Sudoku Mahabharat & Puzzle Ramayan Finals
LMI Launches Puzzle Expo!
Posting a reply to: [CTC] Comments on 2016-01-13's Tapa
Anonymous: No
Disable HTML


You are replying to:

CTC Comments

CTC Comments posted @ 2016-01-16 12:03 AM

Comments on 2016-01-13's Tapa
Prasanna16391 commented at 2016-01-13 00:17:04
Saw the logical way on the resolve but I don't really regret not waiting for it because its really narrow and obscure. A nice puzzle once you figure it out though.
rob commented at 2016-01-13 00:49:37
Hmm, care to share a hint? I only saw the wrong way out of the bottom left fail when the puzzle became non-unique in the top left... (And then it does fail due to connectivity, but that's going quite far.) It's nice when the required case distinction is near the start, though, because then I can just "Clear Puzzle".
Prasanna16391 commented at 2016-01-13 00:51:10
Its a lot of connectivity stuff, thinking that something can't come out from two separate places at the same time. The bottom 3-1-1 can also be proven to be straight or curved early on which helps.
Prasanna16391 commented at 2016-01-13 00:53:40
Specifically, the top left part of the bottom 3-1-1 will have to be part of the "3" of it and the top right of it can either come out upwards through the middle or through the right, meaning the bottom has to come out from the right. Does require some forward thinking but not much if you spot the right patterns.
kiwijam commented at 2016-01-13 06:24:03
I've done this a few times now, but still don't see much logic at the start. If you guess wrong then there's no immediate contradiction, and you can continue happily until connectivity fails later. But I suppose guessing is probably still faster than trying to find a logical track.
kiwijam commented at 2016-01-13 06:29:24
PS I found the top-right corner a faster place to start guessing.
john_reid commented at 2016-01-13 06:29:57
Yeah - it was just TOO hard.
Para commented at 2016-01-13 09:32:19
That was confusing. My brain really kept excluding correct options. Surprised I'm still so far at the top.
rajeshk commented at 2016-01-13 12:41:03
For any guess I made, contradiction came very late and I have to redo this puzzle many times. Very nice puzzle to get your brain ticking.
anderson commented at 2016-01-13 21:15:35
I kept thinking I had a contradiction when guessing my way through it, and it turned out that my very first guess wasn't actually a contradiction but I thought it was. Oops! Is there a solution that doesn't involve thinking more than two steps forward at any point?
rob commented at 2016-01-14 00:56:34
I'm somewhat closer to something reasonable now (trying out chess notation): A3,B4,A5 shaded, DE6 blank. Can't go through below E2 (else D3 can't escape). Wall between CD4 (else D2 can't escape). There's now two options for C3: B3B4C4 shaded, in which case D3 has to escape along DEF3, or D234 shadedin which case D1 must also be shaded (say for A3 to escape). So we also have two options for E2, and in both cases, F1 is a shaded 1.
rob commented at 2016-01-14 01:03:08
... Now we see that F3 has to be part of G4's 3, and its group can either escape up along the center, or to the bottom right (but not both). So F3 has to escape along the right of I3. Now H4 must be blank, so G4 connects with G5's 3, hence G6 is blank. So our bottom right group needs to escape via I78 and EFG10. H8's 1 must now be H9, so we must shade H567, and finally we're ready to resolve the bottom left.