A new approach for Bitcoin pool-hopping detection

Mining pool hopping is a phenomenon taking place in cryptocurrency networks such that a miner changes of pool overtime, in a regular or recurrent way, to increase its gains related to mining work rewards by the mining pools. This phenomenon is not well understood, also because of the lack of a precise pool-hoppers detection solution. In this paper, we propose a methodology for detecting the pool-hopping behavior in the Bitcoin network; we propose a deterministic framework exploiting the different time windowing phases (rounds, epochs) involved in the rewarding process. Our methodology includes a new algorithm to identify the miners, and a new algorithm to trace the revenue stream distribution. We assess the performance of our approach for the five mining pools with the highest hash rates during two three-month period in 2020 and 2021: we show that the phenomenon is still advantageous in terms of overall gains for the pool-hoppers. We also assess the fairness in stake: the phenomenon was known to be unfair in the beginning of the Bitcoin network due to the simple rewarding methods in place at that time, with single rewards higher for pool-hoppers than for miners; we show that this is no longer true and that the new rewarding policies now guarantee that pool-hopping is fair with respect to miners that do not perform pool-hopping. Nonetheless, we also show that the cumulative gain over time of pool-hoppers can be higher by 33% on median than static miners.