I wanted to take a look at how this past season compared to previous seasons in terms of ATS numbers and results. The source for my data may not be the same as the source for your data (if you try to compare), as the line shifts and I don't know if you are using opening or closing lines, and what book you use.
But all my numbers are consistent w/ themselves (same source) and so the analysis should show similar results, even if you use a different source.
First I looked at the percentage at which favorites covered. Knowing the public likes taking the favorite, and knowing that you only need to hit just over 51% to win money w/ consistent betting practices (getting better juice can drop this number), I just wanted to see what years the favorites did well and what years the underdog did well.
I looked at both regular season and post season numbers.
The second thing I looked at was the number of games in which the points come into play. This is from an underdog perspective. If a favorite wins and covers, you don't have to worry about the points, you lost your bet. If an underdog wins, you don't have to worry about the points, you won your bet. The time the points come into play are those times where the favorite wins but did not cover. Then the underdog bettor won, but the points did matter.
I have heard various people claim how the points only matter in such a small % of games. So the column for "points mattered" shows the % of games where the dog lost but covered the spread (or the underdog won but failed to cover the spread).
As you can see, I only took my sample thru the 1991 season. I could have gone further back, but I felt that was a large enough sample.
At any rate, here are the numbers, and my conclusions follow:
Conclusion 1: Betting underdogs in the long run will make money.
Most bettors know this. The problem is, you aren't betting across the board, you are picking select games, so that shifts the percentages. However, looking at all 16 years, or even at the last 5, underdogs cover slightly more than favorites.
But, if you look at individual seasons, you will see how much it can shift from one season to the next. The largest shift took place this season. In 05-06, favorites covered at 58% - killing a sharp bettor and helping the public. This season, the favorites only covered 45%. That's a 13% swing!
There have been spans (from 92-97 and from 99-02) where for 4 years or more, favorites covered at less than 50%. Significantly less (47% and 46% respectively, and those numbers are significant). But the past 4 years, we have had 2 years where favorites covered more than underdogs.
This season it dropped low, to 45%. My gut feeling is we should see another season where more dogs cover, and we won't see a season like 05-06 (58% of favs covered) for quite some time.
Conclusion 2: Favorites tend to cover more in the postseason.
Of course, if you look at the last 5 years, that has not been the case. However that avg is really brought down by only 36% of the favs covering in 2003-04.
But for 11 of the 16 seasons, the fav covered more in the postseason than in each regular season. That is 69% of the seasons saw more favorites cover in the playoffs than the regular season. Another number tough to ignore. Now, by how much more did they cover?
In those 11 seasons, they covered an average of 9% more games. Which is a sizable amount.
Even if you take the average % increase using all 16 seasons (including those where the fav covered less in the postseason) you still are looking at an increase of 4%.
There was no real correlation between the regular season and the postseason.
So while this won't matter for almost another full season, remember that on average, favorites do fare better than underdogs ATS in the postseason.
Conclusion 3: The points are extremely significant.
The way I calculated how much the points mattered was:
I took the number of games won SU by the favorite and subtracted the number of games where the favorite covered the spread. So we are left with the number of games where the fav won but the underdog covered (and the points mattered). I divided that number by the total number of games played that season.
[Note: I looked at backing out those games that "pushed", but on average the "pushed" games only amounted to 0.6% of the total, so I ignored those games]
What we see is that the points matter in about 20% of all games played, and this is pretty consistent from season to season. In fact, the most it has ever mattered was 23%, and the least was 17%.
So the big overriding conclusion here is the points are huge, and don't let anyone tell you that they don't matter. Now, maybe getting into specific pointspreads, they don't matter as much.
One time the points don't really matter is if there is a home favorite of between 1 and 2.5 points. The points only mattered in 2 games out of 200, since 1995. In this situation, either take the fav and don't worry about laying the points, or take the dog SU. Because in only 1% of the games will the fav win but not cover. So if you want to look at specific situations, the points may not matter as much in those situations.
But in general, the points are a big part of gambling, and getting the most points you can get for a dog (while still having value in the juice) can affect around 20% of the number of bets you might win.
Conclusion 4: It is hard to predict how much the points will matter in the postseason.
We had 3 seasons where the points did not even matter (you could either take the fav and lay the points, or take the dog SU, there wasn't a fav who won but failed to cover).
We had almost half of the 16 seasons where the points mattered in 10% or less of the games, which is about half the amount it mattered in the regular season.
But then again, we have had those postseasons where the points mattered in 36% of the games, much higher than the amount it mattered in the regular season.
There is also no real correlation between the regular season and the postseason.
So how did 06-07 compare to the average?
We had more dogs cover in both the regular season and the postseason than average. And while the points didn't matter quite as much as average on the regular season, the points mattered a lot more in the playoffs than average.
So what to expect for next season?
At times an analysis does not really produce any incredible trends, and that is the case here. Going in, we should already know that the points matter, and we should know that dogs do better than favs on average. And that is all we really found out. The only thing I could surmise for next season is a similar number of dogs covering, as 2005-06 was really an aberration. Other than that, I'll be looking forward to another challenging year of handicapping in 2007-08.
