Sticking with the thread sub-topic of "Anti-skating, WTF?", starting circa 1965, Dual likely did more to explain things in plain language than any other manufacturer. With the introduction of the automatic Dual 1019 (definitely not your '80s CS505 @Mr. Pig) came a marketing tome that explained the following...
"The Dual 1019 was designed in every aspect for 0.5 gram tracking. This dictated that the bearings in the tonearm pivot had to be well nigh friction-free. (The friction is actually under 0.04 gram.) This added a new dimension to an old problem.
Whenever tonearm bearing friction is less than 12% of tracking force, any stylus mounted in an angled tonearm head tends to run toward the centre of the record. This is caused by friction between stylus and rotating record, and deflects the stylus against the inner groove wall and away for the outer wall. And this is what skating is all about."
Dual called their solution to the skating problem "Tracking-Balance Control" and came up with a device called "Skate-O-Meter", initially as a production tool for quality control of Tracking-Balance Control calibration, as well as being provided as tool to Dual authorized service agents (and also eventually released as accessory). Basically, this is a dial gauge that clips into that period Dual headshell in place of the usual cartridge carrier and contains a 0.7mil conical stylus which was the default stylus tip size and type that the scale on Tracking-Balance Control was calibrated to.
Where this ties in with the discussion of friction, within the 1019 user manual was provided a table of equivalent settings for stylus tips ranging in sizes from 0.4mil to 0.9mil conical, plus 0.2 x 0.9mil elliptical. It is this that mostly explains the friction aspect of skating as the larger the tip size (i.e. contact patch area) the less the friction, and therefore the less the skating force one need compensate for (i.e. less friction due to g/sq mil pressure reduction), with elliptical requiring the highest compensation via having the smallest contact patch.
A couple of others have since produced similar skating gauges, one of which came with an assortment of interchangeable stylus shapes and sizes which prove the friction (and hence amount of skating force) relating to contact patch size aspect.
An interesting side note here is that by the 1970s Dual had shifted their concentric (with the arm pillar) Tracking-Balance Control over to a dedicated Anti-Skating dial with two scales on, one for conical, the other for elliptical, with the manual including columns of alternative lower values for 'wet' play with each tip type. From this it seems obvious that happiness comes with a fat wet one.