BetUpset

BetUpset is a cross-platform soccer double-chance (1X, 12, X2) Dutching scanner with a web dashboard for reviewing opportunities, placing bets, and tracking outcomes across Polymarket and Kalshi.

What It Does

BetUpset scans prediction markets on Polymarket and Kalshi, groups equivalent events, then looks for three-outcome soccer markets where the cheapest outcome can be rejected and the remaining two outcomes can be covered across platforms at a combined cost below 1.0.

The core strategy is a selective Dutching setup rather than pure three-way arbitrage: the app rejects the least attractive outcome and covers the other two when pricing and safety filters are favorable.

For each surfaced opportunity the app calculates:

• best available prices per outcome
• rejected outcome and reject price
• ROI if either covered outcome wins
• score and win probability estimates
• per-platform stake split and liquidity information

Setup

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Create a local .env file from .env.example and fill in your credentials and key paths.

Main Entry Points

Web app

python web.py
python web.py --demo

The web app exposes:

• GET / — frontend dashboard
• GET /api/scan and POST /api/scan — scan results
• GET /api/bets and POST /api/bets — tracked bets
• POST /api/bets/execute — guarded live execution
• GET /api/balances — platform balances
• POST /api/resolve — bet auto-resolution

CLI monitor

python main.py
python main.py --once
python main.py --demo

Continuously scans, detects Dutching opportunities, records them, and emits alerts. Use --once for a single pass, or --demo with synthetic data.

Configuration

All runtime settings live in config.yaml. The config loader (config.py) substitutes ${ENV_VAR} references with actual environment values and falls back to sensible defaults if the file is missing.

Strategy

Controls when the scanner considers an opportunity worth surfacing.

Key	Default	Description
strategy.min_gap	0.03	Minimum price gap (3 cents) to consider a trade
strategy.max_reject_prob	0.25	Never reject an outcome above 25% implied probability
strategy.bet_fraction	0.10	Risk 10% of current bankroll per trade

Risk Limits

Hard caps that prevent over-exposure regardless of what the strategy suggests.

Key	Default	Description
risk.max_exposure_per_match	50	Max USD on any single match
risk.max_total_exposure	3000	Max USD across all open trades
risk.max_matchday_exposure_pct	0.15	Max 15% of bankroll on any single matchday

Bankroll

Key	Default	Description
bankroll.starting	10000	Starting bankroll in USD used for position sizing

Scanner

Key	Default	Description
scanner.interval_seconds	60	Seconds between scan cycles
scanner.leagues	See below	List of league slugs to scan

Default leagues: premier-league, la-liga, champions-league, serie-a, bundesliga, ligue-1.

Execution Safeguards

The web execution path enforces these checks server-side so execution safety does not depend on browser-side controls alone.

Key	Default	Description
execution.dry_run_only	false	Block real order placement, keep simulation mode only
execution.max_stake_per_trade	50	Hard cap (USD) on any single execution attempt
execution.max_scan_age_seconds	600	Reject execution if the scan snapshot is older than this
execution.max_liquidity_fraction	0.05	Max fraction of covered liquidity allowed per platform

In addition, live per-platform balance verification runs before any order is placed.

Platforms

Each platform has an enabled flag and credential fields that reference environment variables.

platforms:
  polymarket:
    enabled: true
    private_key_path: "${POLYMARKET_PEM_PATH}"
  kalshi:
    enabled: true
    api_key_id: "${KALSHI_API_KEY}"
    private_key_path: "${KALSHI_PEM_PATH}"

Alerts

Key	Default	Description
alerts.console	true	Print alerts to stdout
alerts.telegram_bot_token	—	Telegram bot token for push notifications
alerts.telegram_chat_id	—	Telegram chat ID to send alerts to

Output

Key	Default	Description
output.csv_path	opportunities.csv	CSV file for logged opportunities
output.db_path	trades.db	SQLite database for trade tracking

Environment Variables

BetUpset expects credentials to come from the environment rather than from committed source files.

KALSHI_API_KEY=your-kalshi-api-key
KALSHI_PEM_PATH=kalshi_bet.pem
POLYMARKET_API_KEY=your-polymarket-api-key
POLYMARKET_PEM_PATH=polymarket.pem
APP_PASSWORD=
APP_SESSION_SECRET=change-me-if-you-enable-password-auth
TELEGRAM_TOKEN=              # optional, for alert notifications
TELEGRAM_CHAT_ID=            # optional, for alert notifications

If APP_PASSWORD is set, the web dashboard requires that password before it will load scans, bets, balances, or resolution actions. The login state is stored in a signed session cookie.

Project Structure

.
├── web.py                 # FastAPI backend and web entrypoint
├── static/index.html      # Single-page frontend
├── scan_service.py        # Scanning logic
├── main.py                # CLI monitor loop
├── scanner.py             # Platform scanning pipeline
├── matching.py            # Cross-platform match grouping
├── detector.py            # Arbitrage detection
├── risk.py                # Risk checks and stake rules
├── tracker.py             # SQLite and CSV-backed tracking
├── polymarket_client.py   # Polymarket integration
├── kalshi_client.py       # Kalshi integration
├── config.py              # Typed config loader and defaults
├── config.yaml            # Runtime configuration
└── tests/                 # Unit tests

Testing

pytest tests -v