dailysudoku.com

[pendrith]

I'm stuck on the July 6 puzzle. Can anyone explain the logic required to continue??

942---567
173562948
856-4-312
-29---673
78----25-
63--2-48-
3--28-79-
--8-7312-
2-7---836

[David Bryant]

To proceed you need to analyze where a "1" can fall in row 4.

-- There cannot be a "1" at r4c1 because of the "1" at r2c1.

-- Therefore the "1" in row 4 must fall in the middle center 3x3 box. This means that you can eliminate the possibility of a "1" in the rest of that box -- at r5c4-6, and at r6c4&6, in other words.

-- Now you can find a naked pair {7, 9} in column 4 -- at r3c4 & r6c4. So you can set r8c4 = 6, and the rest should be relatively straightforward. dcb

[pendrith]

Hi David

Thanks for responding.

I'm still confused however.

No doubt finding where the "1" falls in row 4 would help solve the puzzle.

I understand how there cannot be a "1" at r4c1

However I don't see how you can eliminate a "1" at r4c4 or r4c6

What am I missing??

[Marty R.]

[ravel]

942|xxx|567
173|562|948
856|x4x|312
-----------
x29|xxx|673
78x|xxx|25x
63x|x2x|48x
-----------
3xx|28x|79x
xx8|x73|12x
2x7|xxx|836

To repeat in other words, what David said:
In row 4 the 1 must be in column 4,5 or 6. So box 5 must have the 1 in this row and it cannot be elsewhere in that box (this is called box/box interaction, Locked Candidates 2 or Claiming). Therefore only 79 remain for r6c4, the same we have in r3c4.
Same thing for the 4: In row 9 the 4 must be in column 4 or 6. Therefore it cannot be in r7c6 or r8c4.
Looking at the remaining candidates in r8c4 now, it cannot be 4 and it cannot be 9 from the 79-pair in r36c4. Then only 6 is left.

[pendrith]

Got it!

Thanks guys

[TKiel]

[ravel]

I agree, that its not the best name (as for some other techniques). I suppose, it comes from the fact, that with a "box/line interaction" (locked candidates 1) you can eliminate candidates in the line (row/column), whereas with "box/box interaction" you can eliminate them in the box.

[TKiel]

Ravel,

Thanks. Your explanation was good enough to make the name seem logical.

[keith]

Here is another explanation:

http://www.sadmansoftware.com/sudoku/technique4.htm

I think most of us find these in the course of looking for line - box interactions.

Keith

[ravel]

Ah, you found the link ! And yes, i think, it is unusual to spot box/box in the desribed way. This made it hard to understand the name.

[TKiel]

The explanation on the Sadman site seems totally backward to me, although if one did manage to look at it that way, the name would make more sense.

If we can have umpteen UR's labeled Type 1, Type 2, etc... it seems we should be able to keep track of box/line Type 1 (locked candidates 1) and box/line Type 2 (locked candidates 2). Or maybe just box/line and line/box would work. Nah, too many people would think that was a mis-print. Maybe it's time to organize a Sudoku board of governours to standardize/logisize terminology.

[keith]

Tracy,

You are correct. I don't think a box-box interaction is something different to learn. I also did not understand the Sadman explanation when I originally found it. I thought I was stupidly missing something.

Also, I'd like to see more standard terminology. I tried to impose one on myself:

"Box" not "block" or "square"..
"Candidate" not "possibility".
(Box - Line) "Interaction" not "intersection".
"Line" not "row or column".
"Naked single" not "forced square".
"Cell" not "square".
"Hidden single" not "pinned square".

It's tougher than I thought, to consistently use one set of terms.

By the way, my wife's cousins were here from Kalamazoo for (a family reunion and) the Woodward Dream Cruise this weekend. In just a few years, the Cruise has become an incredible event and spectacle.

Best wishes,
Keith