One athlete-year, run twice.
The same deals, the same market, two ways: the process as it works today, and the same deals on NIL33's rails. Rules and market figures are real and sourced. The simulated athletes are not real people, and every simulated number is labeled as model output.
Press play. The page switches sections and runs the simulations as the narration reaches them. You can pause, skip with the chapter list, or turn off "move the page" and just listen.
Narration is a computer voice reading the script shown in the captions. Download the MP3 · Script and timings (JSON)
Instructions
Pick the athlete
Open Run a simulation. Choose the school's revenue-sharing pool, the sport, and where the athlete sits on the roster.
Set the agent
Set the agent's fee. Turn on the 5% cap to apply it to the NIL33 column, the cap two federal bills propose.
Run
Press Run. It plays up to 5,000 athlete-years on the same random draws for both columns and replays one athlete's deals week by week.
Compare and check
Read the leakage lines side by side, then open Method to see which inputs are cited and which are assumptions.
Everyone around the athlete
Who touches a college athlete's money, what usually happens today, and what NIL33 puts on the record. The same parties appear in the simulation.
The rules and the real numbers
Published facts only. Each line carries its source and date. Last checked 2026-11-20. These figures set the simulation's base rates; see Method.
Estimated revenue-sharing cap per school for 2026-27, up from $20.5M in 2025-26 under the House settlement's 4% escalator.
Inside the Loudhouse · Columbus Dispatch, 2026Paid directly to 34,915 athletes by 307 schools. Only 68 of 319 opted-in schools reached or came within 5% of the cap.
The College Investor, 2026Cleared versus rejected since launch on June 11, 2025. May-June 2026: average cleared deal $14,792; average rejected deal $51,593.
ESPN, CSC report of July 8, 2026Share of deals resolved within 24 hours, and within seven days once all required information was in.
AP via Fox Sports, Jan. 12, 2026| Topic | What the record shows | Source |
|---|---|---|
| Threshold | Every third-party NIL deal of $600 or more goes through NIL Go, run by Deloitte for the College Sports Commission. | theScore, June 2025 |
| Why deals are rejected | No valid business purpose; pay not in line with similarly situated athletes; no meaningful activation of the athlete's NIL. | ESPN, July 2026 |
| Arbitration | In May 2026 the CSC won an arbitration over deals between about 18 Nebraska football players and Playfly Sports. | ESPN, July 2026 |
| Data corrections | The CSC's first report said $79.8M cleared; it corrected that to $35.42M a day later, citing a clerical error in Deloitte's data. | On3, Sept. 2025 |
| Collectives | After CSC guidance that collective deals needed another third party, House settlement lawyers reached a basic agreement on collectives. | Front Office Sports, Oct. 2026 |
| Agent fees, federal | No federal cap today. The Protect College Sports Act of 2026 would cap agent fees at 5% and require state registration. The SCORE Act, which also had a 5% cap, was pulled from the House twice. | Senate Commerce · Morgan Lewis, June 2026 |
| Executive order | Executive Order 14400 (April 3, 2026) calls for a national student-athlete agent registry. | Knight Commission, May 2026 |
| FTC | On Jan. 12, 2026 the FTC sent letters to 20 NCAA schools asking whether agents representing their athletes complied with federal law. | AFCA · Duane Morris, 2026 |
| States | More than 40 states have a version of the Uniform Athlete Agents Act. Louisiana now requires NIL agents to register with the state Department of Justice, pass a background check and complete training. | Wiley · WAFB, Aug. 2026 |
Reported athletes, agencies and brands
Real people and companies as reported in the press. None of them is a NIL33 client or endorser, and none appears in the simulation.
| Who | As reported | Source |
|---|---|---|
| How valuations work now | On July 1, 2026 the On3 NIL Valuation moved from an algorithm to a deal-based model that reflects a player's current contract value. | Sports Illustrated, July 15, 2026 |
| Darian Mensah, Miami QB | No. 1 On3 NIL Valuation for 2026-27 at $6.5 million. Ohio State's Jeremiah Smith was No. 4, the highest-ranked non-transfer; six of the top ten were basketball players. | Sports Illustrated, July 2026 |
| Brands, 2025 season | Powerade ran an NIL campaign with LSU QB Garrett Nussmeier among others. adidas's college NIL roster includes Ohio State WR Jeremiah Smith and Nebraska QB Dylan Raiola. | Arizona Sports, Aug. 2025 |
| Agencies, first NIL year | CAA signed Bryce Young and Paolo Banchero; WME signed Olivia Dunne and facilitated her Vuori deal; Wasserman represented Paige Bueckers; Klutch signed Kedon Slovis. | Front Office Sports, 2021 |
Simulation output. Modeled athletes, not real people. Deterministic: the same inputs and seed always give the same result.
