Asian Handicap

Soccer Asian Handicap Model

Spread predictions for EPL matches using Skellam distribution to model goal margin probabilities with xG-based team ratings.

Methodology

The Asian Handicap model projects expected goal margins using the Skellam distribution - the difference between two independent Poisson variables. This gives us precise probabilities for each possible margin outcome.

// Projection Approach

Expected goals calculated for each team

Margin distribution via statistical modeling

Factors:

• Team attack and defense strength

• Home field advantage

• Recent form weighting

Output:

Probabilities for each handicap line

Margin Modeling

• Skellam distribution for margin probs
• Support for margins -5 to +5
• Quarter line split bet handling
• Std dev typically ~1.8 goals

Confidence Tiers

• MAX (2.0u): Highest EV picks
• STRONG (1.5u): Strong EV picks
• STANDARD (1.0u): Solid EV picks
• All picks must meet minimum thresholds

Key Factors

Projected Margin

Expected goal difference based on team ratings. A projected margin of +1.2 means the home team is expected to win by ~1 goal.

Margin Standard Deviation

Captures uncertainty in the projection. Higher variance matchups require larger edges to maintain confidence.

Line Shopping

Model compares odds across multiple books to find the best available price for each handicap line.

Home Field Advantage

EPL home teams average ~0.3 goal margin advantage. This is baked into the projection model.

Common Handicap Lines

Draw No Bet

Push if draw

-0.5

Win to cover

Most common

-1.0

Win by 1+

Push if win by 1

-1.5

Win by 2+

No push possible

Quarter lines (-0.25, -0.75, -1.25, etc.) split bets between adjacent whole/half lines.

Settlement Examples

Example 1: Man City -1.5 vs Brighton

• Final Score: Man City 3-1 Brighton (Margin = +2)

• Man City -1.5: WIN (margin ≥ 2)

• Brighton +1.5: LOSS (margin > 1)

Example 2: Arsenal -1.0 vs Newcastle

• Final Score: Arsenal 2-1 Newcastle (Margin = +1)

• Arsenal -1.0: PUSH (margin = 1, stake returned)

• Newcastle +1.0: PUSH (margin = 1, stake returned)

Example 3: Liverpool -1.25 vs Wolves

• Final Score: Liverpool 2-1 Wolves (Margin = +1)

• Liverpool -1.25: HALF LOSS

→ Half on -1.0: PUSH (returned)

→ Half on -1.5: LOSS

→ Net: -0.5 units

Frequently Asked Questions

Asian Handicap is a spread betting market that eliminates the draw outcome by giving one team a head start. For example, if Man City has a -1.5 handicap against Brighton, they need to win by 2+ goals for a bet on them to win. Brighton +1.5 wins if they win, draw, or lose by exactly 1 goal. This creates a 50/50 market with roughly even odds on both sides.

The Skellam distribution models the difference between two Poisson random variables - perfect for goal margins in soccer. Given expected goals λ₁ for home team and λ₂ for away team, Skellam gives us the probability distribution for all possible margins (-5 to +5). This is more accurate than simply subtracting expected goals because it captures the full variance of both teams' scoring.

We first project expected goals (xG) for each team using attack/defense ratings. These feed into the Skellam distribution to calculate P(margin = k) for each possible margin k. To get P(Home -1.5), we sum probabilities where margin ≥ 2. For quarter lines like -1.25, we split the bet: half on -1.0, half on -1.5.

Quarter lines like -1.25 or +0.75 are split bets. A bet on Home -1.25 is half on Home -1.0 and half on Home -1.5. If the margin is exactly 1, you win half (the -1.5 portion loses, the -1.0 portion pushes). This reduces variance compared to full goal lines while offering slightly different odds.

Attack and defense ratings are derived from xG data, measuring team performance relative to league average. We weight recent matches more heavily and apply regression for small samples. This captures current form while maintaining statistical stability.

Edge represents expected value (EV) as a percentage. Calculated as: (Model Probability × Decimal Odds) - 1. An edge of 6% means for every $100 wagered, you expect $6 profit on average. Only picks with positive EV above our minimum threshold are surfaced.

Confidence tiers (MAX, STRONG, STANDARD) are based on expected value thresholds. Higher EV picks are assigned to higher tiers and warrant larger unit sizing. Only picks meeting our minimum requirements are surfaced.

Picks are generated in the morning on match days UK time. Picks lock before kickoff to capture closing line value. The model compares odds across multiple bookmakers to find the best available price.

CLV measures line movement in your favor after locking. If you took Home -1.5 at -110 and it closed at -125, the market moved toward your position - positive CLV. Consistent positive CLV indicates your model is identifying value before the market fully prices it in. It's the best predictor of long-term profitability.

Settlement is based on the final margin plus the handicap. Home -1.5 wins if actual margin ≥ 2. Home -1.0 wins if margin ≥ 2, pushes if margin = 1. Quarter lines settle as split bets. For example, Home -1.25 with margin = 1: half loses (the -1.5), half pushes (the -1.0), net result is -0.5 units.