Feedback on a puzzle design

[Br1312te]

Hellooo,

I recently posted a puzzle on LMG: https://logic-masters.de/Raetselportal/R...?id=000PXE

For my standard this puzzle received a very low rating, and I found this quite surprising. I can usually tell by how much fun I had designing the puzzle, how fun it would be to solve. And this one basically designed itself; so I expected it to be a smooth solve too. It's also a sequel of sorts to a puzzle that indeed was very popular.

I'm not writing this to complain. I hope someone could give some feedback. Clearly I'm not seeing something and it's shook my confidence a little. 

Love, Brigitte

[Dandelo]

I think 75% means "good". So 81% is better than "good".

[dumediat]

Hello, Brigitte. I see you've posted several well-received puzzles on the portal, so you seem to be experienced in building good logic into your puzzles. Having just solved this puzzle, I thought I would share some of my thoughts.

I'm not sure how many of your puzzles use global constraints, but putting multiple into a single puzzle usually narrows the audience of people who might enjoy it. There's nothing wrong with doing so, of course, but you might be more likely to see a lower rating (or maybe fewer solves) for such puzzles.

With multiple global constraints, you might also run into an issue where the intended solve path involves global ideas or patterns. This logic may be intended as smooth, but there's always a chance a solver won't find your idea. For harder puzzles, this might mean the solver simply can't complete it. However, for easier puzzles, there is usually at least one way around the intended logic that can feel case-checky. People who don't find the intended path may not like the puzzle as much. That appears to be the case for at least some people based on the comments on this puzzle.

Finally, there are individuals on the puzzle portal who are routinely harsh with their ratings. If they solve your puzzle early, your initial rating will be lower than if they don't. Perhaps that's the case here, since the current rating of 88% is very respectable.

My best advice is to not get attached to beauty ratings. In my opinion, it's OK to use the rating as a barometer for when there might be something "wrong" with the puzzle, so in that regard it's good that you're seeking out feedback. But don't let ratings shake your confidence, as much as possible. Comments (kind and constructive) are a much better source of feedback.

[Br1312te]

This is very helpful, thanks dumediat!