Monthly Archives: November 2016

Basketball Insiders Podcast with Eric Pincus, November 20, 2016

These graphics are for my podcast with Eric Pincus for November 20, 2016. You can find the podcast itself HERE. In one segment, we were talking about Eric’s recent article looking at the league’s current financial landscape. You can find this article HERE. In it, Eric looked at player salaries as a percentage of the total team payroll (i.e., a $25 million salary has a much different effect on, say, Cleveland’s books than Philly’s), and he broke out the money that went to each team’s healthy starters.

Before the podcast I did some quick visualizations of a couple different factors. During the podcast I brought these up and went through them. Here they are, in the order in which we talked about them. All data (including team winning percentages) is current as of today, November 20.

Visualization 1: Team Salary vs. winning percentage (size & color indicates average age). This is the baseline — does spending more money on players equate to more wins? This shows that it does (the trend line slopes upward) although there is a lot of variance (lots of teams outside the one-standard-deviation band). Teams above the band can be described as getting more bang for the buck (eg: Golden State, LA Clippers), while teams below the band aren’t getting their money’s worth (eg: Dallas, Washington). Cleveland and Brooklyn, despite being on opposite ends of the spectrum, and none the less consistent in that both teams are getting what they are paying for.


Visualization 2: Average Age vs. winning percentage (color & size indicates team salary). How much of a team’s success is simply a matter of the age of the team? Should we expect a team of older veterans to out-perform a team of young players? Here the team’s average age is on the X axis, and wins remains on the Y axis. Note the similarity between this visualization and the previous one. Even the slope of the trend line is about the same.


Visualization 3: Percentage of team salary to starters vs. average age of team (size & color indicates winning percentage). Here’s where we start to look at Eric’s premise. The first thing I wanted to figure out was whether the percentage of payroll that goes to starters is simply a factor of age — i.e., would you expect a lower percentage from a younger team (because more players are on rookie scale contracts), and a higher percentage from an older team (because they’re off their rookie contracts, they have higher maximums, etc.)? And sure enough, the trend is that the older the team is, the higher the percentage that goes to the starters. And look at where all the best teams are — Golden State, San Antonio, Cleveland and the LA Clippers are all around that 65-75 percent range.


Visualization 4: Percentage of team salary to starters vs. winning percentage (size & color indicates average age). How does the percentage of payroll that goes to starters compare to team success? Again, there’s a positive trend here, and despite a few outliers, the correlation is pretty strong — as teams pay a higher percentage to their starters, they win more. Note that we can’t confuse correlation with causation — we can’t say that teams win more BECAUSE they pay a higher percentage to their starters. But it does show the effect we were looking at — you find all the best teams in that sweet spot around 65-75 percent. The Warriors, Spurs, Cavs and Clippers are all clustered pretty closely together in this chart.

So what does it mean to be on the right-hand side of this chart? It isn’t that you’re paying your players more — remember, this is the PERCENTAGE of your total payroll. Philly (lowest payroll in the league) would still be on the far right of the graph if 80 percent of its payroll went to its starters. Rather, the effect seems to be that as teams mature and pick up top, veteran players (players who are off their rookie deals), those players (whose salaries are at or near the maximum) tend to drive the payroll. Remember, basketball is driven by the top of the roster, not by the average or the bottom (for those interested, there’s an interesting podcast from Malcolm Gladwell examining this phenomenon, which you can find HERE (this particular podcast episode is mostly about university philanthropy). Given that successful NBA teams typically have the best players (not just a lot of good players, or a few very good players, but one or two of the best players), the effect of these players on the team’s roster, coupled with the supporting casts needed to make their teams winners, tend to gravitate toward a 70-30 distribution of starter salaries to overall payroll.


Anyway, I did this entire thing, from having the idea to entering the data to completing the visualizations to throwing them up into this blog, in about 45 minutes before we recorded, so apologies if I missed any other interesting things to look at.