This topic comes up all the time and is debated to death, I don't get it why people don't understand what a basically useless statement "all you have to do is pick the winner" is. First off I'm not a great math guy and yet I get how wrong that statement is. I'm going to paste an article from a great capper and number cruncher, his name is Number Freak and maybe you've heard of him. It's a response where he addresses this topic. This paticulair article is good but he has another post I can't find now that really went into the mathematical aspect of this. This post is over three years old now but it should shed some light on this for some.
--------------------------------------------------------------
I've seen it posted on this site and others many times that "all you have to do is pick the SU winner because the point spread only comes into play 16-19% of the time."
This statement is both true and false. A classic example of how numbers can mislead if you don't analyze them correctly.
While it is true that the SU winner fails to cover the spread 16-19% of the time (lines vary, I have it at 19% over the last 20 years), the situations where we would ideally look to this for reassurance would be when playing favorites. Cappers are trying to justify or rationalize their wager on the favorite, not the underdog. It's common sense that all SU dog winners cover and that warps the numbers and misleads the public. So, to analyze correctly as I mentioned, throw that 19% out the window.
We need to only analyze favorites first off, and second, we need a breakdown of each point spread range. How can we assign one number (19%) to every point spread? We can't, but that's what people post on these sites. It's simple mathematics that the spread comes into play more in 10-point lines than 3-point lines.
So, here's the real story. That 19% number is for favorites AND dogs and also an average of ALL lines. The average line is 5 points so 19% would be accurate at 5 points and 5 only. However we can only measure favorites which raises that average number considerably to 29% over all lines but to get this accurate and usable, we need a range breakdown.
The spread comes into play this often when betting FAVORITES....RS = regular season; PS = post season
For lines Pk-2.5...7% RS; 18% PS (only 11 PS samples)
For lines 3.0-6.5...25% RS; 22% PS
For lines 7.0-9.5...39% RS; 29% PS
For lines 10.0-13.5...42% RS; 25% PS (only 24 PS samples)
For lines 14.0-16.5...52% RS; 50% PS (only 4 PS samples)
For lines 17.0-19.5...57% RS; 0% PS (only 23 RS & 2 PS samples)
For lines 20.0+...67% RS; N/A PS (only 6 RS & 0 PS samples)
Since the playoffs samples are limited in certain line ranges, here are all lines combined:
29% RS; 25% PS
You can see the drastic increase from less than 3 points, to more than 3 points, to more than 7 points, dispelling the popular myth that the line matters only 16-19% of all games. What this all shows me is that there is no angle to benefit us here. Obviously, the line matters more at higher spreads than lower and therefore nullifies any advantage to just predicting the SU winner. It is very easy to predict SU winners in double digit spreads but very difficult to do so at <3.
