NHL back to back games

[element1286]

I created a model that predicts the goals for and goals against for each team, each night, based on their current lineup. I use this to generate my own lines, and thus far they have been extremely accurate, and I have been able to find some inefficiencies.

I have had one problem, that is my lines are completely off for the second game of a back to back situation. Does anyone have the records (preferably team by team) in the first game and second games of back to back situations for the last few years?

Or does anyone use a standard adjustment for the second game of the back to back situation?

[Justin7]

Two things you might do.

On average, how much do these teams under perform (or do opponents of them over perform) based on your model? And if you have a line history, how much does it appear the line is adjusting for it?

[Pokerjoe]

Someone somewhere out in googleland did the work. I know I stumbled upon it once.

But if you have your own DB, why don't you just crunch the numbers yourself? You say your numbers are off for the second game of a b2b; the average amount they're off answers your question.

[element1286]

Two things you might do.

On average, how much do these teams under perform (or do opponents of them over perform) based on your model? And if you have a line history, how much does it appear the line is adjusting for it?

Thanks for the ideas, I actually hadn't thought of either, but I want the input to be independent of my model itself.

[element1286]

Someone somewhere out in googleland did the work. I know I stumbled upon it once.

But if you have your own DB, why don't you just crunch the numbers yourself? You say your numbers are off for the second game of a b2b; the average amount they're off answers your question.

I think I am going to have to crunch them myself, just though I'd ask if anyone had them on hand.

The problem is that some night they are off 10 cents, some nights 40 or 50 cents. And I don't know how much off this is due to the line being calculated different (on my end or the books) and how much is due to the market.

[Stocks]

Adjust them more for 3 games in 4 nights where they would play like on monday have tuesday off and then play again on wednesday and thursday.

The other deal with back to back games is starting goalie, some teams start they're backup on the first game of B2B while others start they're backup in the second game.

[element1286]

Adjust them more for 3 games in 4 nights where they would play like on monday have tuesday off and then play again on wednesday and thursday.

The other deal with back to back games is starting goalie, some teams start they're backup on the first game of B2B while others start they're backup in the second game.

I haven't thought of 3 games in 4 nights, but I will look into that.

I have it controlled for goalies, as I I have separate input for the stater and the backup.

[mathdotcom]

http://forum.sbrforum.com/handicapper-think-tank/478997-looking-market-inefficiency-vs-predicting-line.html

[subs]

mathsy does that actually work for you? have you been able to find any variables or is this just advice on how to factor in B2B games, kinda like what pokerjoe said?

just curious?

[Peregrine Stoop]

not all b2bs are created equally