Author: Gary Mulder 2015-07-30 16:56:23
Published on: 2015-07-30T16:56:23+00:00
In an email sent on July 30, 2015, Jorge Timón proposed a blocksize hardfork that uses median time instead of block.nTime for activation. He noted that this was different from previous proposals and suggested discussing the use of height for all proposals in a separate thread. In response, Gary suggested using a statistician to help select the most appropriate statistical algorithms for the proposal, considering the consequences of choosing a "bad statistical model." He also mentioned the possibility of using an exponentially weighted moving average, which is less volatile and harder to manipulate than a simple moving average but still responds to short term drivers. The email includes links to resources on percentiles.
Updated on: 2023-06-10T04:27:57.001708+00:00