Greyhound Market Making Lifecycle

Market Making Lifecycle

An algorithmic analysis of greyhound bookmaking, from origination to settlement.

Objective

This interface deconstructs the black box of sports trading. It is designed for quants and model developers to understand how odds are compiled, managed, and exploited. Navigate through the modules to simulate the lifecycle of a single market.

Key Core Concepts

Pricing: Logistic regression & Poisson models.
Risk: Distinguishing 'Sharp' vs 'Retail' flow.
Margin: Non-linear application of vig (Fav-Longshot bias).
Latency: The millisecond war for stale prices.

Note for Analysts: The simulations below use simplified deterministic models for demonstration. Real-world production environments utilize complex stochastic differential equations and real-time liquidity aggregators.

1. Tissue Pricing (Origination)

The "Tissue" is the opening price set by the bookmaker. While historical manual compilation was common, modern origination is 90% automated using regression models. The model predicts the probability of winning ($P_win$) based on quantifiable features.

Model Input Variables

Speed Rating (Weight)

Low ImpactHigh Impact

Trap Bias (Weight)

NeutralHeavy Bias

Recent Form (Recency)

LTD OnlyLast 3 Runs

Simulated Dog A (Favorite) Stats:

Split: 3.45s (Fast)

Trap: 1 (Inside)

Probability Construction

Calculated True Odds

3.50

Methodology: The chart above represents a simplified linear combination of factors. In production, this would be a logistic regression output converted to implied probability. $$ P(win) = \frac{1}{1 + e^{-(\beta_0 + \beta_1 Speed + \beta_2 Trap)}} $$

2. Dynamic Liability Management

Once prices are live, the bookmaker manages the "Book". Price movements are dictated by two forces: Weight of Money (Balancing) and Sharp Information (True Price Adjustment).

Trading Desk Actions

Current Odds

4.00

Simulate incoming order flow:

Price Ladder Movement

Retail Impact Sharp Impact

Analysis: Notice how retail bets cause minor, gradual decay (auto-balancing exposure). A "Sharp" bet triggers an immediate, drastic cut (steamer) because the model interprets it as signal, not noise. The bookmaker creates a new "True Price" instantly.

3. The Overround & Vigorish

The "Overround" is the sum of implied probabilities above 100%. This is the bookmaker's theoretical margin. Crucially, this margin is not applied uniformly. It is heavily weighted against longshots (outsiders) to protect against volatility (The Favorite-Longshot Bias).

Market Margin (Overround)

105%

100% = "True Fair Market". 125% = High margin early market. Note how the Red bars (Book Odds) diverge from Blue bars (True Odds) as margin increases.

Fav True Price: 2.00

Fav Book Price: 1.90

Outsider True Price: 20.00

Outsider Book Price: 15.00

4. Market Microstructure & Latency

Modern pricing is derived from Exchanges (e.g., Betfair). The Exchange Price is the "source of truth". Fixed-odds bookmakers trail this price. "Steamers" occur when bots snipe the bookmaker's stale price before their API polls the exchange update.

Latency Sequence

t=0ms: The Event

News breaks (e.g., Trap 1 withdrawn).

t=200ms: Exchange Moves

High frequency bots re-price Betfair.

t=400ms: The Arb Window

Bookmaker is STALE. Snipers attack.

t=1500ms: Book Update

Bookmaker API polls and updates price.

Latency Arbitrage Visualization

Defense Mechanism: To protect against this, bookmakers use "Bet Delay" (e.g., 4-8 seconds in-play) or "Profiling" to reject bets from accounts that consistently beat the 1500ms update window.