The visual theme of pi sudoku was very nice, but as David mentioned it was a puzzle where it was much easier to solve by guessing the long sequences at the bottom than using logic to know what went where so I skipped over it in my first pass. I was two bifurcations deep (for the 3 and the 4 digit sequences on the bottom) in my notation, but the first things I tried worked probably because the number of usable sequences is so low. So I think the challenge of breaking into the puzzle may have been the biggest factor, but I could also see confusion about the instructions keeping others from trying it too.

Edited by motris 2010-09-27 8:54 PM

Agreed - I can't for the life of me work out how to solve these with anything approaching elegance and fluidity, or even without scribbling down lots of sums in what is essentially trial and error. I have a mess of notes and calculation by the side of my printed page, and I'm impressed by anyone who claims to solve these relatively cleanly. I'd like to know how it's done! Ditto calculo-doku or whatever the current Krtek cup is.

N.B. as a maths researcher you should take it as given that my arithmetic is lazy and prone to all sorts of errors ;)

Pi sudoku solved.
1234+567+89 solved too. Not very cleanly , but perhaps I didn't thought enough about logic of these puzzles (like: when the last digit of a sum is 5, sum of the 3 units is 15 (5 or 25 isn't possible),...)

I didn't like calcdoku, but I think I'm in progress .

Quite often the 1000's place column of 1234 gets solved immediately.

That is a very interesting observation indeed.
Thanks for pointing out.