Stars, Comments, and Flags

One top goal of this site is having really high quality problems. This means a few different things, in priority order:

1. The problem must be valid to begin with. In other words, if you are supposed to kill a group, it must actually be killable.
2. The problem must be fully correct: all the solution paths are covered, with no incorrect solutions.
3. All reasonable failure paths are covered. This is a little bit of an art, knowing what is appropriate for a problem.
4. A general feeling of how nice a problem is. This is subjective. This is represented by the stars.

User Feedback relies on the user community to give feedback on problems. It’s hard for a problem creator to get every problem exactly right every time. We ask users to leave flags, comments, and star ratings. In turn, we ask problem creators to diligently monitor the comments and flags and respond to any items that need addressing, and to clear flags that have been fixed or aren’t correct.

Invalid Problems

Some problems are simply invalid. For example, you are asked to live, but there is no sequence that creates life. Or it’s a non-sensical problem that isn’t even close to a real problem. These should be flagged with the category “Invalid”, and a brief reason given. This flag is reserved for problems that can’t be easily fixed without changing the problem significantly.

Some examples of things that do NOT belong in this category:
- Problems with a rating you think is not what it should be — ratings are calculated automatically
- Missing variations or solution paths

Problems needing fixing

The most common issue needing fixing is missing a valid solution path, an alternate valid sequence. The ideal user feedback works like this: add the variation to the tree with a comment, something like “this also works”. Select the comment type “Correct”.

Other common issues for needing fixing:
- A computer response doesn’t correctly refute, even though it could
- A position is labeled something like “seki” when it’s not seki
- A sequence was marked as correct but doesn’t actually work
- Typos in the text
- Some sequences that really should be covered aren’t in the problem

Adding a “Needs fixing” flag is usually appropriate for these situations. However, if you added a new variation on the tree with a comment type of “Correct line” or “Wrong line” this is enough information without the flag. It doesn’t hurt to add the flag too though.

Duplicate problems

No matter how nice a problem is, it should only exist one time on the site. There are mechanisms to help catch this at creation time, such as Pattern Search. However, it’s hard to catch it all, and so we have a mechanism of flagging problems as duplicates. Simply enter the ID of the duplicated problem on the Duplicate flag. Whichever problem was entered first into the site will get to stay.

The exact definition of a duplicate problem is a bit hard to pin down, but essentially if all key stones, shapes, and solutions are the same, it’s a duplicate. We ignore of course transposition and changing player colors, as well as the extra boundary stones.

Problem Stars

On every problem, users can rate the problem from 1 to 5 stars. The site calculates an overall rating based on all users’ scores. This is where subjective elements come in.
Some common reasons to rate a problem highly:
- A nice elegant tesuji
- The problem may have a fun story accompanying it

Some reasons to rate a problem low:
- Solving the problem is a pain, or a chore, without corresponding benefit. For example, playing out a long ladder without an interesting end.
- Having to read some complicated but not interesting instructions.
- Lots of extra stones that take your brain extra time to parse the problem, but don’t make it any more interesting.

Important: do NOT rate a problem lowly for things that should be covered by flags and comments. For example, missing some variations. These things can be fixed. The stars rating should apply only to the intangible subjective quality of the problem.