If we think of how long ago a specific event occurred, for instance, the last Presidential inauguration, we can measure how far back it was down to the second.
If we think much further back, I can still find events measurable to the day (Assassination of Julius Caesar, March 15, 44 BC) close to the introduction of the Julian calendar. But then other major events close in time like the Crucifixion of Jesus are murkier.
Not too long before that, I am having more difficulty in coming up with an event measurable to the day. I can find things measured to the month, like the coronation and death of Xerxes.
Is there a specific set of calendars we can chain backwards to correlate the earliest possible event to the day? What event might that be?
Total eclipses are excellent in this regard, and accounts of them in history allow us to calculate delta T, which is the total amount of 'drift' in the Earth's rotation.
The trouble with going too far back to tie these to an event that occurred on a specific day, is there are often multiple such eclipses that could be involved. Additionally, some eclipses are recorded in a relatively timid fashion compared to the sorts of declarations we get from a totality, leading some to believe they may refer to annular eclipses instead.
The earliest date I know of that can be assumed to be accurate to near certainty would be the Battle of the Eclipse, during the Eclipse of Thales. No other eclipse can qualify for this event. There are some claims that it may refer to a lunar eclipse instead, but I am not sure if the people making this claim have ever experienced totality to compare.