Betting the spread is what most people mean when they think about sports betting. It seems a little more manly than just picking a winner; any chump can pick a winner (in theory), but it takes real knowledge to know how much a team will win by.
The spread for a game involves two numbers: the spread itself and the pay odds. For example, right now the Packers are -14 at even odds against the Bills, who thus have the complementary values of +14 at -120. The spread is the amount a team would have to win by in order to cover or beat the spread. In this example, the Packers would have to win by more than 14. Looking at it from the other side, the Bills would cover if they win, or lose by up to 13 points. You can think of the spread as an adjusted score. If we take the actual final score for the Packers and subtract 14, would they still have won? If the answer is yes, they covered; if not, they did not (and thus the other team covered). If adjusting by the spread would result in a tie (e.g. the Packers win by exactly 14), that is called a push. In the case of a push, all wagers are returned and no one wins or loses any money. Otherwise bets are paid off according to the pay odds. If you picked the Packers to cover and they do, you would get even money on the bet and thus receive whatever amount you bet; if you bet $100, you would win $100. If they didn’t, you would lose your $100. If you picked the Bills to cover and they do, you would get paid back at odds of 100/120. So if you bet $120, you would win $100; if you bet $100, you would win about $80.
How do you know which side to take? Well, historically it’s been the case that the winning team also covers the spread 80% of the time. So if you can pick winners really well, you can also pick them to cover. But since most games have pay odds of -110, you need to pick against the spread at better than 52.4%, meaning that you have to pick winners with about 65.5% accuracy. This is pretty difficult for anyone to do with any consistency. Another method is to develop some manner of power ranking, then convert it to points when comparing two teams. Maybe you think the Packers are a 109 and the Bills are a 98, so you favor the Packers by 11; you would then pick the Bills to cover since 11 is less than the spread of 14. The way that I’ll do it is with my mathematical model. It predicts how many points the home team will win by, and if that is greater than the spread I pick the home team.