it works according to power, RPM, and ground clearance.
The optimum propeller would be a single blade to create a maximum active thrust area.
This theoretical ideal prop would, of course, make for one rough ride if it stayed together, or rather, a very rough ride till it came apart.
There was in fact a single blade prop made...It was tried out on a Cub...The one side had the full length of a prop. The other side was just a weighted stub, to counter balance...Not too good of a idea, but it did work.
WuzzFuzz
two blade is the real world ideal propeller, as it can be balanced and provide the largest thrust area.
The glitch is that as engine power gets up there, that acting disk grows as well, until it extends into the tarmac in ground operations.
To put the power to use, we add more blades to offer a suitable load for the engine, while keeping the prop out of the ground.
Add RPM into the mix and theres more fun to be had.
When the RPM is high enough, they can, and do, drive the prop blades into supersonic speeds.
The result of this is a dramatic loss of efficiency.
Thus for a higher RPM, the prop size must be reduced in order to keep the tips out of the sound barrier. Again, looking back at engine load requirements that must be met, the only option is to add blades.
this, omits the attribute of prop pitch .. which is yet a whole other bag of rats adding to the prop selection equation, which was purposely omitted as the above is sufficient to gain a fair grasp of the general concept