What is a good sports bet? The answer to the question involves a really quick refresher on the principles of value betting.

Just as in all other types of successful gaming, the idea in sports betting is to find "value" in a betting proposition. That means finding a proposition in which a given outcome in the real world (sometimes called the "true odds" or the "physical expectation") is more likely than the implicit financial terms contained in the betting proposition. To review this quickly in the context of the coin toss, the physical expectation is 50%. If the bet provides a payoff of more than 1:1, then it is a good bet, and in the long run will add to the bankroll. If the odds are 1:1 or less, the bet is a long run waste of time or worse, a loser.

Thus, in explicit terms, "value" is the prospect of a monetary reward that exceeds the risk of loss by at least enough to compensate for the transactions costs. These transaction costs, in the context of the sports book, are fairly specific. They are the "vig" collected by the book.

Return on Investment - Edge

Another way to think about "value betting" is to think about "edge". "Edge" is the expected profit from a bet. It is often expressed as a percentage, and is often called the "return on investment."

"Edge" is calculated by taking the net profit from the bet and dividing it by the cost of the bet. How is "profit" calculated? It is the expected monetary value of the outcome (total proceeds times the probability of winning them) less the cost to play.

Consider the example of a totals prop on Sunday's match up between the Browns and the Bears in Chicago. The book is prepared to pay 10:11 odds ($100 of winnings on a $110 bet) if the total number of points scored in the game is more than 40. Sometimes it is hard to articulate how the bet is framed in terms of a spread or total. When the total is "40" or the spread is, say, "-3" it means that the wagerer bets against the house prop. Betting against the spread means that the favorite will win my more than three points. Betting against the house on the total proposition is to wager that the two teams will score more than the number proposed by the house. A tie with the house at 40 total points might trigger a refund, or it might go to the house, according to the rules of the book.

To beat the house vig in this proposition, the physical expectation of the winning outcome would have to be more than 11/21 (the $110 cost of the bet divided by the $210 payoff) or 52.4%. If handicapping shows a lower probability, say 46%, then the edge is negative. But if you conclude instead that your best estimate of the probability of 41 or more points in this game is at least 56%, then you might consider the wager. You will need to consider the likelihood of an outcome at exactly 40 points if the house refunds bets on a tie. Suppose that this calculation is 3%. (The chance of a close game like 23-17 is remote, but certainly reasonably possible.)

The edge is calculated first by taking the "real" probabilities of outcomes times their returns and summing them up. Thus: 0.56 times $210 (win) plus 0.03 times $110 (tie) plus 0.41 times $0 (loss). The total is $117.60 plus $3.30 or $120.90. (Purists will want to round this to the nearest dollar, as we're only working with three significant digits.). So the "value" of the ticket is $121. The profit, therefore, is $11, obtained by subtracting the ticket's cost from its expected value. The "edge" expressed in percentage terms is 11/110 or 10%. Not bad.

So, to summarize, the "essence" of successful sports betting is to find those betting propositions that offer a positive edge greater than the house vig, and bet them.