I am currently analyzing the reward distribution over several cycles for bakers and delegators. To start with I tried to figure out a mean reward for the bakers over all cycles (baseline) to compare the actual rewards with. This "baseline" is the mean reward for all bakers over all cycles and indicated as the blue line in the graph. Furthermore, I have the real rewards which are the actual rewards averaged over all bakers per cycle. The interesting thing is that there are several significant changes in this graph, which show either a big increase or a big decrease on rewards at a certain point in time (i.e. at a certain cycle). For example at block 150 and at block 200 there is somehow a big change. Also in the beginning the line is very straight (showing that the rewards are really equal all the time until cycle 150). I explain this behavior with a change in the staking protocol mechanism. Do you know how I can find out if and when the staking protocol changed ? Thanks for your help. Best regards
Indexers do list protocol parameters, here you have tzstats for example. Block rewards, and other protocol parameters have indeed changed with the evolution of the Emmy family of consensus algorithm, and general protocol evolution. You will find there the protocol change level, to match with your data.
PS: I suspect that the peak in the middle could come from not accounting from cycle length duplication with Florence? Would this alter your measures? -- I had not run the numbers.