Standard Tools/Skills Used in Front Offices

[Eternal]

Based on my personal experience:

1) Database server and SQL (most common are SQL Server, MySQL, PostgreSQL)
2) R has gotten very common and is growing rapidly
3) Other programming - this I know the least about, but for us it's R (again), Python, Ruby and some C/C++ (as needed)

If you're working with data from a dozen or more sources you need a database server if you're trying to do serious work. R is no longer really an option for this type of work, even if you use other tools, too. Python and Ruby are the kings for web scraping.

Data modeling (linear regression, mixed models, logistic/multinomial regressions, splines/local regressions), machine learning/classification, graphing/visualization.

Understanding model fit and validation is very important. Back-testing, cross-validation, checking model assumptions.

-Chris

[Crow]

Any particular areas of basketball analysis where you would use splines/local regressions or machine learning/classification? Would you suggest anything related to either as some form of enhancement to the current state of art RAPM?

Anything you would say briefly to layman or trained analyst about supervised vs semi-supervised learning approaches?

Anything you (or others) can pull and summarize anything out of these articles for application to basketball analysis?

http://projecteuclid.org/DPubS/Reposito ... 1034276635
http://image.sciencenet.cn/olddata/kexu ... 964602.pdf
http://web.mit.edu/9.520/www/spring08/P ... -ML-04.pdf
http://www.kyb.mpg.de/fileadmin/user_up ... 271[0].pdf