Moe Zhoost

I made no statement to validate. Nor do I have a strong opinion about the topic one way or another. Had I posted an opinion, though, I would have wanted to be able to back it up. I asked for documentation specifically because I like to learn all I can about the engineering and technical aspects of bicycles.

Archibald Sharp, in his book "Bicycles and Tricycles - An Elementary Treatise on Their Design and Construction", has a great discussion related to bearing friction due to mutual rubbing of the balls. This book is a bit dated (1896) but shows that this question was being discussed very early in the modern bike era. His bottom line is that caged or uncaged, the friction due to ball-to-ball or ball-to-cage was minimal. He did mention that cages were quickly abandoned in manufacture specifically because a less balls could be put in the bearing rendering it less reliable than uncaged. He points out that wear on well adjusted and lubed bearing balls, caged or not, is not significant.

Personally, I usually use caged balls, but when replacing balls will sometimes opt for loose when I don't feel like cleaning and refilling the cages. I often use loose balls for headsets because I think it helps avoid indexing, but also because I love the challenge of getting everything together with so many small items that go easily astray.