dailysudoku.com

keith · Posted: Sun May 27, 2007 9:39 pm Post subject:

Glassman,

At the risk of irritating you, the Type 3 UR in R46C68 says R5C6 is <2>.

Otherwise, an XY-wing 18-48-14 takes out <1> from R1C3 an R3C5.

Then, there is another XY-wing, and an X-wing ...

Neat puzzle!

But, I do not see how to complete the puzzle without a chain ...

Keith

Mogulmeister · Joined: 03 May 2007 Posts: 1151

Glassman,

I don't see it straight away either - I do see an XY wing though.

Mogulmeister · Joined: 03 May 2007 Posts: 1151

Eureka !

Take the XY Wing that Keith identified - crucially by removing the 1 from r3c5 this reduces that square to a <46> bivalue.

Now we can eliminate that 6 that is troubling you:

ALS - Set A is r3c5 <46>.
Set B is r4c4 <67> and r6c5 <47>

Locked common is 4. The 6 in r1c4 can see both 6's in A & B and is thus eliminated.

NB these cells will also make up an XY chain - pincers at r3c5 and r4c4.

Mogulmeister · Joined: 03 May 2007 Posts: 1151

All this having been said there is a way to go to complete the puzzle.........

Glassman · Joined: 21 Oct 2005 Posts: 50 Location: England

Keith — I got side-tracked when I realised that your xy-wing eliminated the 1s at r2c1 and r2c3. That 57 was too tempting a prize not to be of use.

Mogulmeister — I prefer yours as a simple xy-wing — I struggle with ALSs.

Oh, and Keith, thank you for your diplomacy. However, has Andrew Stuart ever said that all his puzzles have a unique solution? ... or that he never makes a mistake? And, if so, would you believe him?

Glassman Cool

Mogulmeister · Joined: 03 May 2007 Posts: 1151

I put in the ALS because it got rid of your 6 but if x-y chain works as well so be it.

Personally, I would definitely always trust Andrew Stuart - he worked with and has now succeeded Michael Mepham(RIP) whose Sudoku puzzles always have one solution.

I think he does excellent work.