | 0.43% | Nearly impossible |
| 2 | #9 | 0.70% | Still feels safe |
| 3 | #8 | 1.15% | Barely registering |
| 4 | #7 | 1.89% | Fans start paying attention |
| 5 | #6 | 3.13% | Tension building |
| 6 | #5 | 5.26% | Genuinely nervous |
| 7 | #4 | 9.09% | Real sweat |
| 8 | #3 | 16.67% | Heart pounding |
| 9 | #2 | 33.33% | Coin flip territory |
| 10 | #1 | 100% or 0% | Climax |
Nearly impossible in round 1. Still comfortable in round 2. Barely registering in round 3. By round 8 the worst team’s rep is on the edge of their seat. By round 9 it’s a coin flip. That escalation is the show’s dramatic arc, mapped mathematically.
Overall odds
Across all possible lottery outcomes, here’s how the #1 pick probability distributes (approximate, based on simulation):
| Rank | Overall Odds of #1 Pick |
|---|---|
| 1st worst | ~58% |
| 2nd worst | ~38% |
| 3rd worst | ~20% |
| 4th worst | ~9% |
| 5th worst | ~5% |
| 6th worst | ~2% |
| 7th worst | ~1% |
| 8th worst | ~0.3% |
| 9th worst | ~0.1% |
| 10th worst | 0% |
The worst team wins the #1 pick most of the time. Traditional draft order is preserved in the aggregate. But roughly 4 out of 10 years the top pick goes elsewhere — and that uncertainty is the entire product.
The bottom five teams don’t need to win the lottery to benefit. Every upward move matters in a deep draft class. This creates distributed drama across all ten fanbases simultaneously — every rep on stage has something genuinely worth rooting for.
| Rank | Real Goal | What Counts as a Win |
|---|---|---|
| 1st worst | Win #1 | Survive every round |
| 2nd worst | Steal #1 | Land top 2 |
| 3rd worst | Top 3 | Any top-2 finish is an upset |
| 4th–5th worst | Jump 2–3 spots | Real roster impact |
| 6th–8th worst | Avoid their seed | Any climb is a victory |
| 9th–10th worst | Lightning strike | Surviving round 1 is already a moment |
The Production
The math is the skeleton. The production is the show.
The setting
Sunday night, closing weekend of the NFL Combine in Indianapolis. Ten illuminated towers standing in an arc behind a central lottery drum on a purpose-built stage. A live audience of combine attendees. A host at center stage flanked by a two or three-person analyst desk. Ninety minutes of primetime.
Each team sends a fan-favorite player or franchise legend as their representative — not an executive in a suit. The rep is the human stakes. When a tower goes dark, a person rises and walks off stage to a crowd acknowledgment. When a tower blazes at the end, the rep is overcome with the weight of a brutal season transformed into genuine franchise hope.
The dispensing towers
Because re-seeding is fully dynamic — every team’s ball count in each subsequent round depends on who was eliminated before them — you can’t pre-fill anything beyond round 1. The only certainty in the entire system is that the worst team always has exactly 1 ball in whatever round they’re playing. Every other team’s count is only known once the prior elimination is confirmed.
The solution is a purpose-built automated dispensing tower system.
Ten illuminated transparent towers stand in the arc — one per team, lit in their colors. Each tower connects to a central computer pre-loaded with every possible re-seeding combination, certified by an independent auditing firm before the event. The moment a ball is drawn and a team is eliminated, the system calculates the correct Fibonacci count for every surviving team and dispenses that exact count into each tower’s staging cylinder — visibly, audibly, simultaneously.
The reset sequence is part of the show:
- A ball pops — a team is eliminated — the crowd reacts
- That team’s tower extinguishes as their rep rises and walks off stage
- Surviving towers pulse briefly as the system recalculates
- Colored balls drop into each cylinder — the sound of one ball clicking into the worst team’s cylinder against a cascade rattling into another tells the entire odds story without a word spoken
- A broadcast graphic confirms each count as cylinders fill
- Crew pours each cylinder into the next drum with team auditors watching
- Drum is sealed — next round begins
The entire reset takes under three minutes. It’s the between-round content — no dead air, no hiding the logistics. Transparency is baked in from the start: all re-seeding logic is published publicly before the event and certified by an independent auditing firm. Multiple broadcast angles cover every draw and every dispense with no cuts. The transparent towers and cylinders make every ball count visually verifiable by team representatives, auditors, and viewers at home simultaneously.
