FairPlay Rates

How this separate prototype works

Sportmonks confirms teams, format, start time, winner and match status. FairPlay builds its own format-specific Elo-like ratings from decided results; it does not copy bookmaker quotes or the private reference engine. Teams start at 1500, a first-named side receives a 30-point home adjustment, and an independently chosen K-factor of 24 and logistic scale of 400 turn ratings into probabilities.

Below eight rated results for either side, no winner estimate is shown. From eight through nineteen, the estimate is pulled toward 50% by the lower team count divided by 20. The model is replayed chronologically with a later holdout; winner previews require at least 100 eligible holdout matches and a Brier score below 0.250. Context such as venue and form is not secretly added to the estimate. Test matches have no three-way result price because there is no validated draw model.

Proposed margin and automatic withholding

For eligible T20/shorter or one-day match winners, or a theoretical 50/50 toss, the proposed two-sided book target is 105%. Decimal values are rounded down to this prototype’s disclosed 0.01 ladder, not the reference’s undisclosed ladder. Prices outside 1.01–26.00 or the proposed book band are withheld. The margin is not approved for betting.

A visible preview version lasts at most 15 minutes without reconfirmation. Match-winner previews stop at the scheduled start; toss previews stop 45 minutes before start or on a recorded toss, whichever is earlier. A delayed source, changed fixture or failed check suspends the preview instead of leaving a stale price available. Corrections to verified results change the next model version.

Read the actual data cadence and limits