I’ve noticed a weird interaction. It makes sense given the model, but feels like it creates the wrong behaviour:
If I enter a data point for yesterday, then ratchet, this creates a different gap between my current data and the road than if I do these operations the other way around: In the former, the ratchet absorbs the difference, in the latter the ratchet does not.
Intuitively what “should” happen is that if I enter a data point for before the ratchet, it should be absorbed by the ratchet as much as possible.
This isn’t a major problem for me, but its lack means that I am semi forced to never backdate data on a particular goal, because I have it designed to auto ratchet to the two week mark and in normal operation it should always be hovering around that mark (it’s a rolling budget for certain classes of spending which are very high variance on a week by week basis and I don’t want to allow the budget to build up too much on “quiet” weeks).