Saturday, January 15, 2011

My new projection system

I've have spent hundreds of hours trying to get this projection system to work. I think I finally have a usable system that I can be proud of. Using nothing more than a stat line, I run it thru a system that places into a ranking. The next step is to age adjust these ratings using historical information to approximate what players should do. The lower the player is in the majors, the worse he is likely to do, but some players project well, even from low a and rookie ball. I have created a percentage that is a realistic approximation of how likely a player is to be a major leaguer. It isn't based on if he was actually called up for a cup of coffee, it just shows how likely he is to be promoted. If a player is over 80% it is likely he will get a shot. Over 100% means he could make a solid impact. I have this percentage for both hitters and pitchers.
I don't have a good automated projection system for pitchers yet and I think that will take more time, but I'm going to work on it.
As for hitters, I have created a system that predicts what a player could do in the majors right now. It isn't a full stat line, I just created a BA/OBP/SLG slash line for each player. I have this for players from DSL all the way up to the majors. I haven't done a ton of work to verify the accuracy. With the players I have looked at, the best season a player has in his minor league career, typically his AA or AAA season, correlates fairly well to his future career average slash line.
The two slash lines I created for each minor league player are what a player could likely do in the majors currently and what his future output could be.
For major leaguers it is likely the worst they will do and the best they will do.
I will post some of this info on the blog, maybe projecting the future lineup for the Royals, since they are so stacked.

No comments:

Post a Comment