The graduated drums
A single drum can’t handle this cleanly. A lottery drum is engineered to randomize a specific volume of balls. With 231 balls in round 1 the drum performs as designed. By round 9 there are as few as 3 balls rattling around in a chamber built for hundreds — the randomization is compromised, the draw looks awkward on camera, and the mechanical reliability of the most-watched moment of the night is genuinely at risk.
The solution is a set of purpose-built graduated drums, each sized and calibrated for its assigned range:
| Drum | Rounds | Ball Count Range |
|---|---|---|
| Drum 1 | Round 1 | ~231 balls |
| Drum 2 | Rounds 2–3 | ~90–140 balls |
| Drum 3 | Rounds 4–5 | ~32–53 balls |
| Drum 4 | Rounds 6–7 | ~11–19 balls |
| Drum 5 | Rounds 8–9 | ~3–6 balls |
Each drum is calibrated — chamber volume, airflow, or tumbling speed — to produce consistent genuine randomization for that range. The draw mechanism works identically at 231 balls and at 3, because each drum is purpose-built for exactly that load.
The drum transition is its own theatrical beat. After each round’s dispensing sequence, the next drum — visibly, obviously smaller — is already positioned and illuminated on stage. The retiring drum is ceremonially removed. By round 8 the drum is almost intimate. Small enough that individual colored balls are visible through the casing from broadcast cameras. Three balls in a purpose-built small drum is a riveting image. Every person watching can count them.
The stage itself becomes the visual narrative of the entire show. Ten towers blazing and a large drum at the start. Towers extinguishing one by one. Drums shrinking round by round. By round 8 the stage is mostly dark — two towers lit, one small drum at center stage with a handful of balls visible inside it.
That image — two lit towers, one intimate drum, three colored balls — is the show’s defining visual. It’s a broadcast thumbnail. It’s the moment that gets replayed every year.
The ninety-minute format
The breathing room is intentional. Each round is a chapter. Between draws, analysts break down what the result means for each remaining team — who they might target, why moving up changes the offseason calculus, what this draft class offers at each slot. The lottery becomes a live stakes event wrapped around a draft preview show. By the time the #1 pick is announced, the audience has consumed a fully produced draft preview and ten distinct emotional moments — one for each fanbase watching.
A rough breakdown of how the ninety minutes breathes:
| Time | Content |
|---|---|
| 0–15 min | Opening video package, team rep introductions, system walkthrough using the towers, drum sealed |
| 20–35 min | Rounds 1 & 2 (picks #10, #9) with analyst segments before and after each draw |
| 40–58 min | Rounds 3–5 (picks #8, #7, #6) — the upset zone, where mid-tier teams survive and the stage thins |
| 63–78 min | Rounds 6–8 (picks #5, #4, #3) — stage mostly dark, worst team rep visibly anxious |
| 83–88 min | Round 9 (pick #2) — two towers, three balls, near silence, no lengthy analyst segment needed |
| 88–90 min | Round 10 (pick #1) — no draw, last tower standing, confetti, live call to the coaching staff |
The final two rounds don’t need analysis. They just need to breathe. Round 9 is two towers, two cylinders, one quiet click versus two rattling balls, three balls in a drum, the host saying almost nothing. Round 10 is no draw at all — the last tower is the winner. The host walks over. Confetti drops. A live phone call to the coaching staff over the arena speakers.
The draft is real now. The combine week has its closing ceremony.
What I Don’t Know
I’ll be honest about the limits here. I’m not a statistician and I don’t have a background in lottery mechanics. Most of what I built out I’m reasonably confident in at the conceptual level, but the randomization integrity at small ball counts is the piece that would absolutely need expert input before this goes anywhere real.
Three colored balls in a drum engineered specifically for three balls should work — that’s precisely why the graduated drum system exists. But “should” isn’t the same as “does,” and proving the mechanism is provably fair at every ball count across every round is something that has to be validated by people who actually certify lottery equipment for state governments. I believe it’s solvable. I can’t solve it myself.
Same goes for the independence of the auditing process, the certified pre-loading of every possible re-seeding combination, and a dozen other details that matter enormously when the integrity of the draw is the single thing you cannot undermine at any point in the production.
What I can say is that nothing in this proposal requires exotic engineering. The same firms that build certified drawing machines for state lotteries and game show randomizers have the precision manufacturing and regulatory certification experience to build exactly this kind of system. The NFL would commission the hardware once and use it annually for decades. Against the broadcast rights and advertising revenue a new primetime event generates, the capital outlay is negligible.
Whether any of this ever gets in front of the people who could actually make it happen — I have no idea. But it’s the lottery I’d watch. And I think a lot of other casual fans who wrote off the combine years ago would too. If you think the whole thing is a terrible idea, that’s fair. It was genuinely fun to think through either way.