User Guide for Variance Calculator

This is the user guide for the tool here – https://bestforguide.com/sports-betting-variance-simulator-free-tool/

What it does

  • Simulates bankroll outcomes for a sequence of bets priced in American odds (+150, -120).
  • Uses your EV% edge and odds to derive the win probability and run Monte Carlo simulations.
  • Visualizes variance: final bankroll distribution, sample bankroll path, risk of ruin, drawdowns.
  • Supports fixed, percent, and Kelly-based bet sizing.
  • Optional commission on wins.

Key concepts

  • American odds: Enter +X for underdogs, -X for favorites (e.g., +150, -120).
  • EV% (edge): Expected profit per bet as a percent of stake.
    • EV% formula: EV% = d_eff × p − 1, where d_eff is decimal odds after commission, p is your true win probability.
    • The simulator derives p from your EV% and d_eff: p = (1 + EV%)/d_eff.
  • Commission: Reduces payouts on wins; enter as a percent.

How to use

  1. Inputs
  • Starting bankroll: Your initial capital.
  • Number of bets: Bets per simulated run (season).
  • Simulations: How many runs to simulate. More = smoother charts, slower compute.
  • American odds: Entry price, like -110 or +150.
  • EV% (edge): Your estimated edge per bet (e.g., 2.0 for 2%).
  • Commission/Vig %: Optional haircut on winnings.
  • Bet sizing:
    • Fixed stake: Same dollar amount each bet.
    • Percent of bankroll: Stake = X% of current bankroll (compounding, higher volatility).
    • Kelly fraction: Stake = user_fraction × Kelly*, with Kelly* = (b·p − (1−p))/b and b = d_eff − 1.
    • Kelly fraction f: Only used in Kelly mode (e.g., 0.5 = half-Kelly).
  • Random seed: Optional; set for reproducible results.
  1. Run
  • Click “Run simulation.” The app draws one sample bankroll path and then runs all simulations in chunks.
  • “Stop” cancels a run in progress.

Outputs and charts

  • Expected profit (theoretical): Approx nBets × initial stake × EV%.
    • Note: For percent/Kelly staking, actual results compound; this KPI is a baseline.
  • Median final bankroll: 50th percentile of final bankroll across sims.
  • Risk of ruin: Fraction of sims where bankroll hit zero before completing all bets.
  • Avg. max drawdown: Average worst peak-to-trough decline per sim.
  • P(ending down): Probability your final bankroll is below the starting bankroll.
  • Final bankroll distribution: Histogram of ending bankroll across sims. A red dashed line marks the starting bankroll.
  • Sample bankroll path: One simulated trajectory showing swings over your n bets.

Bet sizing details

  • Fixed stake: Lower volatility, easier to interpret expected profit.
  • Percent stake: Scales with bankroll; can reduce risk as bankroll falls but increases variance when bankroll is large.
  • Kelly mode:
    • The app computes the Kelly fraction using your implied p and b = d_eff − 1, then multiplies by your “Kelly fraction f.”
    • Full Kelly (f=1) maximizes long-term growth but can be very volatile. Many use half-Kelly (f≈0.5).

Tips

  • Calibrate EV% carefully. Small changes (e.g., 1–2%) can dramatically affect ruin risk and drawdowns depending on stake size and odds.
  • Higher odds (big underdogs) increase variance even with the same EV%.
  • Use a seed when you want to reproduce the same charts for screenshots or presentations.
  • If results look “too smooth,” increase the number of simulations for a more stable distribution view.

Common pitfalls

  • Entering non-integer American odds: Use integers like -110 or +150.
  • EV% mismatch: If you type a large EV% with short odds, the implied win probability may get clipped to [0,1]. Keep EV realistic.
  • Commission: It applies to payouts on wins; this lowers d_eff and reduces your effective edge.

Formulas reference

  • Decimal from American: d = 1 + A/100 if A > 0; d = 1 + 100/|A| if A < 0.
  • Effective decimal with commission c%: d_eff = 1 + (d − 1) × (1 − c).
  • Edge to probability: p = (1 + EV)/d_eff, where EV is EV% in decimal.
  • Kelly fraction*: f* = (b·p − (1 − p))/b, with b = d_eff − 1.

Leave a Comment

Your email address will not be published. Required fields are marked *