I categorize games 3 ways
A) completely Fresh
B) 3rd in 4(did not play last night, but played the 2 nights prior)
C)Back to Back(played last night)
I give A and extra point over B and an extra 2 points over C. B gets and extra point over C.
Is it completely accurate? It is hard to say but I find it works well enough. and found no reason to change it when fitting the data into results.
Regarding a half point, I addressed this in a thread a 6 months back or so in the mess hall. Cannot find it. Basically a half point is worth about 1.9%. A little higher with road favorites, also I probably a little higher in the 2-4 range as so many games that go down to the wire will end up with a margin of victory of 2-4.
Actual data I have over the last 4 seasons on 3 point home favorites is
56-81-6
the 6 pushes would either add to the losses or add to the wins depending on whether you were giving or getting an extra 1/2 points.
56-81=40.87%
Getting would bring you up to 43.356%
Giving would bring you down to 39.16 %
for the sake of getting a larger sample size-converting all home favorites of -2.5 to -3.5 to -3
We get 170-191-19 actual psr (47.1 %)
becoming 189-191 when getting an extra 1/2 point (49.73 %)
becoming 170-210 when giving an extra 1/2 point (44.73%)
On road favorites we get the following(converting all 2.5 through 3.5's to -3)
143-124-12 actual(53.55%)
meaning
155-124 when getting an extra 1/2 point((55.55%)
143-136 when laying an extra 1/2 point(51.25%)
By the way the actual stats on 3 point road favorites are
52-46-5
 |