How to Read Prediction Market Odds Like a Pro

A Polymarket contract trading at $0.65 means the market estimates a 65% probability of YES. That sounds simple — but the implications for profitable trading take time to internalize.

Prices Are Probabilities

On binary prediction markets:

• Price = probability (roughly)
• $0.65 per YES share = 65% probability of YES
• If YES resolves: your $0.65 share pays $1.00 (profit: $0.35)
• If NO resolves: your $0.65 share pays $0.00 (loss: $0.65)

Expected value: EV = (0.65 × $0.35) + (0.35 × -$0.65) = $0.2275 - $0.2275 = $0

At the market price, EV is zero — that's an efficient market. You only have positive EV if you believe the true probability differs from the price.

Converting Between Formats

def decimal_to_implied_prob(decimal_odds: float) -> float:
    """Convert decimal odds (common in Europe) to probability."""
    return 1 / decimal_odds

def american_to_implied_prob(american_odds: int) -> float:
    """Convert American odds to probability."""
    if american_odds > 0:  # Underdog
        return 100 / (american_odds + 100)
    else:  # Favorite
        return abs(american_odds) / (abs(american_odds) + 100)

def prediction_market_price_to_odds(price: float) -> dict:
    """Convert prediction market price to various formats."""
    prob = price  # Price IS the probability on prediction markets
    
    return {
        "probability": f"{prob:.1%}",
        "decimal_odds_yes": f"{1/prob:.2f}",
        "decimal_odds_no": f"{1/(1-prob):.2f}",
        "american_yes": f"+{int(100/prob - 100)}" if prob < 0.5 else f"{int(-100*prob/(1-prob))}",
        "payoff_per_dollar": f"${1/prob:.2f}",
    }

# Example: Polymarket showing 0.72 for "BTC above $80k by year end"
print(prediction_market_price_to_odds(0.72))

Finding Your Edge

The only way to profit consistently is having better probability estimates than the market. How?

Method 1: Reference class forecasting

Find the base rate. How often have similar events resolved YES historically?

"Will BTC be above $80k at year end?" — Look at: historically, how often does BTC end the year above the starting price? Above a 50% gain? Build from data.

Method 2: Information arbitrage

You know something the market doesn't. An industry expert, someone with connections, early data access.

Method 3: Market inefficiency

Some markets receive less attention and are priced by fewer sophisticated traders. Niche markets on Manifold, new markets on Kalshi — prices can be far from true probability.

The Vigorish (Market Fee) Effect

Prediction markets take a fee. On Polymarket, it's 2% on wins. This means:

If true probability is 65% and market price is 65%:

• Buy YES at $0.65: EV = (0.65 × $0.33) + (0.35 × -$0.65) = -$0.013 (negative!)
• The fee makes breakeven ~68% conviction needed to profit on a 65% market

def minimum_probability_to_bet(market_price: float, fee_pct: float = 0.02) -> float:
    """
    What probability do you need to believe in order to have +EV?
    Accounts for the platform fee on wins.
    """
    # After fee, win pays: (1 - market_price) * (1 - fee_pct)
    net_win = (1 - market_price) * (1 - fee_pct)
    loss = market_price
    
    # Break-even: p * net_win = (1-p) * loss
    # p * net_win = loss - p * loss
    # p * (net_win + loss) = loss
    # p = loss / (net_win + loss)
    breakeven_prob = loss / (net_win + loss)
    
    return breakeven_prob

market_price = 0.65  # 65% YES
breakeven = minimum_probability_to_bet(market_price, fee_pct=0.02)
print(f"Market: {market_price:.0%} YES | Need: {breakeven:.1%}+ conviction to +EV bet YES")
# Output: Market: 65% YES | Need: 67.2%+ conviction to +EV bet YES

Common Beginner Mistakes

1. Betting with the crowd: If a market is at 80%, the crowd already priced it there. You need strong specific reasons to bet Yes (agreeing) or No (disagreeing).

2. Confusing volatility with opportunity: Prices moving a lot doesn't create edge. Only disagreement with the true probability creates edge.

3. Overtrading: Each bet has fees. Neutral-EV bets eat capital over time. Only bet when you have clear edge.

4. Not sizing by edge: A 60% belief vs 55% market deserves a small bet. An 80% belief vs 50% market deserves a large bet.

Understanding odds is the foundation. The rest is calibrating your probability estimates and maintaining discipline to only bet when the edge is real.