In 2017 I have created betting model for basketball and after 1373 bets we made a profit of close to 70 units. In other words, $100 bettor made $7000 of profit in last 2 seasons. And more than 1000 bets is not small sample size anymore. And what I am most happy at this moment, is that since I started tracking CLV numbers, we beat the line in more than 80% and CLV is more than +4%.
The beauty of the model is simplicity. Anyone can use it, anyone can implement for any league that he wants and I saw some great results from my members, who have this model, also in some other sports (rugby, handball).
So, I was thinking how to take the ideas from basketball betting model and use it for NFL betting.
NFL is a new sport for me and I wanted to give some simple betting model to my members, so they can project the numbers before they bet. It still amazing how many people bet without numbers and without using any statistics at all. But more and more people are aware that betting without numbers will not make any profit on the long run.
After all, lines, spreads, totals, odds and handicaps are nothing else than some kind of probabilities turned into the numbers. And those numbers directly represent the prices on betting market, that you pay (bet) at the end. So, it is crucial to have statistical method, that will estimate those prices even before bookmakers.
This is not rocket science. It is basic concept of making money. You need to know the price before you buy. You need to know if you can sell it later (so you can make profit) for a bigger price. In sports betting that means, that when you see Green Bay Packers -7, you need to know if this is good spread to bet or not. You need to know what is the last spread that you are willing to take Green Bay Packers (in this case).
So, I wanted to get those spreads and totals before bookmakers open their lines.
And this is how I combined my basketball model with NFL stats. It gives me my projected (or expected) lines and totals. Then I simply compare those spreads/totals with bookmakers numbers.
When I see big enough difference between my numbers and their numbers, I simply make a bet.
Because of simplicity and because of very small number of games in one season, I look for big difference and I bet 1 unit per game on each game.
After 11 Weeks the results are:
35 -17, +16.64 units
The question is… can this continue?
From my perspective I simply used the ideas of my basketball betting model and put it here with some adjustments. So, at the end of the day, I get my projections (spreads/totals) before I make final betting decisions. And I am not even NFL bettor. I think this is much more powerful to those who bet NFL on a regular basis, so they get the numbers before placing bets.
The only concern is CLV number so far. But I would also say, that NFL market is very special, because a lot of people with zero betting knowledge (and a lot of NFL “sports fan knowledge”), who don’t use any statistics, analytics or any other mathematical/statistical method bet huge amount of money. It is very small sample size of games and a many people are simply gambling. So, the market is definitely little bit different than NHL or MLB betting market.
However, I will try to improve this betting model in the future and will track my progress to see where we can improve it. I will keep informing bettors what is important to win and on what they should focus, when they bet NFL.
Cheers and good luck in Week 12