View previous topic :: View next topic |
Author |
Message |
LloydB
Joined: 09 Jul 2007 Posts: 21 Location: Skagit Co., WA USA
|
Posted: Wed Jul 09, 2008 9:47 am Post subject: July 9 VH |
|
|
I can't believe I'm the first to comment.
An X-wing on 9 followed by a 689 XY-wing solved it for me. |
|
Back to top |
|
|
George Woods
Joined: 28 Mar 2006 Posts: 304 Location: Dorset UK
|
Posted: Wed Jul 09, 2008 10:35 am Post subject: same only different |
|
|
Yes - the same X Wing and a W wing in boxes 7,8,9 |
|
Back to top |
|
|
cgordon
Joined: 04 May 2007 Posts: 769 Location: ontario, canada
|
Posted: Wed Jul 09, 2008 11:42 am Post subject: |
|
|
A funny one. There were so many basic eliminations, I thought it would end up being easy. But I had to use the x wing on 9; a skyscraper on 6 and a BUG + 1. |
|
Back to top |
|
|
Deb
Joined: 30 Jun 2008 Posts: 8
|
Posted: Wed Jul 09, 2008 1:27 pm Post subject: Can You Explain? |
|
|
I'm new to Sudoku. I find the "hard" ones easy but the "very hard" ones impossible...until today's. I completed it, but I'm not sure how; it wasn't a lucky guess, I used logic, but I just happened to start the logic thread at the right place. So, I was wondering if you could help formalize what I did.
After the easy stuff, I saw that the "69" pair in R5C1 and R9C1 and the "89" pair in R5C9 and R9C9 would eliminate any 9's in their path, which left me with boxes 7, 8 and 9 looking like this:
31.7.45.2
75.23.641
.24.1573.
R8C3 and R8C6 were a "89" pair so I just used the logic of "if R8C3 is an 8 how does it flow? Then, if R8C3 is a 9 how does it flow?" I realized that it would be impossible for R8C3 to be a 9, so the puzzle was solved.
Can you tell me what that configuration was in boxes 7, 8 and 9? And how I would recognize it again? |
|
Back to top |
|
|
cgordon
Joined: 04 May 2007 Posts: 769 Location: ontario, canada
|
Posted: Wed Jul 09, 2008 2:20 pm Post subject: |
|
|
Hi Deb:
I believe your first statement relates to the x-wing mentioned in previous posts.
The grid below shows the unsolved candidates for 9. Look at C1 and C9 where there are only two 9’s in each column.
Either R5C1 or R5C9 has to be a 9 – so you can eliminate all other 9’s in R5 (marked 9X). Same thing in R9.
I think your next statement suggests a guess – though I have been known to be wrong. You should refer to the various techniques for solving the Very Hards, e.g. x, xy and xyz wings, skyscrapers, UR’s and many more. I don’t have up-to-date websites but I’m sure others could help.
And I would suggest copying the grid below (incl. the code symbols) for use when posting numbers and grids.
Good Luck.
Craig
Code: |
+-------+-------+-------+
| . . . | . . . | . . . |
| . . . | . . . | 9 9 . |
| . . . | . 9 9 | . . . |
+-------+-------+-------+
| . . 9 | . . . | 9 . . |
| 9 .9X | . . . | .9X 9 |
| . . . | . . . | . . . |
+-------+-------+-------+
| . . 9 | . 9 . | . 9 . |
| . . 9 | . . 9 | . . . |
| 9 . . | . . . |9X . 9 |
+-------+-------+-------+
|
|
|
Back to top |
|
|
Marty R.
Joined: 12 Feb 2006 Posts: 5770 Location: Rochester, NY, USA
|
Posted: Wed Jul 09, 2008 3:48 pm Post subject: |
|
|
One W-Wing on 89 with pincer coloring was all I needed. |
|
Back to top |
|
|
Johan
Joined: 25 Jun 2007 Posts: 206 Location: Bornem Belgium
|
Posted: Wed Jul 09, 2008 4:16 pm Post subject: |
|
|
I used the 26 UR* in R45C36 to solve this one, the digits marked + (= possible solutions to avoid a 26 DP) in the 26 UR pins <8> in R8C3.
