In-Play Totals: Live Goals Markets

## Totals in the Live Environment In-play totals markets add a real-time dimension: rather than predicting the final total from the start, you are predicting the remaining total from the current game state. ## The Current State Update At any moment during a match, the live goals market prices the probability of reaching various total-goals thresholds from the current score: **Example:** 0-0 at 65 minutes. Initial total line: 2.5. Now the live market offers: - Over 1.5 at 1.60 (remaining ≥2 goals expected? Implied 62.5%) - Over 0.5 at 1.20 (at least 1 more goal? Implied 83.3%) Your live model should estimate the remaining expected goals per unit of time remaining, then generate the probability of each remaining-goals total. ## The Key In-Play Totals Calculation Remaining expected goals ≈ pre-match expected goals × (1 − fraction of match elapsed) × intensity adjustment This is a simplification — actual remaining xG should use accumulated in-game xG data for better accuracy. ## Common In-Play Totals Edges - **Under after high xG half:** A match with high xG but no goals through 45 minutes is "owed" goals statistically, but the Under 0.5 for second half may be mispriced - **Over in injury time:** Pressure from a trailing team creates disproportionate attacking output in injury time — over/under markets often underprice this - **Post-red card goals:** Markets often over-adjust for red cards — sometimes under prices become good value immediately after a red card when teams settle into their new formation ## The Suspension Window In-play totals markets frequently suspend around goals, red cards, and half-time. Acting before suspension is the execution challenge — the same speed constraints as all live markets apply.