Posted: 5/6/2013 4:27:10 PM
Originally Posted by BuckeyeKaptn:
That is ^^^ if no one is willing to step up to the plate, I guess I can do it, if the spreadsheet for Buddha is like the others.
Don't agree until you are sure you believe in the methodology of the system
I uploaded a file that has the system details all in one document and the URL's of the original threads on Covers.
The system is based on Sagarin Ratings, which makes it difficult to backtest since they are not archived and the math involved to calculate Predicted ERA, MOB and RICKEY is highly advanced - so to recreate the formulas in excel may be difficult - but that being said - if someone does know how to calculate those 3 things then this system can be backtested. (but I doubt it)
So if this cannot be backtested - the only option is to test for the season.
The premise of the system is to look at the results of like situations using Starting Pitcher Stats and Odds. The author uses the theory that in the same situation similar teams will perform the same (i.e Win or Lose)
He uses the Sagarin Stats to come up with a Composite Score for both pitchers and classifies the FAV team as Low, Med, High Fav. and records the result of the game (Win or Loss). He builds a DB of these type of games until he has a enough games to use to predict the outcome of a game with similar numbers.
Favorite Type: HIGH
Dog Pitcher Score: 40
Home Pitcher Score 50
Let's say the above situtation has come up 5 times and all resulted in a WIN for the home favorite and a game comes up that has the exact same numbers. He would bet on the Favorite - since in those situations - it has yielded a high win rate. (100% to be exact)
Now I have no idea if this works or not. The premise is interesting - and it sounds logical. Does it work? No freakin' clue. Would like to know if it works? Defintely. Do I have the time to run a thread and track the results with all the other projects on my plate? Unfortunately no.
So before you jump in and say you are willing to track this (which I highly commend by the way - I admire people who don't sit back and expect things to be spoon fed to them, give them some guidance, advice, tips and support and let them run with it. You will learn more in process and it may spark creattive constructive thought in the forums) - I advise you to read the system doc, and then read the orginal posts on covers, then read it again.
Then ask yourself - "Will I devote the entire season tracking this, and share any insights them comes out of it with the group on a daily basis"
Remember - I will provide the spreadsheet - you will just have to click update each day - and look at what the plays should be and record if the predicted plays win or lose.
If the answer is "YES" let me know. Just writing this post got me even more intrigued if this works or not - so I hope the anwer is "YES"