Suppose that you have a tipster who is expected to maintain a particular over-all strike-rate, and a particular average winning dividend, betting into a particular market medium with a particular constant total market %. Given a tip with a particular ( I love this word ) available dividend in a market with a particular total %, how do you calculate the probability of a win? Well, I gather that when a bookmaker or a capper frames a market, she works from a set of expected contestant performance differentials, and assumes that for every x amount of expected difference between 2 contestants, their probabilities differ by a FACTOR of y. Eg, c1 is x lengths superior to c2: P1 = yP2, & with 3x lengths difference, P1 = (y^3)P2. My tip strength indicator concept is based on the idea that the expected performance of a tip is a particular constant amount higher than that expected by the market. Tip Strength ( T ) is defined as follows: T = .01Ps( MDa - 100 )/( 1 - Ps ) ( Ps = expected strike-rate of tipster ), ( Da = expected average winning dividend ) & ( M = market % in which the preceeding dividend figure is achieved. ) The probability of success of a tip for which the available dividend is D, in an M market, is given by: P = ( .01( MD - 100 )/T +1 )^-1 Soon I will be posting more formulae related to this concept.
Situation: There is a different tip from each of 2 different sources on the same event: P1 = [ .01*( D1*D2*M - 100*D1 )/( D2*T1 )*( 1 - 100*T2/( D2*M + 100*( T2 -1 ) ) )^-1 - 1/T1 + 1 ]^-1 For P2, apply symmetry. More later.
Situation: There is more than 1 tip from the same source in the same event: Calculate overall dividend ( Doa ): Doa = ( 1/D1 + 1/D2 + 1/D3 + . . . + 1/Dn )^-1 ( n = number of tips ). Calculate overall probability ( Poa ): Poa = ( ( M*Doa - 100 )/( 100*T ) + 1 )^-1. The probability ( Pi ) of an individual tip is given by: Pi = Poa*(( ( M*Di - 100 )/( 100*T ) + 1 )^-1)*( ( ( M*D1 -100 )/( 100*T ) + 1 )^-1 + . . . + ( ( M*Dn - 100 )/( 100*T ) + 1 )^-1 )^-1
SBR Top-Rated Sportsbooks
Tip Strength
Reply With Quote