So for a senior econ class a buddy and I are trying to write a regression that will help determine the "BEST NBA PLAYER OF ALL TIME". Everything that has ever had to do with ranking players is based on polls of journalists, players, coaches ect... however that is extremely subjective and will lead to biases if no one likes you.

We are trying to objectively, as best we can, figure out who the best player ever was and we some help coming up with variables. Again we want this to be as accurate as possible so we need a small error term which means we need variables and variable and variable!!!

The problem we are running into is how to judge say Bob Cousy vs Allen Iverson. Athletic ability now vs athletic ability back in the day.

Thanks!

I would start by taking a look at Hollinger's PER formula, and either tweaking it to your liking, or just using it as a reference to get an idea of what you can do with stats.

What variables have you guys come up with so far?

http://www.82games.com/

this site may help

Another useful tool, which I'm sure you guys have already looked it, is basketball-reference.com.

They have their own probability model for the Hall of Fame which will probably share a lot of the same statistical variables with a model you're trying to create.

http://www.basketball-reference.com/about/hof_prob.html

[This message has been edited by Head Ninja In Charge (edited 4/9/2014 9:12a).]

Win shares. You're welcome.

quote:

---

The problem we are running into is how to judge say Bob Cousy vs Allen Iverson. Athletic ability now vs athletic ability back in the day.

---

Its not just athletic ability, but also rule changes and quality of teammates and opponents.

quote:

---

quote:

---

The problem we are running into is how to judge say Bob Cousy vs Allen Iverson. Athletic ability now vs athletic ability back in the day.

---

Its not just athletic ability, but also rule changes and quality of teammates and opponents.

---

Haha! We are debating whether or not Tim Duncan or MJ had better teammates