if the lower bound of a blocks timestamps is the median of the previous blocks, how can the network difficulty ever decrease?

Question

according to bitcoin developer documentation the lower bound for a blocks timestamp is the median of the previous 11 blocks.

and also according to the documentation the network difficulty is calculated based on the median blocktime of the past 2016 blocks to ensure that the average block is mined every 10 minutes

but if difficulty is calculated by the median blocktime of the past blocks, and the lowest possible timestamp a block can set itself to is the very median timestamp that resulted in the current difficulty, how can the difficulty of the network ever actually decrease?

Answer 1

The lower bound for block timestamps is simply a measure against miners setting block timestamps too far back into the past – in an idealized scenario of one block every 10 minutes, the median of past 11 blocks is 1 hour ago. Similarly, there is an upper bound on the timestamp, nodes will not accept a block with a timestamp more than 2 hours into the future (measured against the network-adjusted time).

Notice that these bounds are never actually violated if everyone's clocks on the network are synchronized and miners are always putting the correct timestamp into a block.

Answer 2

So i misunderstood the documentation

i assumed it was talking about the difference in time between each block, however see now that its probably just the actual epoch timestamp.

So the lower bound is NOT the median amount of time it took between each block to be mined, its the actual median time ago those 11 blocks were all mined "together".

So a miner can just set a blocks timestamp to be the actual current epoch timestamp. I thought it had to set it the median blocktime ahead in the future.

Answer 3

The difficulty of the network will decrease when the median block time over ~2 weeks rises increases over the ~10 minute target, usually as a result of a decrease in total mining hashpower.