Randomness breaks starting from block 400,000. It correlates with reaching the block size limit. Observing it has something to do with 10^29 (~seconds since Jan 1 1970) another user speculated some mining pool uses timestamps to find nonces and does something wrong (I don't fully understand the conversation).
There is a speculation that some entity simply found an ASIC optimization not known to others.
Other people did same plots for Monero (discovered even more oddities that coincided with covert ASICs) and Ethereum (nothing, all random noise).
Finally, this study found that a lot of nonces are close to zero which suggests many miners start brute forcing the nonce from zero and move linearly (50% of winning nonces come from 0.002% of the search space).
Usefulness for Decred I can imagine:
it's simply interesting how the distribution looks for Decred
xaur
commented
5 years ago
Seemingly useless but fun research idea.
One researcher found oddities in Bitcoin nonce distribution (which was expected to be random): https://twitter.com/100trillionUSD/status/1081217034485149697
Randomness breaks starting from block 400,000. It correlates with reaching the block size limit. Observing it has something to do with 10^29 (~seconds since Jan 1 1970) another user speculated some mining pool uses timestamps to find nonces and does something wrong (I don't fully understand the conversation).
There is a speculation that some entity simply found an ASIC optimization not known to others.
Other people did same plots for Monero (discovered even more oddities that coincided with covert ASICs) and Ethereum (nothing, all random noise).
Finally, this study found that a lot of nonces are close to zero which suggests many miners start brute forcing the nonce from zero and move linearly (50% of winning nonces come from 0.002% of the search space).
Usefulness for Decred I can imagine: