A Technique to Improve Your Handicapping Results:
by Rick Janov
Two Plus Two Magazine, Vol. 7, No. 12
Years ago, when I first started handicapping, I used to spreadsheet what I felt were the significant variables i.e. yards rushing, yards passing, turnovers etc. I would spend many hours each week to come up with games where the line seem to be off based upon what the stats showed in my spreadsheets. I was always dismayed when I came up with a side that seemed clear cut and it lost.
Then, computers came into play. Handicapping software packages came out that used the same stats I was spreadsheeting and crunched the numbers based upon the stats. What a time saver! But the results were the same.
It took me awhile but I came to the conclusion that the best way to find value is to use techniques that were completely contrary to how the betting public viewed the games. The public was the key since the people who make the lines seem to incorporate a bias based upon their perception of how the public feels regarding the game.
In handicapping the NFL, I have found that a systems approach that incorporates a variety of variables most of which is not considered by the public works best for me.
Some of the variables I look at include (in order of importance)
An Analysis of Public Betting High Percentage Situational Angles Line Moves as they Relate to the Public Betting Several Neural Models to Approximate Value for the Game (although this seems to be the least significant) An Analysis of Public Betting Of the above a techniques, if you use it properly, public betting percentages can give you a good head start in finding value. It’s available in all sports but I have found it to be the most predictive in the NFL.
This is most likely because there is no other sport where the public unloads their money than professional football. The public’s betting tracks are clearly visible.
First, you have to determine where the public is on a side in the game and to what extent. I use Sports Insights as I find it to be the most reliable and they accumulate the information real time from a variety of off shore books.
It is costly however with a yearly subscription approaching $600 a year. The only other option I have found is Covers site but their percentages are not based upon actual money wagered but predictions by members. Sometimes the two diverge significantly.
The threshold numbers I use are under 20 percent on the home team and under 40 on the road team for games under +7. For games +7 and over, the threshold numbers would be under 40 both at home and road.
In other words, for home teams if the public is under 20%, I put them in my games to handicap. For road teams, if the public is under 40%, they go into my subset of games to handicap.
Now if a team is a +7 or more dog, then for either home or away the threshold number is under 40%.
Practically, what this means is if you’re laying -110, you have cut the bookie’s advantage of 5% in half. If you’re laying -105, you have a slight edge in your picks from the start. If you’re betting other bettors with no vig, you have about a 3% edge. And this is before you start your further handicapping on the game.
Sports Insights publishes figures that are updated each year as to where the sweet spot is in interpreting the public numbers in most sports. The below figures are taken directly from sports insights studies covering the last 8 years.
The first chart covers all games regardless of the line for home and away teams.
Table 1: Betting Against the Public (2003-2010 NFL Seasons)
Betting Percentage Home Team Visiting Team < 40% 48.7% 54.4% < 30% 50.6% 53.9% < 25% 52.3% 52.7% < 20% 54.5% 36.4% (few games) The second chart covers games where either the home or away team is getting +7 or more. Table 2: Betting Against the Public on 7-Point or More Underdogs (2003-2010 NFL Seasons) Betting Percentage Home Team Getting 7+ Points Visiting Team Getting 7+ Points < 40% 54.6% 54.2% < 30% 55.1% 53.7% < 25% 53.7% 52.6% < 20% 54.5% 33.3% (few games)
Now a caveat. This alone will not do the trick. But if you have a solid methodology that you use in handicapping with some success, adding this filter will improve your results substantially. High Percentage Situational Angles There are many publications that come out yearly that includes high percentage setups for each game in the NFL.
I have a database that I have developed over the years that I feel is the best for finding value. For me, I need to either have them on the game I am picking or at least neutral for me to consider it a play. A caveat here. There are many web sites that show trends for the games that they call angles. These I have found to be completely worthless and of not predictive value. Line Moves as they Relate to the Public Betting A pattern I look for is a line move opposite the way the public is going. So if we get a line move from -9 to -8.5 with the public over 60% on the favorite it would be a stronger indication that the dog is the right side.
The idea behind this is that for the line to move opposite the way the public is going there must be some “smart money” moving it in the opposite direction. Again, this is not conclusive but an indication of where value might be. Logically if the public is say 90% on the favorite you would expect the line to move but if it goes opposite this is a move contrary to what logic might tell you should happen. The books do not move their lines randomly. Another situation I look at is when the public is on the dog. The public has a strong tendency to be on the favorite. When they are heavy on the dog, then for me it’s either the favorite or a pass. Several Neural Models to Approximate Value for the Game This I have found in my experience to be the least predictive- similar to my spreadsheet experience early in my handicapping.
One of the assumptions I make in handicapping is that the lines that the books put out are pretty close to being an accurate prediction to the outcome of the game. But I have found it helpful to have a reliable method of independently coming up with value. And my neural models do that for me. Many handicappers use power ratings. Most of these do not even come close to showing a profit at the end of the season. There is a web site that tracks most of the power ratings (thepredictiontracker.com). It’s a real eye opener in showing the lack of predictive power in most power ratings.
My approach in college football is identical with the NFL with the exception that data over the last 8 years shows the public numbers are only significant in predicting value on home dogs where the public is 23% on the home dog. (Below data taken from sports insights study) Table 1: Betting Against the Public (2003-2010) Betting Percentage Home Team Visiting Team < 40% 49.6% 49.4% <30% 50.9% 49.9% < 25% 51.8% 49.8% < 23% 53.6% 50.2% < 20% 54.5% 45.4% I have also had success with incorporating small line moves opposite the way the public is going with both home and away teams.
But again, this is using the above in my handicapping routine and not stand alone. Another way to determine if you’re on the right track with your handicapping methods is once you come up with what you consider the right side, double check the public numbers. If you find yourself with the public most of the time then you need to find a different handicapping technique.
Now you might ask what kind of profit can you expect using the above techniques. On my web site I have been tracking my plays using the above techniques for almost 7 years. I update the results daily. Prior years records can be found on my site, RickJ’s Handicapping Picks.