Sports Betting

1 September 2019

By John Grochowski

My friend Gary is a retired science teacher I’ve known since college days. He’s not really a casino player, but he does love football with a little something riding on the outcome.

Mostly, that takes the form of confidence pools where you pick winners in order of confidence and 100-square pools.

The spread of legal sports betting has Gary intrigued enough that casino ventures are on his horizon. He says he’ll make some football bets, but he’s curious on one point:

How much does Vegas really know?

By “Vegas,” he means odds makers, and they’re not all in Vegas. They could be in New Jersey, Delaware, Mississippi and other states, and they could be offshore or overseas for online sites. But legal sports betting in the United States was confined to Nevada for so long the sports bettors’ mantra of “Vegas knows” is well entrenched.

“It seems to me that every time there’s a late score that puts the result right on the point spread or within a point or two, the guys I watch the games with put on their serious face and say, ‘Vegas knows.’

“Nobody ever says ‘Vegas doesn’t know’ when somebody beats the spread by a couple of touchdowns. It started me wondering how much of ‘Vegas knows’ is selective memory. How often is the result really within a couple of points of the spread?”

I told Gary I really didn’t know, that I thought “Vegas knows” was a product of people expecting Vegas to know, commenting when the spread was spot on and ignoring when it wasn’t.

After all, sports books’ profits aren’t dependent on the narrowness of difference between spread and result. When setting a spread, the goal is to attract something close to the same amount of action on both sides. If that happens, the 10 percent vig assures the house of a profit.

But that was just a feeling. I didn’t have any numbers. So Gary and I decided to see what we could see, and I got out my trusty laptop.

One of the first things we found was a site called betsportslabs.com, and it had done a study of the 2010 season that was just what we were looking for.

It found that in 1,494 games, the average distance between the spread and the result was 10.5 points – hardly what we’d think of as “Vegas knows” territory.

You can have a 10.5-point average and still have large numbers of instances where the line is spot on because of the effect of blowouts. In a small sample of 10 games, if there are seven in which the spread and result match and three games in which the result exceeds the spread by 35 points, then the total difference between spread and result 1s 105 points, or 10.5 per game.

But that’s not what was happening in those 1,494 games.

Instead, only 2.7 percent of results matched the spread, and only 15.9 percent were within two points.

The median was between seven and eight points. The spread and median were separated by seven points or fewer 46.2 percent of the time, and by eight points or fewer in 50.9 percent of games.

“A full touchdown-plus is not what I call ‘Vegas knows’ territory,” Gary said. “But like you say, they’re not trying to predict the result. So you wouldn’t say they’re getting it wrong.”

No, they’re not getting it wrong at all. As long as the line is drawing the needed action for the house to make a profit, they’re getting it right.

Look for John Grochowski on Facebook (http://tinyurl.com/7lzdt44) and Twitter (@GrochowskiJ).