RealGM

Hey all

I'm in a Econometrics of Sports class right now at college, and being a big basketball fan, I wanted to do a paper topic related to basketball. Any ideas?

I should note that while I understand the basics of regression analysis and have a written a research paper using econometrics before, I've only really taken one Econometrics class so far, and would consider myself a novice, certainly not my area of expertise. Looking to do something manageable yet interesting enough to do. We are allowed to replicate existing research papers or base our papers on them since replication of work is itself an important tool.

Doesn't have to be complex, ground breaking stuff, just enough meat in it to write a paper, any cool ideas to point me in the correct direction would be helpful. Right now, I'm thinking of doing something related to team performance or salaries and individual performance.

Thanks in advance!

Have you thought about examining the relationship between salaries and points per game? I would look at the following scenarios:

-points per game in 1, 2 and 3 seasons before a new contract (and then the correlation between that contract value and ppg)
--I would run the same analysis on scoring related stats, like ppg + FG%, ppg + TS%, USG%, USG% + FG% and even PER.
--I would also control for team record. This means you can examine the relationship b/w ppg and salary in general, but also based on how good the team was for the player producing the high scoring.

Finally, you can test these data against the idea of a contract year.

Also, where was "econometrics of sports" when I was in college?

bballcool34 wrote:

Spoiler:

Hey all

I'm in a Econometrics of Sports class right now at college, and being a big basketball fan, I wanted to do a paper topic related to basketball. Any ideas?

I should note that while I understand the basics of regression analysis and have a written a research paper using econometrics before, I've only really taken one Econometrics class so far, and would consider myself a novice, certainly not my area of expertise. Looking to do something manageable yet interesting enough to do. We are allowed to replicate existing research papers or base our papers on them since replication of work is itself an important tool.

Doesn't have to be complex, ground breaking stuff, just enough meat in it to write a paper, any cool ideas to point me in the correct direction would be helpful. Right now, I'm thinking of doing something related to team performance or salaries and individual performance.

Thanks in advance!

Another idea could be the correlation of depth of "high-value players" and team success. When I say "high-value" I'm not talking strictly about playing ability, I'm talking about the number of players on "good deals" a team has.

Without diving too deep this could mean the number of 20+mpg players on a roster along a paradigm of PER (either of current or prior season) & Salary (assign the paradigm weights yourself) with team win totals for RS and PS.

Does stacking the deck with 7-8 high-value guys breed better results or are there, for example, diminishing returns after 4 or 5 players? Or maybe teams with medium-value high production players (due to bloated salaries) are more valuable then teams with several high value/medium production players? I'd limit the research to data post-'99, prior CBAs might skew some of your research.Just some potential things to explore.

Whether or not you use any ideas found here I'd love for you to post what the topic you eventually select ends up being. The fact that there's a class about the econometrics of sports out there fascinates me.

ElGee wrote:Also, where was "econometrics of sports" when I was in college?

Seriously, right?

^Trying to do something similar to ElGee's idea: new contracts and mix of advanced/normal box score stats under the last CBA. Having trouble downloading any files with TS%, asst%, etc. for each player season by season, if any of you guys know a site that has this to download by season rather than by individual team per season (which is what bballref seems to do), let me know.

And yeah, it's definitely a cool class, solid senior time

basketball-reference.com should be able to help you