dailysudoku.com

jayagee · Joined: 26 Feb 2006 Posts: 13 Location: sunny north dakota

Puzzle as posted:

XXX XXX 37X
XXX 8X5 94X
XXX 2X7 8X5

4X9 XXX X5X
X5X X8X X3X
X7X XXX 2X4

2X5 3X4 XXX
X43 9X8 XXX
X16 XXX XXX

I got this far:

5XX 1X6 372
X27 835 94X
XXX 2X7 8X5

4X9 72X X58
X52 48X X3X
X7X 65X 2X4

2X5 3X4 XXX
743 9X8 52X
X16 572 4X3

Now I’m stuck. I think the answer relates to the 1’s at r4, c6 &c7
and at r5, c6 & c7. Doesn’t seem like this is a true x-wing since there are 3 1’s in r5 and 3 1’s in box f or 6. Can someone shed some light???

Thanks!!

Marty R. · Posted: Wed Aug 23, 2006 4:23 am Post subject:

This is the grid with all candidates filled in based on the position you reached:

TKiel · Joined: 22 Feb 2006 Posts: 292 Location: Kalamazoo, MI

I got to this point, which differs from the posted one in that the (1,6) pairs in box 2 and column 4 aren't resolved yet. How did you do that?

David · Joined: 02 Jun 2006 Posts: 58 Location: Bedford, UK

I also arrived at the same point as Tracey and have the same question.

David · Joined: 02 Jun 2006 Posts: 58 Location: Bedford, UK

I have a question.

If you look at R7C28 and R9C18, it looks something similar to a unique rectangle. Can you therefore assume that cell at R7C8 has to be either a 1 or a 6 ?? The solution will certainly fall out from this assumption, but of course I may be using false logic.

keith · Posted: Wed Aug 23, 2006 3:48 pm Post subject:

This is the puzzle we talked about over the weekend

http://www.dailysudoku.co.uk/sudoku/forums/viewtopic.php?t=1325

Keith

Marty R. · Posted: Wed Aug 23, 2006 3:54 pm Post subject:

TKiel · Joined: 22 Feb 2006 Posts: 292 Location: Kalamazoo, MI

Marty,

That's what I thought you did, but I tried to solve it from the beginning and got stuck at the point indicated.

Keith,

Thanks for pointing that out, I didn't realize they were the same. I solved that one shortly after you posted it, but when I tried to recreate what I did I got stopped at this same point. I know the first time I used an XY-wing that excluded the 9 from r5c9 (I think it was (1, 6 & 9) with r6c4 as the stem cell. But that means that r6c8 needs to lose the 1 and I can't figure out how I did that). Has Sudoku finally made me lose my mind?

keith · Posted: Wed Aug 23, 2006 11:23 pm Post subject: A mind to lose?

Tracy,

I saw your note about an XY-wing, I could not find it but did not comment.

I spent hours staring at the 1's and 6's. I was hoping you would come through with a multicoloring solution that I could not see!

Keith

jayagee · Joined: 26 Feb 2006 Posts: 13 Location: sunny north dakota

To Tracy and David. There is an x-wing of 1's at r3,c3 & c8 and r6, c3 & c8. This eliminates 1 at r6,c4 and makes r1,c4 a 1 and of course a 6 at r1 c6.. Those were my last two numbers before being stumped.

jayagee · Joined: 26 Feb 2006 Posts: 13 Location: sunny north dakota

To Marty R. Thanks; I 'm not very good with the x-y thing.. I might have noticed it if the 16 in r5,c1 had been in c2, but then again maybe not. that logic is a bit beyond my experience level. Just an old geezer trying to keep up with the young generation.. Thanks again; I,ll file this one away for future reference.

Marty R. · Posted: Thu Aug 24, 2006 1:55 am Post subject:

TKiel · Joined: 22 Feb 2006 Posts: 292 Location: Kalamazoo, MI

Jayagee,

I don't see the X-wing, either. My column 8 has a third 1 in r7c8. How did you get rid of that?

Have made halting progress. Once the (1,6) Type 1 UR exclusions are made from r6c6, there is a multi-colouring chain on 6's that excludes 6 from r5c6. That gets me 2/3's of the way to that elusive (possibly) phantom XY-wing.

Steve R · Posted: Thu Aug 24, 2006 3:34 am Post subject:

One way forward from the position Tracy first posted

David · Joined: 02 Jun 2006 Posts: 58 Location: Bedford, UK

I am not sure if I qualify as an old geezer yet but my wife will not doubt beg to differ.

Anyway, could somebody please look at my previous post concerning the cells at R7C28 and R9C18 on Tracey's posting. At the moment these cells contain the candidates (89), (1689), (89) and (89). If they were all to contain just (89), would this constitute some form of 'deadly pattern' and hence force R7C8 to be (16) ??

TKiel · Joined: 22 Feb 2006 Posts: 292 Location: Kalamazoo, MI

David,

Look at this discussion here (specifically the David Bryant post) for a sort-of answer to your question. They can be removed but not necessarily because it's an almost UR. If an almost UR used the same rules as a UR, then look at cells r7c58 and r8c59. That almost UR says we can remove the (1,6) from r7c8, which would reduce the candidates in that cell to an unacceptably low level.

TKiel · Joined: 22 Feb 2006 Posts: 292 Location: Kalamazoo, MI

After reading Steve R's post (specifically about his XY-chain) and with inspiration born of desperation (and a tip of the hat to David Bryant, Keith and Marty R for the realization that I had to construct some sort of chain) I think I figured out a way to get to the XY-wing I used the first time I solved this puzzle.

I left off here:

jayagee · Joined: 26 Feb 2006 Posts: 13 Location: sunny north dakota

Tracy: I see what you mean.. I eliminated the 16 from r6,c6 because of the 16 pairs at r1,c4,c6 and r6,c4. Then I decided i had a triple in c6 consisting of 13,19,and 39 in r4,5,&6 respectively. Now I think that may have been false logic even tho it proved out. That yeilds 6 at r6,c4 and clears the 16 pairs in box b. Then my x-wing of 1,s works. I'm pretty sure I've lost my wits the more I stare at this #^*@ thing. Time to use the shredder i,m thinkin.

David · Joined: 02 Jun 2006 Posts: 58 Location: Bedford, UK

Tracey

Thanks. That's a good point about the (other) 'almost UR', clearly false assumptions on my behalf. I will now read the piece from David Bryant et al and see if I can follow their logic.

Jayagee

I don't think your triple works unfortunately. I think it's back to school for the both of us.

Victor · Joined: 29 Sep 2005 Posts: 207 Location: NI

Thanks for that clean, concise explanation, Steve - my 1st encounter with an XY-chain.
Nobody mentioned it, but there is a small technique available here. Once you've used a UR to kill the 16 in R6C6, in rows 3 & 6 there are just two 6s. So there can be no other 6s in column 8 - it kills the 6 in r7c8. (This doesn't otherwise help towards a solution.) I think of this little beast as a partial X-wing. Does it merit its own name?