On: NIST ion clock sets new record for most accurate clock

[hoseja]

Well for example you can take several of the clocks and compare them.

[CoastalCoder → hoseja]

Would that require the assumption that any inaccuracies affecting multiple clocks are uncorrelated with each other?