Calculating expected total from a line.

[donkson]

I've searched around for the answer to this for a while but haven't really found what I'm after.

Say you have an MLB total line o7 -112/+100, how do you actually determine an expected total (7.14 or whatever)? I understand you need to look at the push rates of totals etc, but I don't really understand how you go about it, thanks.

[donkson]

cool

[FourLengthsClear]

There is no simple answer to the question without knowing or having a reasonable basis to estimate the push rate on 7.
If I pull a number of 10% from thin air then you can work out the projected total as follows:

Probability of over 7.0 is 51.38% (-112 prorated to zero juice)
Probability of over 7.5 is 51.38% minus 5% (half of the push rate on 7) = 46.38%

Your projected total is represented by a 50% probability so:

7.0+(((51.38%-50.00%)/(51.38%-46.38%))*(7.5-7.0)) =
7.138 runs.

[donkson]

awesome, thank you.

[evo34]

Be careful if you are trying to go from projected avg. runs scored to odds around a given total. That is, if your model projects 7.5 runs on average, that is not the same as projecting 7.5 runs as the median. The median (which you need to create fair odds) will be less than the average due to the distirbution of total runs.

[donkson]

yeah I understand that, but it's not what I was using it for, thanks though.