Playing the numbers by understanding the data

One of my dreams one day is to build a suite of programs used to bet on the outcome of sports games and win. I do know that it is possible to store any and all statistics related to a particular player. You can keep information about teams (which are made up of players) and actually play the teams against each other in a fantasy statistics game. This module would also have a component to run simulations at the team level that would ignore the gritty information at the player level.

Another part would track my betting history and my total expenses and my winnings. This part probably would be used to test the system before betting any actual money. I suspect that the minimum test time should be a season for each sport.

Still another part would scrape information off of the web and compare the different numbers and find the highest paying bet on a particular game I wanted to place money on, comparison-shopping if you will. I could also track the odds that the odds makers give and compare them against the finial outcome, making sure that I remove the poor performers and reward the good ones with my money.

The system could be built, though I have doubts any real money could be made. Though by my calculations, if a small percent were made daily, it would add up very quickly over time (a few years). Just think about what 1 percent compounded interest daily could do with 1000 dollars.

In that spirit, I would like to share this article.

"No, with the title of this column I am not referring to your favorite league. You can decide that on your own.

I’m referring to baseball wagering — specifically, the general difference in results between the National League and the American League. When analyzing baseball, it usually is not a good idea to blend the results of the two leagues into one statistical pot. Here’s why.
There is often a pronounced difference in overall results when comparing the two leagues. For example, let’s examine some differences for the games from the start of the 2002 season through May 20.

In the National League, when the home team has been favored, it has won 59.1 percent of the time (133 times out of 225). In the American League, when the home team has been favored, it has won 63.8 percent of the time (104 of 163)."

I also wonder the effect that technology has had on betting, especially the numerical analysis of the stats and numbers.

Cube