View Single Post
  #10  
Old 12-01-2007, 11:58 AM
centris centris is offline
Senior Member
 
Join Date: Jul 2006
Posts: 232
Default Re: Likely dumb vig question

[ QUOTE ]
Since the relationship between the price and its associated probability is not linear the mean of the two prices is not the mean of their associated probabilities. So no you cannot just take the average of the two.

For each price calculate the associated probability and then take the mean of those two probabilities and translate this back to a line.

Probability asssocaited with -300= 300/400=0.750

Probability associated with -240= 240/340=0.706

(.75+.706)/2=.728

Line= 100*.728/(.728-1)=-268 (versus -260 by just taking the average)

[/ QUOTE ]

Oh yeah, gosh that is embarrassing I made two mistakes here. If you take the average it should just be -270. And then as Thremp wrote this is a totally bastardized calculation. Just go to the linked page and see the correct calculation. But for the sake of comparison:

Probability asssocaited with -300= 300/400=0.750

Probability associated with +240 = 100/340=0.294

Probability of fav winning = .75/(.75+.294)=0.718

Vig-free line = -255

So there is actually a fairly large difference:
Average -270
My bastardized calculation -268
Correct vig free calc -254

As to which one actually works the best I have no clue.
Reply With Quote