Bettormetrics data shows uptime figures can mask peak-moment outages
A US Open study of 18 sportsbooks finds that conventional uptime scores can conceal downtime during the exact moments when in-play wagering is heaviest.
A new analysis from data firm Bettormetrics argues that the industry's standard way of measuring sportsbook uptime can obscure the moments that matter most commercially, using in-play betting on the US Open as a case study.
Conventional uptime is calculated as the share of a match's total duration during which a market was live, treating a quiet opening set the same as a deciding tie-break. Bettormetrics says this flatters operators whose outages happen to fall in low-activity periods while still letting them look strong on the headline figure.
Its alternative measure, wagering-weighted uptime, instead tracks the proportion of matched exchange wagering that occurred while an operator's markets were actually live, on the basis that betting activity is not spread evenly through a tennis match.
The firm compared operator pricing and availability against exchange pricing across 39 in-play men's singles matches and 18 operators at the tournament, measuring performance both by duration and by the share of exchange wagering taking place under each condition.
LeoVegas produced the starkest illustration of the gap. Its duration-weighted uptime was 97.8%, in line with the likes of Superbet, Paddy Power and Hard Rock, but its wagering-weighted uptime was 88.3%, the lowest in the sample and the only figure below 90%.
That 9.5 percentage point difference between the two measures was more than double the next-largest gap recorded across the field, according to Bettormetrics.
A single fixture, Zverev against Tabilo, showed how this plays out in practice. LeoVegas was live for 98.5% of the match by duration, yet those live periods covered only 61.5% of the exchange wagering activity recorded in the match, meaning its downtime fell during periods accounting for more than a third of the betting.
On pricing, Pinnacle carried the tightest average overround in the sample at 3.7%, well below the rest of the field, while keeping its theoretical arbitrage exposure relative to exchange pricing comparatively low, at 4.8% by duration and 6.4% by wagering.
Bettormetrics notes that this combination runs against the tournament-wide pattern, in which tighter margins were generally associated with greater arbitrage exposure rather than less, since a book priced close to the exchange's fair value tends to slip into arbitrage more readily when that price shifts.
DraftKings illustrates the more typical relationship: its overround of 4.5% was the second-tightest in the sample, nearly matching Pinnacle's, yet its theoretical arbitrage exposure by duration was 14.9%, roughly three times higher.
Bettormetrics defines theoretical arbitrage as any instance where at least one sportsbook selection was priced such that a customer could, in theory, back it profitably against the exchange price, while cautioning that this does not confirm every such opportunity was actually executable.
The firm says it is now adding prediction market data alongside its existing exchange data, arguing that both are exchange-based environments producing market-derived pricing and wagering signals, and that the same wagering-weighted methodology can be applied to either where liquidity is sufficient.
Bettormetrics says the approach is intended to give operators and the wider market a clearer, commercially grounded view of in-play performance across a broader range of sports, particularly within the US betting market.