Yes I get all that. But the principle remains the same in influencing the average as it does a rolling average. If the number on a given day/future days is above the cases 14 days ago then so does the rolling average increase a little. And as there were a couple of days of very low figures it is quite possible for that to occur if later days are higher again. AKA the apparent Wednesday spike. It is hardly a negative comment but a quite possible scenario.
I think we’re saying the same thing, but the number 15 days ago is the key, not 14 days.
For simplicity let’s say the rolling average is over 10 days ....
New cases over 10 days of 100, 10, 20, 10, 15, 15, 10, 20, 10, 20 gives a rolling average of 23 per day. The next day the cases jump to 60. The rolling average will come down to 19 per day (as the 100 is replaced by the 60) even though the last day saw a big jump. Now if the 60 is maintained the next day, the rolling average will of course increase.