What is live betting. Why the in-play vig is 2 to 3x pre-game.

Plain-English answer

Live betting (also called in-play betting) is wagering on a game that has already started, with prices that update in real time as the score, time remaining, and possession state change. Books quote live markets with 8 to 12% vig (versus 4 to 5% pre-game) because the latency between an event happening on the field and the price updating gives the book information edge. The three windows where retail consistently mis-prices: pre-halftime second-quarter total in NBA when the pace differs from season average by 6+ possessions per 48, post-touchdown win-probability in NFL when the line ignores the kickoff variance, and goalie-pull mid-third-period in NHL when the empty-net total still prices like a steady state. The latency trap: a TV broadcast is 8 to 12 seconds delayed; the book's data feed is sub-1 second. Betting off TV against an algorithmic price means buying a number that already reflects the play you just saw. Sharp practice: bet only off the box-score data feed and only into windows where the model has a structural edge, not into reactionary clicks.

Cable TV broadcast latency is 8 to 12 seconds. Streaming services run 12 to 30 seconds. The book's data feed (Genius Sports, Sportradar, Stats Perform) updates in under 1 second. By the time you see a play on TV, the live price has already repriced for it.

• Reactive bets off TV: roughly 60% are negative-EV (the price you get already reflects the event you reacted to).
• Pre-planned bets in known structural windows: the model logs edge consistently because the entry trigger is data-feed-clean.
• The book's bet-acceptance delay (5 to 8 seconds in most apps) extends the latency trap because the price you click can move while the bet is in flight.

Pair the live read with steam-move detection so you never bet into a wave the book has already settled, and with CLV to grade the entry independent of the live result.

Use it live

NuroPicks Pro and Elite ship live alerts only into the three structural windows above (Q2 NBA pace divergence, post-TD NFL kickoff variance, NHL empty-net total mispricing). Outside those windows the model stays silent, because clicking reactively into a live price is the fastest way to lose a bankroll.