Full details, alongside the #Python code and the raw data are available in the article here:

towardsdatascience.com/what-statist...

One noteworthy caveat is that for many of the tests performed, statistical power was low. There were so few coaches hired from outside the NBA that it's hard to say if they truly do better or worse.

My dataset only stretched back to 1990, a bigger dataset have be able to provide more clarity.

The article also asks if internal candidates perform better than external hires. After all, both Gregg Popovich and Erik Spoelstra were internal hires!

But the data does not bear this out. If anything, internal hires actually seem to do worse (though this is not stat. significant either)

If anything, the evidence suggests that previous head coaches perform better than coaches from other backgrounds—they stay in the job longer on average and are more likely to outlast their initial contract, though neither of these differences are stat. significant

Lastly, is there any evidence that #NBA teams are "playing it safe" by hiring known quantities and overlooking better candidates elsewhere? Not really.

How long do #NBA coaches keep their job? Not very long—the median coach is out of work after just two and a half seasons. Nearly one in five doesn't last more than one season, and barely 10% are still employed after 5 seasons.

First, who gets hired as an #NBA head coach? Mainly previous head coaches. More than half of hires already had experience in the role. Among newcomers, most were NBA assistant coaches. Only about 10% of hires happen outside of these two categories

See for an example of one such paper from the US that uses estimsted demand curves to measure shortage: docs.iza.org/dp15447.pdf

I actually looked into the methodology, and it’s based on trends in household formation. These methodologies badly understate shortages at the local level since they neglect the potential for mobility.

Papers that use a supply and demand framework often arrive at estimates much, much higher

120K, even if all Vancouver, seems absurdly low to me. I would bet strongly that if Vancouver loosened regulations, they would see an increase in housing supply several times that.

I don’t know what the methodology is here, but this would imply little shortage in Vancouver, which seems implausible to me.