Related Markets

For most American sports, sportsbooks deal three major markets for each game. They deal one total market, and two sides. The total represents a median outcome of the game. Bets on either side are priced at close to 50 percent break-even (with vig added from there).

In football and basketball, the sides are a money line and a point spread. The point spread also represents the median outcome of the game.

In baseball and hockey, the sides are a money line, and a run or puck line, where the run or puck line are fixed at 1.5 and do not represent a median outcome.

Mathematically, there’s nothing special about these three markets. These markets are chosen because they’re interesting to bettors. “Who will win the game?” is the core bet on any sport, so money lines are always interesting. But when one team is a huge favorite over the other, you get into that situation where you have to lay a big price to bet the favorite, and I’ve already covered why that makes the betting less fun for most people.

Both the point spread, and the run and puck lines are designed to offer an alternative interesting bet when the money line break-even percentage on a favorite is say 60 percent or higher. That’s all they do. There’s nothing magic about them.

Whenever the game is equally matched to begin with—say the money line break-even percentage is around 54 percent on the favorite—there isn’t much purpose for the point spread or run and puck line markets. At least not for most bettors. It’s easy to bet either side of the money line, and the second line is sort of redundant. Sports books offer the markets anyway on these games because why not, and people do still bet them.

One thing that makes multiple markets on the same game interesting is that the prices are related. Prices between any two different side bets will always be strongly related to one another. (The pricing on side and total bets are also related, though the strength of the relationship can vary from extremely strong to barely at all depending on the sport and teams involved.)

Think about an NFL game where the point spread is currently the home team -1. Books are offering either side of this bet at -110.

What should the money line market look like?

A money line bet on the favorite and a -1 bet on the favorite grade exactly the same way in all game outcomes except for two: the favorite wins by exactly one point or the game ends in a tie.

If the favorite falls by one, the spread bet pushes, but the money line bet wins. If the game ties then then spread bet loses and the money line bet pushes.

Because the bets are so similar, it’s obvious they should have very similar break-even percentages. The money line bet on the favorite wins slightly more often than the -1 bet. So the money line bet should have a higher break-even percentage than the -1 bet—but by how much?

It depends on the relevant push rates. This is the percentage of time each bet will push—that is how often the game lands on either zero (a tie, pushing the money line) or one, pushing the spread bet.

There are various methods for bettors to estimate push rates. A simple one would be to look at all the NFL games with similar lines from the past and see how often these games tie or land on the favorite winning by one.

For this, you need data. If you want to win at sports you need at least some access to data. I don’t think it’s realistic to have no data at all at your disposal and expect to find any edges consistently. The type of data you need (and the difficulty in obtaining it) varies based on the type of edges you are looking for.

But the most basic dataset you will need is a database of game scores (by quarter, half, inning, or period) and closing lines. For this NFL example above, say you have a dataset of NFL games. For each game over the last ten years, you have each team’s scoring broken down by quarter, and you have the closing point spread and total for each game. (You would choose from the lines at books known for making the NFL markets when choosing what spreads and totals you used.)

How to acquire a dataset like this is a bit beyond the scope of what I want to talk about here. But no doubt you can find simple datasets like the one I described available—and searchable—either for free or pay on websites.

Once you have your data, you look at games where the game closed around pick—maybe games where the home team closed somewhere between +2 and -2. Then you look at what percentage of the time those games ended in a tie or with the home team winning by exactly one point.

For the sake of argument, let’s say 0.5 percent of those games ended in a tie, and 2.5 percent of those games ended in a home team win by one point.

Let’s say you think a 50 percent break-even money line is a fair price. That is, you think the game is an absolute toss-up. What should the break-even price on the home team -1 be?

If the money line is dead even and the game ties 0.5 percent of the time, that implies that the road team wins 49.75 percent and the home team wins 49.75 percent (and a money line bet pushes 0.5 percent).

If we assume that the home team will win by exactly one about 2 percent of the time, then we have our estimates of the four relevant outcomes as follows:

Road team wins 49.75 percent
Tie 0.5 percent
Home team wins by one point 2.5 percent
Home team wins by more than one point 47.25 percent
Therefore, a bet on home team -1 would win 47.25 percent and lose 50.25 percent while pushing 2.5 percent. The break-even percentage is therefore…
BE% = 47.25 / (47.25 + 50.25) = 48.5 percent or about 106 in American odds.
If the money line is even (100) then the home team -1 should be about +106.

Working the other direction, say you know that -1 has a fair break-even percentage of 50 percent. What is a fair money line?

In this case, the four outcomes break down this way:

Road team wins 48.25 percent
Tie 0.5 percent
Home team wins by one point 2.5 percent
Home team wins by more than one point 48.75 percent
The home team will win the game 51.25 percent, lose 48.25 percent, and tie 0.5 percent. Therefore, the money line break-even percentage is
BE% = 51.25 / (51.25 + 48.25) = 51.5 percent or about -106 in American odds.

Final Thoughts

The key concept is that the point spread and money line are related markets. It would make no sense for an NFL home team to be -1 -110 on the spread, but -135 on the money line. There’s a strong relationship between these prices that’s determined mostly by the rules of an NFL game. This concept of related markets is perhaps the single most important one to understand if you want to win at sports betting. ♠

Ed’s newest book, The Course: Serious Hold ‘Em Strategy For Smart Players is available now at his website edmillerpoker.com. You can also find original articles and instructional videos by Ed at the training site redchippoker.com.