@ThePaulBunyanTrophy
Thanks for the interesting insight!
If it's that way it means that the app it's useful for finding someone to date atm but that doesn't mean it would find someone you would end up marrying (If they are into that). After all, dating someone changes you.
So, the algorithm actually answers in this way to data the people input:
The you right now has this possibilities of having a fruitful and compatible relationship with this X person in the next, let's say, 2 to 3 months.
So it's logical that running a simulation again after some time has passed might change the porcentages considerably, depending on how much one has changed.
@dojinshi
I like that is a different take in which the capabilities of the simulation to reach further in the future are more reliable, yet become unreliable precisely because of that reach.
And it case it isn't about reach, then it raises other questions about how the probability works. To give an example:
Maybe this score is not that high because even if the simulation tells us that this is the person that might bring a positive personality change to your life, you won't be able to reciprocate those changes.