NBA salaries
I began wondering, how much does performance affect your salary? If I’m a super-duper programmer, do I get a super-duper salary? It’s really hard to gauge a programmers abilities. There’s so much that goes into being a “good” programmer that it’s a very difficult problem. So much so, that we now have companies that programmers pay to get them up to snuff, straight out of college. The hiring process is broken for programmers. That’s a deep rabbit hole for later.
I decided that athletes’ performances is directly measurable to a good degree. I went out, and found a dataset that gave me a good aggregation of the top NBA players in the spans of 2005-2014, somewhat dated I acknowledge. Nonetheless, It’s an interesting dataset.
The mining Op.
This set was gathered from the folks over at superdatascience. The python scripts used to generate these plots are over on my github. I used numpy and matplotlib for this mining operation.
Salary
In the first plot, it’s interesting to see the divergence between Kobe Bryant (RIP) and the other top players. Even Lebron James didn’t touch Kobe for nine years! Even then, Kobe still made more dough!
Field Goals percentages
Dwight Howard killed it in the “field goals per attempt” arena. He, at one point, had an accuracy of over 60%! I mean, that’s amazing! Basically, if you’re his opponent and he goes up for basket, you might as well spit in his eyes, because that’s the only chance you might have of stopping the guy.
Accuracy
Dwight Howard was also in the median payscale for all these guys, in the beginning. In 2005-6 he was in the top of his class for accuracy, however his “shot attempts” were fewer that the rest of the players. Though, it looks like he had a decent agent. After his performance in the 2005-6 seasons his salary more than doubled and it continued to increase. Even though, his field goals per game varied by a approximately one attempt.
Conclusion
There are many factors that need to be consider for an analysis on this data. Position is a big one, injuries is another and then there’s a lot of smaller factors. Some of Kobe Bryant’s best performances were when we was less of a household name. Meaning, that to some degree, as you get more fame you get covered a lot more. Naturally, if you’re a young upstart prodigy it’s going to take a season or two before other teams start putting their best against you. Once that happens you’re going to have less easy shots available. Which will result in a drop in stats.
These are just some interesting findings and discussion points. These findings would be a great starting point for some serious deep data mining.
Background and thumbnail image by: Olamar Gibson
Copy Link