NCAA Division I men's basketball · Annual analytics license
Fixed-criteria college basketball analytics. Every result published.
Teams and alternate lines selected by fixed statistical criteria across every Division I men's game, published before tip, delivered by email and in the app, and graded against the final score in a public record.
Every qualifying game, all season. Nothing is skipped and nothing is added by judgment, ours or anyone's.
Standard $999 · Elite $1,999 · one year from the day of purchase, never renewed automatically.
- Positive seasons
- 10 of 10
never a losing season
- 2025-26 combined
- +195.2
singles +103.3, parlays +91.9
- 2025-26 max drawdown
- -11.2
worst day -4.9
- Provenance
- Every season is derived from raw game data by the same fixed criteria. How each season was produced and verified ↓
Worst season drawdown across all ten seasons: -28.2 units. Worst single day: -9.5 units.
Past performance does not guarantee future results.
2025-26, month by monthThe 2025-26 record
The 2025-26 record, day by day with the season to date on every line, as of Apr 6, 2026.
What the record establishes
Ten consecutive positive seasons, 2016-17 through 2025-26, applying the same fixed statistical criteria to every season.
Worst season drawdown across all ten seasons: -28.2 units. Worst single day: -9.5 units.
Past performance does not guarantee future results.
Cumulative units, 2016-17 through 2025-26
As of Apr 6, 2026
The steep steps come from the parlay card: its units are uneven by construction, and a small number of days carry a large share of them. The three largest days in ten seasons produced +1,192.1 units between them. The singles, which Standard receives, are the smoother half: in every season their drawdown is shallower than the combined record's. Why some days are much larger →
Source: the QuantSight results ledger. Data through Apr 6, 2026. The x axis counts days with qualifying picks; offseasons are not drawn.
| Season | Singles | Parlays | Combined | Drawdown |
|---|---|---|---|---|
| 2016-17 | +164.0 | +546.5 | +710.5 | -9.9 |
| 2017-18 | +125.6 | +23.5 | +149.1 | -14.7 |
| 2018-19 | +105.0 | +270.0 | +375.0 | -28.2 |
| 2019-20 | +147.2 | +133.7 | +280.9 | -12.6 |
| 2020-21 * | +103.8 | +18.1 | +121.9 | -8.8 |
| 2021-22 | +149.4 | +215.3 | +364.7 | -8.4 |
| 2022-23 | +92.4 | +571.3 | +663.7 | -11.8 |
| 2023-24 | +97.0 | +59.5 | +156.5 | -16.5 |
| 2024-25 | +81.6 | +16.4 | +98.0 | -24.8 |
| 2025-26 | +103.3 | +91.9 | +195.2 | -11.2 |
Drawdown is the season's largest fall from peak, in units, measured on the combined record; the singles-only drawdown for each season is on the results page. Standard receives the singles only. * 2020-21 was shortened by COVID.
2016-17 through 2024-25 was produced by re-running the criteria over historical game data. That re-run covered 52,840 games, re-run by us from raw game data and checked season by season. 2025-26 ran in production and is published as a complete record. Every published day is derived from raw game data by the same fixed criteria as the historical seasons. The production system that ran during the season recorded 98 days in full from December 4 through March 15, and this record matches every one of them exactly. The days before December 4 and the 14 tournament days after March 15 rest on the engine and the game data alone. From 2026-27, every published position carries a publication timestamp taken before tip.
Source: the QuantSight results ledger. Data through Apr 6, 2026.
One day, publication to settlement
Illustrative. Not a published day.
An invented day, rendered the way a license holder sees it. It is a losing day on purpose: the record publishes those at the same size as the wins, and a licensee sees them the same way.
01Published
Jan 18, 2016, 10:02 AM EST
Positions publish in the morning, Eastern, by 10:15 AM ET. The timestamp is the server's and cannot be edited afterwards.
02Delivered
Positions as delivered Side Line Fordhamvs Dayton +12.5 Coppin Statevs Norfolk State +15.5 Presbyterianvs High Point +17.5 On the picks page and by email, 1u on each. The email →
03Graded
Positions after the final scores Side Final FordhamDid not cover 58-79 Coppin StateDid not cover 61-84 PresbyterianDid not cover 66-91 - Ledger row
- Jan 18, 2016
- Day
- -3.0
Each position that did not cover settles at -1.0u, the stake and nothing else. The row is appended to the public ledger and never edited.
