Originally Posted by bellweatherman
What's so great about n-1? Isn't the idea to spread the load as best you can? So, therefore n minus anything is going to spread the load alot worse than n minus nothing. And if your goal is to use less ball bearings, then why even through away the retainer?
I think something that's being misunderstood in this thread is what "n" means. "n" is the number of bearings it takes for the first occurence of the bearing set to jumble. "Jumble" means that when the bearing set is test seated the loose balls do not lie in a perfect circle, because there are too many bearings to allow it. So, you eliminate one bearing until you can get the loose balls to line up in a perfect circle. This will almost always end up with the number of loose bearings at least one more than you would have in a caged bearing set. The goal is to use
more bearings and distribute the load more widely. Not
less, as you stated.
In other words, another way of looking at what "n-1" means is that "[(first occurence of jumble) - 1]" would be optimum for hubs and BBs.