/v1/probability prices any two players over any format. The gap between their ratings gives each player's chance, and the format decides how that chance plays out over a race of legs or sets.
Match length
Whether a longer match favours the stronger player more is measured from the archive rather than assumed, and applied as measured. Every answer says which measurement it used.
Against the market
/v1/markets/value and /v1/matches/{id}/probability put the model beside the closing price with the margin taken out. Over the priced archive the market is the better forecaster, and the endpoints say so rather than presenting every disagreement as value.
Outrights
/v1/editions/{id}/outrights and /v1/simulate/bracket work out each player's chance of each round exactly, through every path of the draw. Nothing is simulated, so the same draw always gives the same answer.