Comments by "" (@jasonreed7522) on "What Black Friday tells us about parking lots" video.

It used to be common to just sleep in your car and drive home (also drunk driving was a lot less taboo in the 80s) now you have to prove your vehicle reached its present location while you are sober or you will get a drunk driving ticket (making it illegal to "sleep it off" in your car). Now its common practice to have a designated driver who stays sober to drive their drunk friends home. Also, many bars are also restaurants that are perfectly normal to go to as a family and not drink any alcohol. This is different from a nightclub the is exclusively about the alcohol crowd, and nightclubs are basically always in walkable nieghborhoods or are walkable from a local college. As far as the math on parking spaces per unit of usage for different types, that math definitely needs an upgrade to reflect modern times and a "driver factor" (basically account for expected share of patrons to drive themselves vs take a cab, uber, walk, bike, or mass transit)
2
As an EE, it seems incredibly outdated to use averages for building type. In the national electric code the last resort for calculations are the multipliers for building area based on the kind of building you have, normally you go with much more project specific calculations. In the case of a Mall a reasonable calculation would start with the expected occupancy and derate for "other mode share", for instance if you have a bike path near the store and expect 20% of customers to bike to the mall replace 20% of parking spaces with bike rack capacity (indoors and watched by cameras). Expected occupancy probably having its own formula based on a combination of historical data (other walmarts) and building nameplate max occupancy. Just saying a box store should have .1 parking spaces per square foot of shopping area is something you do in the 50s before modern data analytics was possible. (And it sounds like they didn't even collect any data to set the number per unit area)
1