R45C3=9 => R8C3=8
R4C6=8 => R8C6=9 => R8C3=8
R5C6=1 => R5C89=89 => R5C1=6 => R9C1=9 => R8C3=8
Code: | +-------------------------+-------------------------+-------------------------+
| 5 9 1 | 4 2 3 | 8 6 7 |
| 2 6 3 | 5 8 7 | 19 19 4 |
| 4 8 7 | 1 69 69 | 3 2 5 |
+-------------------------+-------------------------+-------------------------+
| 1 4 * 26+9 | 68 7 * 26+8 | 29 5 3 |
| 69 7 * 26+9 | 3 5 * 26+1 | 4 189 89 |
| 8 3 5 | 9 4 12 | 127 17 6 |
+-------------------------+-------------------------+-------------------------+
| 3 1 689 | 678 69 4 | 5 789 2 |
| 7 5 =89 | 2 3 89 | 6 4 1 |
| 69 2 4 | 678 1 5 | 79 3 89 |
+-------------------------+-------------------------+-------------------------+ |
|
|
Back to top |
|
|
Deb
Joined: 30 Jun 2008 Posts: 8
|
Posted: Wed Jul 09, 2008 5:33 pm Post subject: Thanks |
|
|
I didn't realize one had to learn so many techniques. I thought it was all logic. Thanks for the pointers. |
|
Back to top |
|
|
wapati
Joined: 10 Jun 2008 Posts: 472 Location: Brampton, Ontario, Canada.
|
Posted: Wed Jul 09, 2008 8:15 pm Post subject: Re: Thanks |
|
|
Deb wrote: | I didn't realize one had to learn so many techniques. I thought it was all logic. Thanks for the pointers. |
If you look back you may see that you don't have to learn even most of the techniques. There are so many! Did you notice that people pretty much all did it different ways? If you use the W-wing it is the only hard step needed, but you could use other tricks first.
Code: |
As cgordon pointed out, your first technique was an x-wing.
Your second was a guess. r8c6 could only be an 8 or a 9.
You tried one, it didn't work, you tried the other it did.
You could do the same with any bi-value cell in any puzzle.
For logical methods there are at least three that use those 89s.
There is a W-wing as marked (aka semi-remote pair).
One of the 89s must be 9. Whichever one is 9 forces one of the 69s to be 6.
Whichever is 6 r7c3 cannot be 6.
.---------------.---------------.---------------.
| 5 9 1 | 4 2 3 | 8 6 7 |
| 2 6 3 | 5 8 7 | 19 19 4 |
| 4 8 7 | 1 69 69 | 3 2 5 |
:---------------+---------------+---------------:
| 1 4 269 | 68 7 268 | 29 5 3 |
| 69 7 26 | 3 5 126 | 4 18 89 |
| 8 3 5 | 9 4 12 | 12 7 6 |
:---------------+---------------+---------------:
| 3 1 89-6| 7 @69 4 | 5 89 2 |
| 7 5 #89 | 2 3 #89 | 6 4 1 |
|@69 2 4 | 68 1 5 | 7 3 89 |
'---------------'---------------'---------------'
There are 2 slightly different xy-wings working with the 89s.
You can see that if (the hinge in) r9c1 is 6 then r9c4 is 8.
If r9c1 is 9 then r8c3 is 8. Thus either r8c3 or r9c4 is 8.
No matter the value for r9c1, r8c6 cannot be 8.
:---------------+---------------+---------------:
| 3 1 689 | 7 69 4 | 5 89 2 |
| 7 5 *89 | 2 3 9-8 | 6 4 1 |
|*69 2 4 |*68 1 5 | 7 3 89 |
'---------------'---------------'---------------'
Here we can use the other 89.
The hinge is r9c4, this time.
If r9c4 is 6 then r9c1 is 9. If r9c4 is 8 then r8c6 is 9.
Whatever the value in r9c4 r8c3 cannot be 9.
:---------------+---------------+---------------:
| 3 1 689 | 7 69 4 | 5 89 2 |
| 7 5 8-9 | 2 3 *89 | 6 4 1 |
|*69 2 4 |*68 1 5 | 7 3 89 |
'---------------'---------------'---------------' |
I learn new methods because it is fun! I hope you can enjoy it too. |
|
Back to top |
|
|
Deb
Joined: 30 Jun 2008 Posts: 8
|
Posted: Wed Jul 09, 2008 8:42 pm Post subject: |
|
|
wapati, thanks so much for the elegant demo. I see now that it really was a guess at the "89".
And now I know what a hinge is.
But I'm afraid the w-wing eluded me. (as did x-wing, xy-wing -- I see I have a lot to learn) You say it is marked, but I'm not seeing it. I'll google it and I'm sure I'll come up with an explanation.
I guess I like things simple so if a w-wing is all I need, that's what I focus on. I like the sudoku as a mental break. Thanks. |
|
Back to top |
|
|
wapati
Joined: 10 Jun 2008 Posts: 472 Location: Brampton, Ontario, Canada.
|
Posted: Wed Jul 09, 2008 9:25 pm Post subject: |
|
|
Deb wrote: |
And now I know what a hinge is.
|
Sorry Deb,
hinge in this case means bending point, or perhaps focus. There are several methods that call a certain cell a "hinge" cell and xy-wing is one.
3D medusa another, Y-wing chains, wxyz-wings and xyz-wings are others. Hinge is a general term for a critical cell.
There is also a technique first called "hinge" but now commonly called ER (Empty Rectangle). It was first called "hinge" because it was L shaped. It was later expanded to + shape as well, thus the new name. The meeting of the arms within the box is the hinge point.