Pooled fund
Most outside-deal money goes to a few athletes in two sports. This tab models one way to share some of it: a small part of each athlete's outside-deal money above an exemption goes into a pool, and the pool is paid to the athletes who earn least. Some people call this a pooled tax. It is not a tax. It is a rule a team, a school or a collective could adopt and athletes could agree to, and the ledger could carry out. NIL33 does not run one today.
Simulation output. Modeled athletes, not real people. Ten simulated departments per run. Deterministic: the same inputs and seed always give the same result.
By sport, average athlete's total for the year
By sport and roster spot
Total = school revenue share + outside-deal money actually received. The pool takes nothing from school revenue share. "Today's process" is the same athletes and deals without NIL33 rails and without a pool. No cost of running the pool is modeled.
Published pooled fund scenarios
Eight scenarios, . Download results-pool.json · Model source
The pool is the same size at a full-cap school and a small one, because it draws on outside deals and not on what the school pays. At 5% above $5,000 it is about a third of the money the rails recover from rejections, delays and non-payment. A star still ends the year ahead of today's process. Raising the floor moves the most money to walk-ons and Olympic-sport athletes; an equal share leaves the fewest athletes below today's process.
Whether a school may require contributions, how a pool sits with Title IX, the settlement's cap and the Commission's review of each deal, and how payouts are taxed are legal questions for counsel. Contributors here pay on deals already cleared and paid. The roster, the Olympic-sport count and each sport's share of school money are assumptions listed on the Method tab.
Published scenarios
Eight scenarios, 5,000 simulated athlete-years each, engine , generated . Simulation output, not measured results. Download results.json · Engine source
Net = school revenue share + outside-deal money the athlete actually receives after agent fees, hidden cuts and dispute costs. Leakage = rejected + lost to review delays + cleared but unpaid + hidden cuts + dispute costs.
Validation
Three checks on the model's own numbers: how much a result moves with luck, which assumptions it leans on, and how many athlete-years it takes to settle. They test the model against itself. They do not show that it matches real outcomes; the calibration check on the Results tab is the only comparison with published figures. . Download results-validation.json · Source
Monte Carlo: the same scenario on 200 different seeds
Each seed is a fresh set of random draws for 1,000 athlete-years. Gain is the athlete's average total on NIL33 rails minus the total under today's process.
Try it on whatever the simulator is set to now. Each press picks a new starting seed, so the figures will differ a little each time. That is the point.
Sensitivity: one assumption at a time
Convergence: how many athlete-years are enough
The published scenarios use 5,000 athlete-years each.
Pooled fund on 60 different seeds
Method
How the engine works, every setting it uses, and where each comes from. "Cited" means a public source listed on the Rules tab. "Assumption" means our estimate, stated so you can change it.
The engine draws the year's outside deals once: how many (by roster spot), who offers them (local business, national brand, collective, group licensing), size, whether the price is far above comparables, whether the deal is tied to an event date, and whether a middleman brought it. Both columns then process the identical deals under their own rules.
Complete paperwork at submission (signed, hashed agreement with deliverables), a fair-market check against comparables before submission, escrow paid on proof of the work, the agent's fee written into the agreement and published, and the introducer on the record. It does not change what the school pays, it does not speed up the CSC's own review, and an overpriced deal that is still submitted is still rejected.
Limits: one season, no transfers mid-year, no taxes, no Title IX allocation effects, no state-by-state rules beyond the $600 threshold and five-day reporting clock. The NIL33 column models what the rails are built to do; it is not a measured result from live athletes. Found a number that looks wrong? hello@nil33.com.