Someone recently asked me about the link between a players rebounds plus assists line and the players individual rebounds line and assists line.

For example

if a player has over 3.5 assists at -125 and over 4.5 rebounds at -130. If I assume that that bookies lines are perfect can that information then be used to determine the correct price of bets such as over 8.5 (points plus assists).

My guess is there is no easy way of doing that.

you would use the lines and prices to estimate distributions for assists and rebounds individually and then use combinations of those distributions to look at the implied probabilities of certain totals of rebounds plus assists.

You would always do better by looking at the individual players stats and making models based on that. But if you were in a situation where you only had the sharpest bookies lines to offer insight into the players performance.

Anyone have any ideas?

individually they are both poisson distribution so you can figure out what the book has for the mean. The problem is that for rebounds + assists they are correlated so that adds a step to figure it out.