Let's say I have a data set which shows me basketball players: list of their games which contains date, number of shots they took and number of shots they scored. For example:

1.10.1992 Michael Jordan took 12 shots - scored 6

I would like to predict what is going to be their scoring success rates in the next games using the historical data set. What is the best way to do it? The real model I'm doing it for has very low success rates < 1%.

  • $\begingroup$ When you say 'predict', what is it you want to get? A mean prediction? A median prediction? A prediction for some particular loss function? Do you need a confidence interval for (say) a mean? Do you need a prediction interval for a new observation? $\endgroup$ – Glen_b Jun 16 '14 at 11:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.