Friday, October 24, 2014

Site Terminology and How to Read the Tables

I have had a view people ask to explain how to read my prediction tables. Using the example table from October 26, 2014 pick below, I will point out to understand what I am presenting.

Picks for 10/26/2014

TeamSpreadProb WinnerKellyBetter's Edge
Nevada-3.053.33%2%1%
Hawai'i+3.0

Under the team names, the visiting team is listed first followed by the home team. If the game is at a neutral site, the visiting team for the purposes of the coin toss are listed first.

The spread is the number of points a team is favoured or an underdog. The favoured team has a negative spread and the underdog team has a positive spread. In other words for spread bets, the winner of the bet is determined by the final score plus the spread on the team you picked. So, say I bet on -3.0 on Nevada, and Nevada wins the game 21 to 14. I would win the bet because the score would be 18 to 14 with the spread (21 - 3 = 18), If on the other hand, I bet on Nevada, and the final score was Nevada 9 and Hawai'i  7, I would lose the bet. On the other hand, if I bet on Hawai'i I would have won the bet because with the spread the score would be Nevada 9 and Hawai'i 10 (or 7 + 3).

The prob winner is the probability my prediction algorithm says the selected team will beat the spread. This value is selected by a proprietary algorithm I developed using college football statistics over back to the 2000 season. The success rate of the algorithm historically has been 56% which enough of an edge to allow for you to make money in the long run.

The Kelly value is amount of your bankroll you should bet on a given game if the total among of action at one time is not near 100% of your bankroll. The Kelly bet value takes in consideration the odds of bet (I assuming -110 odds unless stated otherwise) and the probability of a correct bet.

The better's edge is how much you should bet of your bankroll if the total amount of action you have at one time approaches 100% (I consider that anything over 90%). The better's edge is the probability correct bet minus the book's edge [1/(odds+1)]. I have written about this before, that research has shown that the better's edge is more optimal for when you have lots of bets at the same time and the Kelly bet if you have a few.

I hope this helps you with your handicapping.

No comments:

Post a Comment