Sudoko theory

It looks like a clever way to visualize the problem, and to make the ‘guessing’ easier. Instead of putting for example tiny eights and nines, you simply decide you’re testing nines and run true/false checks to see if it’s possible. Easier than going back and erasing.

In twickster’s puzzle there’s a couple of squares where it can be either a nine or one other number. These have corresponding squares in the same group (row, column or 3x3 square).

Row 1:column 6 is either 8 or 9. A conjugate pair is found at row 3:column 4, also 8 or 9 and mutually exclusive. Looking only at the nines, we find that this has a conjugate pair in R4:C4 - either a 3 (doesn’t matter, only looking at the nines) or a 9. This leads to R4:C3 and R6:C5.

And you keep going, alternating between blue and green until you find an impossible situation with either colour (two of the same colour in the same group). The other colour then is where you want to put your 9s.
Hmm, not as clear as I was hoping. Sorry, it’s early.

I’ll be honest with you, guys – I solve Sudoku for fun, and I don’t find running long hypotheticals fun. I was wondering, mostly, if I was missing something major here, or whether I really had reached a point where nothing more could be determined without testing a string into a dead-end – and it sounds like I had.

I’m okay with having ditched that particular puzzle and moved on to one I could solve.

I’ll just throw in a not so subtle plug for sudokuconquest.com, the newest addition to the puzzleconquests. Also try hitoriconquest and kakuroconquest. Excellent time slayers.