a side-project in Machine Learning that tries to predict good bets on sports using real Vegas odds
- virtualenv
- serverless
- docker (for packaging python dependencies with serverless)
psst, here is a little cheat sheet for using python 3 with serverless.
mkvirtualenv sports-bet -p ${path/to/python3}
pip install -r requirements.txt
Use serverless
export AWS_PROFILE=personal-serverless
serverless create --template aws-python3
Scoresandodds.com has some helpful tips on reading and understanding the published odds. Some important defaults to keep in mind:
- MLB run line (aka, spread) is always 1.5 runs
- Most bets use a default of 20% fee between the favorite and underdog. For example, if it is a 50/50 toss up, then each
team will cost $110 wager to win $100. If one team is favored at
-120then the underdog would be+100
Scoresandodds.com has a roughly 24 hour history of the trends of the odds for a given matchup. You can find an example at http://scoresandodds.com/linemovement/20180620/387097277