Learn
Master the fundamentals of sports betting and analytics.
📖 Betting 101
Spreads, moneylines, totals, odds formats, and implied probability — explained from scratch.
Build your first model
Spin up a no-code model in the browser builder. No Python install, no server upload — weights stay on your device.
Workshop: fork a model recipe
A fork copies the model recipe (architecture and feature choices), not trained weights or private data. Retrain and backtest the fork on your own data before publishing.
Track CLV like a sharp
Closing line value tells you whether your bets are sharp before the result does. The leaderboards rank every public model on the platform.
Feature glossary
EPA, success rate, pace, Elo, target share, CLV, and more — what each number means in plain words.
FAQ
Where should I start if I have never bet before?
Open Betting 101 to lock in the vocabulary (spread, total, moneyline, implied probability), then read the bankroll-management piece in /guides. After that, /picks shows you the kind of plays the platform actually recommends.
Do I need to write code to use the modeling tools?
No. Studio and Workshop are both drag-and-drop or one-click. Model weights are JSON files you can download. Python is only useful if you want to extend the platform itself.
How long until I can trust my own model?
There is no fixed bet count that makes a model reliable. Test it on games it did not train on. Compare predicted chances with actual results, track closing line value by market, and account for uncertainty before drawing conclusions.
What is the difference between Learn and Guides?
Learn is the short-form on-ramp — single concepts with a calculator or surface attached. Guides at /guides goes deeper with multi-section tutorials, evidence ledgers, and links into the data library.
Can I share my progress with friends?
Yes. Every public model on /leaderboards has a shareable card with rolling CLV, Brier, and ATS. Share that card with friends to compare results.