Guide
How the automated repricing engine works, how rolling snapshots refresh (~30 min when new games ingest), and how paper trading fits in. Educational - not financial advice or real securities.
At a glance
You're paper-trading simulated shares tied to player value. Prices blend anchors, history, and each ingestion cycle; the board refreshes automatically (~30 min) when new games arrive - a rolling snapshot, not a live exchange feed.
How a model price comes together
Each listing starts from an opening anchor and prior-season productivity. That keeps prices from whipsawing on tiny samples early in the year.
As regular-season and playoff games enter each ingestion cycle (~30 min cadence), performance feeds the automated repricing engine. The price is a smoothed signal - it carries season context, not just the last box score.
The number you see is the model’s latest quote for that player. It updates when that player logs another game and otherwise stays flat until the next one.
Where to click in the app
On a player page, game score is that night's Hollinger-style summary when a game was ingested. Model price can stay elevated after a soft night because the path carries prior games and season structure - use chart ranges and season averages to separate blips from trends. Flat stretches mean no new game data in that interval.
Practical tips
Paper trading, simply
You start with paper cash. Orders are whole shares at the latest model price shown on that player - there's no bid/ask spread or slippage simulation. Your portfolio rolls up cash, mark-to-model position value, and total equity so you can rehearse allocation ideas on the same timeline as your ingested games.
Lot size
Whole shares only
Fill price
Latest model quote on submit
Balance
Cash + positions = equity
A warning means this player has no minutes logged in 2025-26 in your dataset file - injury, inactive, or data lag. The price may still reflect older seasons or the model anchor; treat it as stale relative to current floor time until your file catches up.
Terms you'll see repeated
Simulation only - paper currency and model-driven prices. Nothing here is investment advice, a prediction of real athletic or financial outcomes, or an offer of securities. If your pipeline or dataset changes, replay history and labels change with it.
