Why Is The Line So Accurate?

Originally published in the Telegram channel on December 9th, 2022…

This gets a little mathematical and technical, but it will help you to see why and how sometimes the oddsmaker nails the line spot on…or very close to the actual outcome.

I was asked how bookmakers can get the number so accurate. It is a combination of math through algorithms…both the book’s algorithms for setting the initial line, then sharp money’s algorithms for how they bet and further shape the line combined with old fashioned standard deviation.

Denver -1 (NBA) yesterday winning by 1, which was the closing line, was not as it wild as it might seem. People always remember when a game finishes on the number. People don’t pay attention to the occasions when the number is off 3,4,5+ points from the actual result. It’s more luck than skill when a number falls on the number.

Want to get behind the numbers? Check out TSP Live at https://TSP.Live/tsp-live!

For example, today (I am just picking a random game) we have Cleveland -5 over Sacramento.

Standard deviation assessment would put the score differential (2 standard deviations from the mean = 95% of possible outcomes) anywhere from Cleveland by 21 to Sacramento by 10. It gives us 31 possible score outcomes…

Cleveland by 21, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1

Sacramento by 10, 9, 8, 7, 6, 5, 4, 3, 2, 1

So, if we assume the outcome of tonight’s game is within 2 standard deviations, then we have a 30-1 chance of just randomly selecting the exact score differential…no skill involved…just luck. Throw a dart…Sacramento by 2!

If we just go out only 1 standard deviation (1 standard deviation means 68% chance of getting the score differential within that window) then the possible score differential window becomes anywhere from Cleveland by 12 to Sacramento by 2.

It makes the possible outcomes…

Cleveland by 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1

Sacramento by 2, 1

There are 14 possible outcomes above (1 standard deviation)…meaning if the game ends with a result 1 standard deviation from the mean, there is a 13 to 1 chance that Cleveland wins by 5.

Furthermore, there is a 3 out of 14 chance that the game falls on or within 1 point of the spread (Cleveland by 6,5,4 since Cleveland is -5). Meaning the book has a 21% chance roughly 5 to 1 of looking really intelligent and getting the spread on or within 1 point of the actual outcome.

When you start to look at all the games on the board, there will undoubtedly be games that fall right on the number for the push. When you zoom out, you will see these games which end right on the number or within 1-2 points of the spread aren’t really magic, it fits within these standard deviations and probabilities. When you are putting prices on so many games per day, says 90 in a college football weekend…the book will likely nail 3 to 6 out of those 90 perfectly. In an NFL football weekend with 16 games, the book will get 1 games on average every two weeks to fall right on the number.

Books can’t see the future, they don’t know anything…well, other than math and probabilities of random outcomes. Sometimes in a game of numbers that’s all you need!

Good luck!!