I was just thinking about this. I know why the random selection exists, but I don’t know how it’s exactly coded. As more nodes come online, the odds of you getting randomly selected is naturally going to go down simply due to volume. Is there a way to set a finite amount of days without getting selected before it is automatically selected? Something like 45 days and then it is put into rotation at least once? Or, even on a sliding scale. Maybe it goes up 1 extra day every 100 or 500 nodes that come online? I was at 65 days, beat my own record, and then had two others immediately beat me. Or, are we way past the point to tweak the selection process?
Maybe I am wrong, but setting a limit like that makes the selection predictable, which is what the random selection was supposed to prevent.
Rotate more than the currently 4 nodes per epoch would make the odds slightly better.
There are pseudo RNGs like sobol that incorporate memory based functions, so the same number getting selected twice goes down… I used this is some of my simulation work to ensure I was getting a random selection of surface sites, but also ensuring the surface was fully selected at some point. So I imagine something like this is feasible, but I dont know how easily a sobol rng can be incorporated into node selection, as I am sure there are many things about the rng and selection process that have to be considered.