Would it be possible to give a “future” account graph, based on previous income/expenses? I feel like it would be nice to have an idea of where the account might be in the future. It might also be useful to optionally exclude categories or transactions from this transaction, to simulate what would occur if some things stopped happening
2 Likes
This is a nice idea 
Do you mean exclude categories from the graph to see how much you could save?
Yes, something along those lines, excluding categories or tags
How far in the future would you like to see?
I imagine the future-seeing lengths would mirror the past ones (1W, 1M, 3M, 6M, 1Y). I’m not sure if it would be too useful to see further than that. I think I’d usually use 3M or 6M, myself.
1 Like
I suppose with one week you would also have to think about which week you would predict it against. For example, if it was the 3rd week of the month you would want to compare that to the third week in your last month.
What would you base the 3/6 months off of? Just the last 3/6 months or that exact time a year ago?
Yeah, you’re right in thinking that the shorter-term futures would need to take that into account. As for the 3/6 months of, I’d personally go for the last 3/6 months, as they’re a more recent reflection of what my transactions.
Setting a time period to project a future from works for both cases I think, just as long as you can set the end of the original period, and its length matches the length of the future we’re “predicting”
Yes I understand what you mean I was just taking into consideration things like holidays Christmas etc. If you were to predict it might skew due to more spending in certain periods across the year!
Ah, that’s actually a super good shout! A year would certainly be the most accurate overall, but if smaller, perhaps it could take into account (say we were looking a month ahead) both the last month, and that month in the given year?
1 Like
Yeah, that would make sense, that way you have both things to compare it against!
Yep! I think a good thing for this are charts like those with confidence/error bands or similar - they have a shaded area to indicate “we reckon this is probably the value, but it could actually be anywhere within this shaded part”, like this image here: https://static1.squarespace.com/static/5a07c63e18b27d4c1ef5e91d/t/5a19c8c6652dea2e1a3ef466/1511639255075/before_and_after_error_bands.png?format=2500w
1 Like
Yeah that works well, it’s good to see examples as well, sometimes it’s difficult to understand what people want just from written descriptions.
1 Like
Amen to that, sometimes I read what I’ve written, and it absolutely does not get across what I’m wanting to say.
It might get a little weird when there’s enough uncertainty though - it might be that after a few months, the shaded area is huge. But I think that depends on how much you’re allowing things to vary in predictions, and how different the two time periods you’re looking at are (the last X period of time, and the X period of time at the same time last year)
1 Like