NHL O/U statistics question

[WallStreetadk]

I’m trying the convert a NHL total with a line into an expected total.

Should I try to use the Poisson distribution for this kind of event?

If so I have an example.
For 6 goals the under line is at +150. So the under should occur 0.4. When I use the Poisson formula do I set it, the cumulative probability of the score 6 or less, equal .4?

How do I solve for λ ?

I have no idea how to do this; any help will be greatly appreciated.

Thanks

[Ganchrow]

No, because total NHL goals scored per game would clearly not be Poisson.

Nevertheless for a manner in which the Possion could be used to describe hockey scoring see this paper.

[WallStreetadk]

No, because total NHL goals scored per game would clearly not be Poisson.

Nevertheless for a manner in which the Possion could be used to describe hockey scoring see this paper.

What kind of distribution do you recommend using?

[Ganchrow]

What kind of distribution do you recommend using?

You're not going to find a single "off-the-shelf" probability distribution that's going to sufficiently account for enough of the contingent probabilities embedded within the outcome of the total score of an NHL hockey game so as to be economically meaningful to the advantage bettor.

Check out the above-linked paper, however, as it gives an example of how one might attempt to modify the Poisson to make it applicable to the prediction of regulation time hockey MOVs.