Take a look at the timeline:
Code: Select all
Timeline: 0..........30..........60__________90__________120 Event: Update 1 Update 2
How come if I've just fetched fresh data and updated the Sheet? Because the data does not tell me when in the dotted (..) period those 100 Pirates were killed. They could have been killed:
- between 0 and 5 minutes
- between 55 and 60 minutes
- between 0 and 59 minutes
The same principles apply then we observe event Update 2. Just before Update 2, that is before I update the Sheet with fresh data, I'll say "The data is two hours out of date." and mean "The data is AT MOST two hours out of date.".
Once the data has been updated again, I'll say "The data is one hour out of date." and mean "The data is AT MOST one hour out of date." returning us back to Update 1 event on the timeline.
Obviously, the data is rarely in the WORST case state and it is far too pessimistic to assume that we are always acting on data at least an hour late, but it is important to know the upper bound: two hours!