About BBMI Hoops
The Benchmark Basketball Model Index is a college and high school basketball analytics project built on the same principles used in professional actuarial forecasting — and documented publicly from day one.
† Record includes only picks where BBMI and Vegas lines differ by ≥ 2 points (1,919 of 2,927 completed games). The Vegas line used by this model is captured at a specific point in time — lines routinely move 1–2 points between open and tip-off, and can vary by a point or more across different books. A difference smaller than 2 points is therefore within normal market noise and does not represent a meaningful BBMI disagreement with Vegas. View full public log →
It started with a family NCAA bracket challenge. I built a quick model to get an edge, the model worked better than expected, and I got nerd-sniped into something more serious. What began as a fun experiment became a genuine forecasting project — one that now covers 1,919+ documented NCAA games and an entire WIAA high school basketball season.
I've spent decades as an actuary building predictive models for healthcare costs and revenue forecasting. The core disciplines — data quality, variable selection, calibration, and out-of-sample validation — translate surprisingly well to sports. Once I noticed the model's projected game lines were consistently closer to actual outcomes than several publicly available Vegas models, the logical next step was to track it rigorously and see if the edge was real.
The goal has always been simple: publish the picks before the games, record every result publicly, and let the cumulative record speak for itself. No cherry-picking. No retroactive adjustments. If the model is good, the numbers will show it over time.
The BBMI generates its own predicted point spread for every game — independently of what Vegas has set. The gap between the BBMI line and the Vegas line is what we call the "edge." The bigger the edge, the more strongly the model disagrees with the sportsbooks.
When the model strongly disagrees with Vegas, it's typically because it's detected something the market hasn't fully priced in — an efficiency gap, a strength-of-schedule discrepancy, or a situational factor. These are the picks worth paying attention to.
The Vegas line used in this model is captured at a specific point in time. Lines routinely move 1–2 points between open and tip-off, and can vary by a point or more across different books. A difference smaller than 2 points is therefore within normal market noise — it's more likely explained by line movement or book-to-book variation than a genuine model disagreement with the market. Only picks with edge ≥ 2 pts are counted in the performance record.
Team strength is evaluated using a blend of scoring efficiency, opponent quality, historical performance, and situational factors. These inputs are weighted and transformed into a projected spread and win probability for each matchup.
Rather than relying on any single metric, the model uses a layered approach — each component contributes a small but meaningful signal. The goal isn't perfection on any one game. It's consistent, repeatable accuracy across a large sample.
The WIAA model applies the same framework to high school basketball, with the acknowledgment that self-reported team statistics introduce more noise. The model is directionally useful but naturally less precise than its NCAA counterpart.
Every pick BBMI has ever made is logged publicly at ncaa-model-picks-history. Wins, losses, dates, spreads, simulated returns — all of it, from the first pick of the season, unedited.
This approach is borrowed directly from actuarial practice: a model that can't be validated against out-of-sample data isn't worth trusting. The public log isn't a marketing tactic — it's the only honest way to evaluate whether the model actually works.
Sports betting is one of the few markets where an informed individual can hold a structural advantage over the house — not because Vegas is bad at math, but because Vegas is solving a different problem than you are.
Vegas has to set a precise number — the line. You only have to decide which side of it to be on. If the true spread is -6.2 and the book posts -5.5, you don't need to know it's exactly -6.2. You just need to recognize it's more than -5.5. That's a fundamentally easier problem.
What Vegas is actually optimizing for
Sportsbooks aren't purely trying to predict the correct outcome — they're managing liability. When 80% of public money lands on one side, the book will shade the line to attract action on the other side, even if their internal model says the original number was right. A model focused purely on accuracy — not risk management — can exploit those gaps.
Where BBMI fits in
BBMI doesn't need to be smarter than the sportsbook's internal model. It needs to be smarter than the posted line — the number that's already been distorted by public money, liability balancing, and market incentives. That's a lower bar, and college basketball is one of the best places to clear it: 360+ teams, thin markets, less sharp money, and more modeling opportunity than the pros.
- Only need to pick a side, not set a number
- No liability to manage — pure accuracy focus
- College basketball is inefficient and data-rich
- Public money distortions create exploitable gaps
- The juice (-110) means you need ~52.4% to break even
- Vegas has faster injury and lineup data
- Sharp bettors move closing lines toward "true"
- Variance is real — even good models have losing weeks
No model wins every bet. The goal is to clear the 52.4% breakeven threshold consistently over hundreds of games — and to bet more when the edge is largest. BBMI's documented record on high-edge picks shows this is achievable, but it requires discipline, patience, and realistic expectations. If someone promises you 70%+ win rates on every pick, they're selling something other than math.
The sports betting information industry is full of services selling picks with no verifiable track record. BBMI was built specifically to be the opposite of that.
| Aspect | BBMI | Typical Tout |
|---|---|---|
| Track record | ✓Public, unedited, full history | ✗Cherry-picked wins, no losses shown |
| Methodology | ✓Documented actuarial approach | ✗Vague claims, no explanation |
| Confidence tiers | ✓Edge scores show conviction level | ✗Everything is a 'lock' |
| Performance filter | ✓Excludes line-movement noise (edge < 2 pts) | ✗Counts everything, including coin flips |
| Bad weeks | ✓Logged and visible | ✗Quietly buried |
| Pricing | ✓$15 trial / $49 monthly | ✗$99–$299+ per month |
| Background | ✓Professional actuary | ✗Unknown / unverifiable |
The honest version of our pitch: the model has a documented 56.9% record on picks where BBMI meaningfully disagrees with Vegas (edge ≥ 2 pts), and 63.3% on high-edge picks — across 1,919+ games. That's real, verifiable, and not perfect. We'd rather you evaluate the actual record than take our word for it.
Major model updates are logged here as they happen. Because picks are frozen before games tip off, any methodology change only affects future picks — never historical results.
Future updates will be logged here as they are deployed. Version history is permanent and will not be removed.
See the record for yourself
Every pick logged publicly. Filter by edge size. Judge it yourself.