For example (I use 8 stats, but will only use 3 for this example):
Let's say Toronto is ranked 17th in G/G, 12th in PP%, and 14th in FO%
That means they have a mean rank of 14.33
Now let's say Montreal is ranked 3rd in G/G, 15th in PP% and 5th in FO%.
That means they have a mean rank of 7.67
The way I would calculate the chance of winning would be the higher rank/sum of both ranks. That would give the chance of the team with the lower rank (meaning better stats) winning.
In the example above, that would be 14.33/22 = 65.14%
First note that it will be more accurate if we use more than only 3 stats, which as I mentioned, I use 8.
65.14% = -187, which would be the 'fair moneyline'
If the line is lower than this (ex. -140) take the favourite.
If the line is worse (ex. -200) play the dog if the dog gives better than +187.
If the actual line was +180/-200, it wouldn't be played, as it meets neither quality there for there is 'no value' betting either team.
I'm going to be playing every game, and have the stats separated into 3 categories:
A) Favourites
B) Dogs where the dogs chance of winning is >= 50%
C) Other Dogs
Without anymore mumbling on, here's today's games.
The fomat will be..
Vis/Home
Team with higher chance, %chance
The Pick
November 8th, 2008
Buf/Bos
Both 50%
Buf +120 [B]
Mon/Tor
Mon 65.46%
Mon -142 [A]
Pit/NYI
Pit 59.66%
NYI +150 [C]
TB/Phi
TB 50.90%
TB +137 [B]
NYR/Was
NYR 57.92%
NYR +109 [B]
NJ/Det
Det 57.46%
NJ +210 [C]
Cal/Clb
Cal 56.15%
Cal -115 [A]
Nsh/Col
Nsh 53.21%
Nsh +132 [B]
Fla/Pho
Pho 52.11%
Fla +144 [C]
Min/Van
Van 51.91%
Min +134 [C]
Dal/SJ
SJ 75.24%
SJ -197 [A]
Stl/LA
LA 52.55%
Stl +122 [C]