Betting Model Historical Records -- All Sports, All Season

Error loading data

Model accuracy by sport -- ML, spread, fave/dog, and O/U going back to 2019

> Every prediction logged, every result published -- wins, losses, and pushes, by sport, season, and model version. The table shows season-by-season ML, spread, fave ATS, dog ATS, and totals records for MLB (since 2019), NBA (since 2019-20), NHL (since 2019-20), NCAAB (since 2019-20), NFL (since 2019), NCAAF (since 2019), WNBA (since 2021), and CFL (since 2022). V1 and V2 (and V3 where active) are shown side-by-side. Green arrows mean improvement over the prior year. Red means decline. If you want to know whether this is worth following before committing to anything, this is the page to start.

↑ Green = improved vs prior year ↓ Red = declined → Amber = flat (±1.5%) · Blue/Pink = first year of model

YEAR

MODEL

SPREAD

FAVE ATS

DOG ATS

TOTALS

2026

3333-3333-100

I'm a paragraph. Click here to add your own text and edit me. It's easy.

2026

MODEL

3333-3333-100

I'm a paragraph. Click here to add your own text and edit me. It's easy.

2026

MODEL

3333-3333-100

Nothing Hidden, Nothing Cherry-Picked

Most betting sites publish records selectively -- the good years, the profitable angles, the winning streaks. We publish everything. The 2021 MLB ML was 54.8%. The 2020 NHL ML was 52.7%. There were season where totals ran flat across multiple sports. The NBA spread market has been roughly 50% most years, and we don't inflate that number.

Publishing the full record -- the hard season alongside the strong ones -- is what makes the improvement trend meaningful. When the MLB V1 ML moved from 54.8% in 2021 to 57.1% in 2024 to 58.7% spread in 2025, that context exists because we showed you every year in between.

What Seven Years Of Data Looks Like

MLB and NFL are both in their seventh season of data. NHL, NCAAB, NBA, NCAAF go back to 2019-20. Those longer data sets show something more valuable than any single strong season: consistency. MLB V1 ML has been 54.6% and 57.1% in every full season since 2019. NHL V1 dog ATS has been above 56.7% every year since 2020, topping 61% in four of six seasons. NCAAF V1 ML has been above 68% in every season since 2019. These aren't outlier years -- they're patterns that have held across different team compositions, different market conditions, and different competitive environments.

How The Models Improve Over Time

Each off-season, the models are recalibrated using the full prior dataset. More season of historical data mean tighter expected probability calculations and more reliable edge thresholds. New sports enter the system V1 first -- broadest selection, first full season of baseline data. Once there's enough history to test tighter confidence buckets, beta models follow: WNBA launched its beta models for 2026. The same expansion is planned for NFL, NBA and CFL heading into their upcoming seasons. The improvement curve here is real, it's documented year by year, and it's still going.