hoshi911: Ok, thanks for your explanation.
With this link, you started to count double or triple occurring
problems only as one problem, which is fair, I guess.
However, you do this recounting only on the user side. On the platform
side, double problems still count as 2, triple problems as 3, and so on (see Collections - Problem Count).
And that is completely unfair! It basically means that in order to reach 75% of the 10.240 problems of your "Problem Count", one has to
solve around 80% of the truly occurring number of problems in the data base, which might be around 9.700 or even less.
Why don't you make a fair problem count on the platform side as well?
For instance use as formula the following:
Adjusted Problem Count = Problem Count (using multiples) - number of double problems - 2* number of triple problems - 3* (and so on).

hoshi911: My comment is on this problem, but it currently applies to many others.
Firstly, I had completed the block Tsumego Dictionary Volume III 1001-1200.
A few days ago, some of these problems pop up as unsolved again. When I
resolve them, it says "Correct" and shows me the SGF file. However,
when I come back again, the platform claims that it is still unsolved.
Furthermore, I observed in recent days, that the hearts system has become completely unstable. Sometimes no hearts get erased although a problem is solved incorrectly, sometimes the number of hearts jumps by many without no reason,
sometimes after using the rejuvenation button, the number of hearts remains zero,
and so on.
It's not a pleasure to use Tsumego Hero these days, I have to admitt.

d4rkm4tter: Should be fixed. Problems change status, because we link duplicate problems.

Ok, thanks for your explanation. With this link, you started to count double or triple occurring problems only as one problem, which is fair, I guess. However, you do this recounting only on the user side. On the platform side, double problems still count as 2, triple problems as 3, and so on (see Collections - Problem Count). And that is completely unfair! It basically means that in order to reach 75% of the 10.240 problems of your "Problem Count", one has to solve around 80% of the truly occurring number of problems in the data base, which might be around 9.700 or even less. Why don't you make a fair problem count on the platform side as well? For instance use as formula the following: Adjusted Problem Count = Problem Count (using multiples) - number of double problems - 2* number of triple problems - 3* (and so on).