Backtest Explanation
The NCAAF model results are built to answer the question a serious customer would ask before trusting a betting process: if this system had been operating during the tested college football seasons, and a member bet every qualified play at $100 per bet, what would have happened?
Walk-Forward Method
A walk-forward backtest moves through the season in order. The model is not allowed to know the final standings, the final third-order rankings, or any information that happened after kickoff. Each game is treated like a live decision point: use the information available at the time, compare the model number to the betting market, apply the governor rules, and then grade the result after the game is played.
That matters because a normal after-the-season review can accidentally become too generous. It is easy to look back and say a team was better than its record, or that a market number was wrong, once the whole season is finished. A walk-forward test is stricter. It asks whether the process could have found value before the result was known.
The NCAAF ROI is not based on opening lines. The active customer card is graded from historical pregame decision-time sportsbook prices, then compared against the close snapshot for CLV. The card shown to customers combines the ATS lane with the Hybrid Best totals lane because that pairing improved volume and ROI without forcing every model lean onto the card.
For customers, that is the most honest way to describe the test. The page is not saying, "we found winners after the fact." It is saying, "under these rules, using only pregame information, this is what the betting card would have looked like and this is how it would have performed."
What The Model Compared
The model's job is to create a fair number. For full-game point spreads and first-half point spreads, that means estimating what the spread should be. For totals, that means estimating the fair game total or first-half total and comparing that number to the sportsbook total.
Portfolio Governor
The portfolio governor is the filter that turns model opinions into customer-facing bets. This is important because a model can have an opinion on nearly every game, but a betting product should not force customers into every tiny lean. The governor is designed to protect the card by requiring enough edge, market support, football context, and the right market fit before a play qualifies.
The model must show a meaningful edge compared with the sportsbook spread or total.
The play must satisfy the market filter, so the card is not built only on isolated model disagreement.
The matchup still needs enough football support and segment fit to avoid weak or fragile edges.
In plain English: the model finds the possible value, then the Three Check process and portfolio governor decide whether that value is strong enough to become a bet.
Customer Card
The active customer card combines two lanes: the ATS side lane and the Hybrid Best totals lane. The ATS lane includes full-game and first-half spread signals. The totals lane adds a smaller but stronger totals group that improved the overall card from +17.3% ROI on ATS alone to +21.4% ROI combined.
The card is built this way because it adds quality volume without turning the product into an every-game prediction sheet. Each bet still needs model value, market confirmation, and football context before it qualifies.
Customer Framing
The results are shown the way a customer would naturally think about them. If a member risked $100 on every qualified NCAAF play, the results page shows how many bets qualified, how many units were earned, and how much money that equals at $100 per bet.
The goal is to make the record understandable, transparent, and useful before a customer joins: not just "the model was good," but how good, over how many bets, in which markets, and with what financial return.
Important Note
The full model research includes older seasons and other bet types, but the public customer card uses 2024-2025 because that is the shared window where the ATS and Hybrid Best totals results can be compared cleanly with the current decision-pricing and CLV audit structure.
Important Note
The NCAAF backtest is evidence of how the model and portfolio governor would have performed under the stated rules. It is not a guarantee of future profit. Sports markets change, injuries matter, pricing can move, and future seasons will not unfold exactly like the test window. The value of the page is transparency: customers can see the method, the rules, the bet count, the wins and losses, and the historical financial result.
The next validation step is tracking the same rules live during the 2026 season, including official posted-card prices, closing-line value, and settled results.