[TryingMyBest]

I often find at least one game a day with a true middle. 5dimes offers 20 cent lines on them, most others use 30 cents. So we're looking at a 5.8% push percentage in order to break even. I noticed when using poisson, for example, middling rebounds per game had a very high push percentage, yet runs per game, which are dealing with numbers below 10 for the most part, are not as high, even key numbers such as 7 and 9. Is this because there are more variables involved with how many runs are scored by two teams in a game as opposed to how many rebounds one single player will record? Can this assumption also be used to assume H/R/E have many variables and push rates will also not be very high?

[FreeFall]

can't use poisson because runs can be scored in groups of 2,3,4.

I'd try making a scraper and a put that in a DB. Then run queries on your data.

[TryingMyBest]

can't use poisson because runs can be scored in groups of 2,3,4.

I'd try making a scraper and a put that in a DB. Then run queries on your data.

I'll try that. I understand you can't use poisson for that reason. I meant I was using it only for the prop middling on rebounds or assists

[TomG]

Well the RHE number is usual around 24-29. What else is in that range? 1st Half NFL Totals. So each half point is probably worth somewhere between 5 and 30 cents.