Edge Spectrum — a growing collection of interactive tools for sports betting, gambling, investing, and odds analysis. Includes the Edge Spectrum expected-return visualization and a 25-year sports bett
翻譯準備中
金融HTML · TypeScriptGitHub @ai-automation-tools上線中營運者未確認
試用過了嗎? 點一下就好,不需要登入。
尚未偵測到更新。GitHub 發布版本後會顯示在這裡。
Interactive demo and full self-hosted toolkit for expected returns, synthetic backtesting, and odds analysis.
翻譯準備中
正在載入完整 README。 以下為已儲存的摘錄。
A unified multi-tool hub for sports betting, investing, gambling, and odds analysis. Compare expected returns across time horizons, backtest sports wagering strategies over 26 simulated seasons, and analyze mathematical edges.
Edge Spectrum is an interactive playground for exploring mathematical edges. It brings together four analytical applications:檢視完整 README ↗Edge Spectrum is an interactive playground for exploring mathematical edges. It brings together four analytical applications:
/spectrum/): Compares 240 financial, investing, trading, and gambling activities over 7 time horizons (Single Event to 10 Years), standardized using the DU/CED framework. Return on capital is the primary axis and is floored at −100%; the unbounded turnover cost gets its own axis so the two are never compared as if they were the same quantity./backtester): Backtests sports wagering strategies across 26 simulated seasons of MLB, NFL, NHL, and NBA games (2000–2025), providing ROI diagnostics, equity curves, live ESPN scores, and an AI Strategy Advisor powered by Gemini./odds): Converts American, decimal, fractional and implied odds, and strips bookmaker margin from any multi-way market via multiplicative, power, or Shin de-vigging./parlay): De-vigs each leg, multiplies out the true joint probability, and compares it with the book's parlay payout to expose how the margin compounds per leg.Served as a static page at /spectrum/index.html (internally routes to public/spectrum/index.html).
data/edge_dataset.md lists every record by category.Served inside the SPA React shell at /backtester.
/backtester?sport=NBA&betType=spread&side=home&…), so any run can be bookmarked or sent with Copy link; a stale or hand-edited link falls back to defaults field by field instead of erroring.A collector that builds your own archive of real sportsbook lines for closing-line-value analysis
(roadmap Actions 5.1 and 5.2). Each run snapshots ESPN's public scoreboard for NFL, NBA, MLB and NHL
and appends one JSON line per game to <folder>/<SPORT>/<day>.jsonl: spread and its prices, total,
moneyline, scores, and a linePhase of open, midday or close. Files are only ever appended to,
there are no keys, and nothing leaves your machine.
[!NOTE] The collector and the live scoreboard read an unofficial, undocumented ESPN endpoint. Edge Spectrum is not affiliated with ESPN. If you run the collector you are responsible for following ESPN's terms of use: keep it to personal, low-volume use (the default is one request per league every three hours) and do not republish what you capture. This project does not publish captured lines.
cd site && npm run capture:lines -- --out ~/edge-lines
以上為 README 的開頭部分。其餘內容請閱讀原文。 檢視完整 README ↗