dailysudoku.com

tlanglet · Posted: Fri Jul 24, 2009 4:19 pm Post subject: Menneske Super Hard (56)

I found a one step solution for this puzzle but otherwise it took me seven steps. I don't always solve these puzzles at this level.

storm_norm · Joined: 18 Oct 2007 Posts: 1741

this might be a little off the wall...
notice that you can create a pair {13} in row 3 with the 13 in r3c1 and the pseudocell {13} created by the type 3 UR69r39c79... all this is true IF the pair {79} is false in r37c1

in other words the {79} pair r37c1 and the type 3 UR69[(13)r3c179] cannot both be false.

tlanglet · Posted: Fri Jul 24, 2009 9:35 pm Post subject:

Norm, your analysis is extraordinary!

My one step solution also used the UR 69 in r39c79, but I just followed the implications that r3c7=3 or r3c9=13.
(3)r3c79 - (3)r3c3 = (3)r5c3;
(1)r3c9 - (1)r45c9 = (1)r4c8 - (1)r4c456 = (1)r6c4 - (1=6)r6c2 - (6=3)r5c3;

Thus, r5c3=3 if the UR constraints are satisfied.

Ted

ttt · Joined: 06 Dec 2008 Posts: 42 Location: vietnam

For AUR(69)r39c79, I did view on other way: at least one of [(9)r3c13, (6)r3c5] must be true
(9)r3c13=(6-7)r3c5=(7)r3c1-(7=9)r7c1 => r1c1<>9

ttt

storm_norm · Joined: 18 Oct 2007 Posts: 1741

ttt, ted,

very nice

Myth Jellies · Joined: 27 Jun 2006 Posts: 64

Not to take anything away from Norm's solution path, which was genius, but here is a little hint I have found that may help the more mortal among us.

When evaluating a UR, if you find yourself considering locked sets with the extra UR digits, take a peek instead at where the base digits have to land when they get moved out of the UR. Oftentimes you can replace an error-prone deduction (with the quantum "cell") with a much simpler equivalent. For example here we have a limited number of escapes for the six and/or nine isolated to r3

daj95376 · Joined: 23 Aug 2008 Posts: 3854

I arrived late to the discussion, but decided to see what I'd have done with the UR. Turns out, my approach is essentially identical to Ted's, but with a different elimination.