Help with a formula...

[tuxStyle]

Hi


I'm starting with: I'm a noob in betting so please, don't hit me hard if i say something wrong

Using the data available on nhl.com, i calculate for the two teams the following averages for last 6 games:
- average goals scored/number of games played (6)
- goals scored/number of shots
- goals/pp opportunities

- average goals received/number of games
- goals received/number of shots received
- goals received/pp opportunities

Now, having this 6 values for each team, what formula ca i use to estimate the chances for win/draw/loose?

I know i can use Poisson to estimate the chances to score for each team but what if i want to add more factors?

To finish, i have to say i already started to read the think tank but will take me a lot of time to read and most important understand everything


Thank you

[BChrisB]

Home and Away Records
Home Points, Away Points

If your using the poisson average for small numbers these listed above are pretty important. You just can't us overall records as the deviation is rather large for Home and Away, especially in a sport such as Hockey.

[BChrisB]

I'm not too sure 6 games is enough. Furthermore, your way too early in the season for getting any decent average for such a small sample of games. This early in the season your going to have to use all games played which is about 30 games so far.

Early in the season of any sport averages are a bit more useless. The deviations between the numbers is too sporadic. It's kind of like a new car, it has to be broken in for the first 1000 miles before you'll really know how it drives.

[tuxStyle]

That is true but i was thinking at the current form.
There are some teams that won first games and then, nothing.
Well, is true, the more games the better.

Thing is, how can i put all those parameters in a formula to get one set of results?

[Flight]

You could start with Bill James' pythagorean theorem and go from there.

[tuxStyle]

You could start with Bill James' pythagorean theorem and go from there.

I know that theorem but i don't see how it can help me to combine all this values.
Yes, to have one more parameter is ok.
Same with poisson.

[Justin7]

In a particular game, you have to estimate the average score. Once you do that, you can use Bill James.

You probably want to know the league average for shot, goals and PPs as well.