They are pool hoppers. If they have no "hoppable" pool to hop to, they have been sending their hashes to us lately.
I know but somehow it seems like they are connected. Just wondering why they come to us almost at the same time.
Most hoppers are using the same database of pools, same algorithms, and same software.
When the code tells one hopper there is no pool which can be exploited and to fall back to the "fair value pool" the same situation is occurring for all hoppers at the same time.
There is some variation because hoppers might be using slightly different algorithms or have slightly different sets of pools but generally speaking when hoppers are falling back to us all the hoppers are falling back to all of their backups pools all across the network. It means for that period of time that they are here, no victim pool presents >1.0 share value (efficiency). As soon as the code identifies a pool to exploit they go there and keep jumping to the most exploitable pool until no pool is exploitable and then they come back here.