What arrives each morning
The Elite view of a day, rendered by the product's own components. Standard receives the singles without the card.
Elite slate · Dec 27, 2015
Illustrative
Published Dec 27, 2015, 9:58 AM EST
| Side | Alt spread |
|---|---|
| Bryantvs Providence | +11.5 |
| Longwood†vs Virginia | +14.5 |
| Queens†vs Wake Forest | +19.5 |
| Stetsonvs Florida | +25.5 |
| Sacred Heart†vs UConn | +29.5 |
- Positions
- 51u each
- Day exposure
- 6u
† A leg of the parlay card: 3 legs, 1u, all must cover.
Illustrative slate. Not a published day.
What following it takes
The figures below are an illustration at a $100 unit, chosen so the arithmetic can be checked. They are not a sizing instruction.
| Item | At a $100 unit | |
|---|---|---|
| One unitthe stake on every position | $100 | the stake on every position |
| A 40-position dayone unit on each; an illustration, not a typical day | $4,000 staked | one unit on each; an illustration, not a typical day |
| The busiest day on record, 66 unitsthe maximum across all ten seasons | $6,600 staked | the maximum across all ten seasons |
| Capital the sizing assumption impliesat 0.25 to 0.5 percent of dedicated capital per unit | $20,000 to $40,000 | at 0.25 to 0.5 percent of dedicated capital per unit |
| Elite license, one yearabout 20 units | $1,999 | about 20 units |
| Standard license, one yearabout 10 units | $999 | about 10 units |
Staked is not lost, and neither figure is capital. A position that does not cover settles at one unit. The 66-unit day is the largest the record holds: an observation, not a ceiling.
- Workload
- In 2025-26 the criteria produced 1,704 positions on 142 days, about 12 on an average qualifying day. Every position is placed by the license holder, at the book they use.
- Arrival
- Mornings by 10:15 AM ET, by email and on the picks page. A delay is announced. Nothing is added after the fact.
- Line moves
- Every position is graded at the published line. If the book no longer offers that line, the license holder takes what is offered or passes. A season followed in part is not comparable to the record.
Following the system takes real capital and real discipline. The record is built one way: one unit on every qualifying game, every day of the season, no breaks, no exceptions. The system has required as many as 66 units of exposure in a single day. Sizing assumes a unit is a small fixed fraction of capital dedicated to this purpose, on the order of 0.25 to 0.5 percent. If a 66-unit day is more than you can hold to without deviating, this license is not for you.
Worst season drawdown across all ten seasons: -28.2 units. Worst single day: -9.5 units.
Longest run of consecutive losing days across all ten seasons: 10 (2018-19) on the combined record, 7 (2016-17) on the singles alone. In 2025-26, 61 of 142 graded days finished negative on the combined record, 45 on the singles alone.
Past performance does not guarantee future results.
How we are paid
QuantSight holds no sportsbook affiliate relationships and takes no sportsbook revenue. We are paid license fees, by subscribers, and by nobody else. We do not accept or process bets.
Every question before a purchase is answered personally within one business day, at support@quantsightanalytics.com or through the contact form.
Licensing
Standard
$999per year
Plus applicable sales tax.
- Every qualifying single, published before tip, all season.
- The daily email.
- No seat cap.
Elite
$1,999per year
Plus applicable sales tax.
- Everything in Standard.
- The parlay card, on days the slate produces one.
- Initial release: 40 licenses. Total cap: 100.
- Term
- One year from the day of purchase. Bought Sep 15, 2026, access through Sep 15, 2027. Never renewed automatically; we email before the year ends, and you decide again.
- Refund
- A license is refundable in full until the first pick is delivered to your account or 14 days after purchase, whichever comes first. After that, the sale is final. If you buy between seasons, the 14 days count from the next season's first slate instead of from purchase. A refund returns money to the original payment method, never credit. The cutoff exists because a delivered pick cannot be returned. Terms.
A free account is optional. It reads the same public record and nothing more. Create a free account →
Disclosures
- 1Every slate, score, and ledger row in the sample sections is invented for illustration and rendered by the product's own components. No sample date appears in the public record.
- 2Dollar figures in the fit table are an illustration at a $100 unit. They are not a recommendation, a sizing instruction, or a statement of results in currency.
- 3Past performance does not guarantee future results. All figures are units, not currency; the price basis of the record is stated on the methodology page.