Since there seems to be some interest in learning how to handicap, let's apply the probability of winning to point spreads. First you need to build your model and test it using the probability percentages as stated in the previous post. That will work for betting money lines, but now let's add point spreads.
Once your model determines the probability of each team winning, then you assign a price to each point. For beginners, make each point worth 10 cents. So let's say your model gives a team a 54% chance to win, which is -117. What should the underdog price be? If you said +107, you are not thinking like a handicapper. The correct answer is +117. The sports book cheats us on the underdog price, but your model is trying to identify fair pricing for both teams, so there is no 10 cent spread or 20 cent spread from favorite to underdog. Your model is a 0 cent line. So a team with a 54% chance to win, should be a pk -117. Now apply a 10 cent value for each point and we get:
-1 -107, -2 +103, -3 +113, -4 +123, -5 +133
Underdog should be a pk +117, gives us:
+1 +107, +2 -103, +3 -113, +4 -123, +5 -133
Now look at the posted line at a sports book. Let's say it's -4 -110. So the sports book has the favorite winning by 4 points 52.4% of the time. Now look at your model's pricing for -4, which tells you it should be +123 so obviously you're not interested in laying the -4. But the underdog is +4 -110, and your model has +4 at -123, which is 55.2%. So there's a 2.8% edge on +4 (55.2 - 52.4). If the line were +5 -110, the edge would be 4.7% (57.1 - 52.4).
The next step is to get more precise with point pricing. A book could be written about this, but the idea is not all points are the same. In football, for example, the difference between 1 and 2 is not the same as between 2 and 3. So starting with a pk, you might add 5 cents to point 1 and another 5 cents to point 2, then add 30 cents to point 3. In basketball the first 4 points are worth more than points 8 thru 12. Once you get this far along in your handicapping abilities, you can start making point pricing adjustments to fit your model.
Hope this gets some of you on the right road to learning how to handicap. If you spend hours looking on websites for trends, streaks, head-to-head data, ATS records, a team's record on Tuesdays, etc., you are not even headed in the right direction. You are like a guy in St Louis who decides he wants to go to California so he heads East on the interstate. He has as much chance of getting to California as you do of long-term winning.