Wild Odds Tracker

A Node.js application that automatically tracks Minnesota Wild moneyline odds, detects sharp movement, and records results — so you can measure your Closing Line Value (CLV) and Expected Value (EV) over time.

Quick start

cp .env.example .env
# Fill in your API keys (see Environment Variables section below)
npm install
npm start

The server starts on http://localhost:3000. On startup it immediately syncs the Wild schedule from SportRadar, then runs all three cron jobs on their normal schedules.

Project structure

wild-tracker/
├── index.js                 Express app + REST API entry point
├── src/
│   ├── calculator.js        Pure math: implied prob, CLV, EV, de-juice
│   ├── db.js                SQLite schema + all query helpers
│   ├── mailer.js            Nodemailer Gmail alert composer
│   ├── metrics.js           CLV/EV orchestration per bookmaker
│   ├── odds-fetcher.js      The Odds API (DraftKings, FanDuel, BetMGM, Pinnacle)
│   ├── results-fetcher.js   SportRadar: schedule, results, context, standings
│   └── scheduler.js         node-cron jobs
├── data/                    SQLite database (git-ignored)
├── .env.example
└── test.js                  Smoke tests (npm test)

Environment variables

Copy .env.example to .env and fill in the values.

Variable	Required	Description
SPORTRADAR_API_KEY	✅	SportRadar NHL API key
ODDS_API_KEY	✅	The Odds API key
ALERT_EMAIL	optional	Address that receives email alerts
GMAIL_USER	optional	Gmail address used as sender
GMAIL_APP_PASSWORD	optional	Gmail App Password (16 chars, not your regular password)
PORT	optional	Express port — default 3000
DB_PATH	optional	Absolute path to SQLite file — default ./data/wild.db
CRON_TZ	optional	Timezone for cron schedules — default America/Chicago

Email alerts are silently disabled if GMAIL_USER, GMAIL_APP_PASSWORD, or ALERT_EMAIL are absent.

Database schema

Table	Purpose
games	One row per Wild game; schedule, scores, context, result, and user's my_moneyline estimate
odds_snapshots	Every poll per bookmaker, tagged opening / hourly / closing
game_metrics	Per-bookmaker CLV, EV, opening/closing lines; upserted on settle
sharp_move_events	One row per detected ≥10-cent line move; tracks alert status
model_predictions	User/model win-probability estimate per game (0–1), used for calibration

Context columns on games

Column	Description
wild_days_rest	Days since Wild's previous game
opp_days_rest	Days since opponent's previous game
opp_wins / opp_losses / opp_ot_losses	Opponent's current season record
opp_last10_wins / opp_last10_losses	Opponent's last-10 record

REST API

Game data

Method	Route	Description
GET	/games	All Wild games (newest first) with per-book metrics
GET	/games/:id/snapshots	Full odds-snapshot history
GET	/games/:id/sharp-moves	All sharp-move events for one game
POST	/games/:id/estimate	Set your opening moneyline estimate { "moneyline": -130 }
POST	/games/:id/predict	Store a win-probability estimate { "my_prob": 0.62 }

Analytics

Method	Route	Description
GET	/stats	Win rate, avg CLV per book, avg EV, ROI
GET	/stats/calibration	All (my_prob, outcome) pairs for calibration curve plotting

Manual triggers

Method	Route	Description
POST	/sync	Re-sync Wild schedule from SportRadar
POST	/poll-odds	Capture an odds snapshot right now
POST	/settle	Try to settle all outstanding games

Cron schedule (Central Time)

Job	Schedule	What
syncSchedule	06:00 daily + startup	Upserts Wild games + contextual data from SportRadar
pollOdds	Top of every hour	Captures opening/hourly/closing snapshots; detects sharp moves
settleGames	Every 30 minutes	Fetches final scores; calculates CLV/EV; sends settle email

Metrics explained

Implied probability

De-juiced win probability derived from both sides of the market (vig removed).

$$\text{prob}_i = \frac{\text{raw}_i}{\text{raw}_1 + \text{raw}_2}$$

CLV (Closing Line Value)

Measures whether closing_prob validated your opening estimate.

$$\text{CLV%} = \left(\frac{p_\text{close}}{p_\text{mine}} - 1\right) \times 100$$

Positive CLV → market moved to confirm your view.

EV (Expected Value)

Expected return per unit at the opening line, given your probability.

$$\text{EV%} = \bigl(p_\text{mine} \times d_\text{open} - 1\bigr) \times 100$$

where $d_\text{open}$ is the decimal equivalent of the Wild's opening American odds.

Sharp move detection

Any line movement of ≥ 10 American-odds points between consecutive snapshots triggers a sharp_move_events row and an email alert.

ROI

Assumes 1 unit wagered per game at the Pinnacle closing line (falls back to DraftKings).

Bookmakers tracked

• Pinnacle (sharpest market, CLV reference)
• DraftKings
• FanDuel
• BetMGM