I am posting my excel sheet with weekly nfl projections and the record thus far. I have been tracking for the year and have not actually made any wagers. I was looking for some feedback on how well it is actually doing or not doing. Do I have something or is it luck and regressing back to a coin flip? It has not done as well recently as it did in the beginning.
Your attachment is just your results which you already stated. Unless you are lying to yourself and fabricating results it looks very promising considering you have a fairly large sample size.
My rough calculations show your results are roughly 1 in 100 to be winning because of luck. Perhaps I did this wrong though so get somebody else to confirm.
This issue is addressed in each of the three major books on sports betting (Stanford Wong, King Yao, Justin7). The Stanford Wong book has a table titled "Rarity of Good W-L Records" and for 500 samples, it shows the chances that the following W-L records were created by chance:
Wong recommends that if you're using historical data, hold out for 1:1000 chance of a W-L record being by chance alone. At 1,000 sample, the 1:1000 record works out to be 549-451.
NFL Projections - Feedback Please
Reply With Quote