Odds comparison

The product is a grid, one row per selection and one column per bookmaker, with the top price highlighted and a link out to each bookmaker's own event page. Sites earn on those clicks, so a price that no longer matches the bookmaker's page costs trust and referral rebookmaker at once.

Take a tennis player priced at 1.80, 1.83 and 1.90 by three bookmakers. The best price is 1.90. A 10-unit stake there returns 19.00 units, against 18.00 at 1.80, which is one unit more for exactly the same bet. That unit is the reason comparison pages exist.

Comparison goes wrong when the columns hold different bets. A price only ranks against the same market and, on line markets, the same line. Age matters as well. On the raw feed last_update says when a bookmaker's market last changed, and on the matched feeds meta.last_seen says when each bookmaker was last read.

In the OddsRelay feed

On OddsRelay's matched feeds one outcome is one row of a grid. Its back[] lists every bookmaker's price for that selection, best first, each with bookmaker, price and link, so the columns are already lined up. The raw feed carries every bookmaker's prices per event in one schema, with includeLinks=true adding each bookmaker's event page, but each book's outcomes keep that book's own labels, so reconciling them into one row per selection is the raw consumer's own work.

The matched feeds carry a ready-made ranking. Each outcome's back[] lists bookmaker, price and link, best price first, so the first entry is the top price among the bookmakers on that feed.

The two shapes answer different questions. Matched pairs are gated on lay liquidity, so a bookmaker's price appears on a matched feed only where an exchange lay beside it has money behind it, and a selection no exchange prices has no row there. Raw lists every bookmaker quoting the event, those selections included, and leaves the matching of labels to you.

Related