So I’m very open to feedback on this, but my general idea behind construction and word choice is that I permit 1 obscure word because you don’t ultimately need to know it to solve the puzzle. In theory, you only need to know half of the clues. Issues arise when two obscure words intersect, but I do my best to avoid those.

It’s a bit of a trade-off between being able to make grids, and keeping all the words common and general knowledge. Being able to fit a specific grouping of letters into a grid is tricky, which is why I allow words to be backwards. While it’s a nice quirk to differentiate the puzzle, it’s actually very necessary for me to make grids in the first place!

However, I’m always trying to expand my dictionaries, and fine tune them so that grid solutions are fair.

In the future, I want to be able to allow people to opt in to dictionaries - so you can choose puzzles that only have general knowledge words, but, for example, if you know everything about Lord of the Rings, I can add in puzzles where I needed “Eowyn” or “Gimli” to end up creating a grid.

I’m rambling now I fear, but back to your point, I currently feel okay adding in 1 obscure word because it doesn’t ultimately block the puzzle solve, it lets me make more grids, and I think there is an element of fun learning new words - however, if the obscure word does end up ruining the experience of the puzzle overall, then perhaps a recalibration is needed on my end, at least in the interim until people can opt in to certain dictionaries! Would love your thoughts on this!