The way I look at xy-wings is that is a 3 cell locked set of bi-value cells with a bend in it. The hinge cell sees both of the other cells but the others do not see each other.
W-wing, I like the name semi-remote pair as it is descriptive.
The basic idea is that you have two bi-value cells with the same two candidates but they can't "see" each other. We will assume that they are both 12 and we can put one in r1c1 and the other in r9c9.
As it stands they have no effect on each other. We look for a strong link on 1s or on 2s where one end can see r1c1 and the other r9c9.
Suppose it is a strong link on 1s. One of these must be 1 which means one of r1c1 or r9c9 (or both) cannot be one and must be 2. In the general case any cell that sees both the semi-remote pairs cannot be 2. In this case both r1c9 and r9c1 cannot be 2.
Code: | +-----------------------------------+
| 12 . . | . . . |*14 . -2 |
| . . . | . . . | . . . |
| . . . | . . . | . . . |
|-----------+-----------+-----------|
| . . . | . . . | . . . |
| . . . | . . . | . . . |
| . . . | . . . | . . . |
|-----------+-----------+-----------|
| . . . | . . . |*13 . . |
| . . . | . . . | . . . |
| -2 . . | . . . | . . 12 |
+-----------------------------------+ |
Hope this helps! |
|
Back to top |
|
|
Marty R.
Joined: 12 Feb 2006 Posts: 5770 Location: Rochester, NY, USA
|
|
Back to top |
|
|
cgordon
Joined: 04 May 2007 Posts: 769 Location: ontario, canada
|
Posted: Wed Jul 09, 2008 9:42 pm Post subject: |
|
|
At the risk of alienating my peers (again) I would suggest the x wings, xy wings and xyz wings and holding the w's (I don't find them user friendly).
try http://www.sudocue.net/guide.php#XWing |
|
Back to top |
|
|
Deb
Joined: 30 Jun 2008 Posts: 8
|
Posted: Wed Jul 09, 2008 10:25 pm Post subject: Holy Cow |
|
|
Medusa, Skyscraper, W-Wing! Who knew?!? Thanks all. I've bookmarked your links and I'll read (Ah, STUDY) them. Thanks for helping out a newbie. |
|
Back to top |
|
|
nataraj
Joined: 03 Aug 2007 Posts: 1048 Location: near Vienna, Austria
|
Posted: Wed Jul 09, 2008 10:33 pm Post subject: |
|
|
Craig, I'm far from being alienated. But if it is necessary to have a discussion about good/bad/legitimate solving techniques, why not have it now?
wapati has given an excellent concise description of the w-wing.
The w-wing, speaking from my own experience, was an acquired taste. I ignored it when it was first proposed. Maybe I was lazy, maybe I did not grasp the simple underlying principle, maybe - I don't know - maybe I felt that the world did not need yet another "bla,bla"-wing. That has changed.
Knowing and using both the x-wing and the w-wing I'll make a strong point in favor of the w-wing: it is much faster to check a puzzle for w-wings than for x-wings.
x-wings could be hidden in ALL cells of the puzzle, sometimes using a cell with five candidates. Very hard to see and one has to go over all 81 cells repeatedly. "user friendly" ? not by my definition ...
w-wings, on the other hand, use bi-values only (easy to see), and only those which are identical (at least two of a kind must be present in the puzzle) and don't "see" each other.
That makes -what? 3, maybe 5 pairs of cells to check. And we know exactly what to look for: either a strong link in the first of the two candidates - or a strong link in the other candidate. Not exactly rocket science, is it?
Are w-wings necessary to solve puzzles in the "daily" section of dailysudoku.com? definitely not.
Could they be useful in solving sudokus published somewhere else? for sure.
Should we therefore agree on a taboo - banning w-wings or other techniques from our arsenal when solving the "daily" - I don't think so.
Don't get me wrong: I am not talking about limiting YOUR arsenal of methods or to force you into using any particular method. Feel free to solve these puzzles any way you like. I have my preferred methods, too, my "modus operandi". I am talking about NOT limiting the methods that we - as a community - might want to try and talk about in this forum... Yes, in the "daily" section. |
|
Back to top |
|
|
keith
Joined: 19 Sep 2005 Posts: 3355 Location: near Detroit, Michigan, USA
|
Posted: Thu Jul 10, 2008 2:39 am Post subject: |
|
|
Nataraj proposes that W-wings are easy to recognize. I would add that M-wings and other variants of remote and complementary pairs are also relatively easy to recognize.
I will also add that for these techniques, the arms war belongs to us, the pencil & paper solvers. As has been pointed out over and over again, a W-wing can make toast of a multiple-step "Very Hard" puzzle. And, those that make and rate the puzzles, do not yet have W- and M- wings in their arsenals.
We may have this advantage for a while, for I think that these wings may be difficult to program.
Keith |
|
Back to top |
|
|
keith
Joined: 19 Sep 2005 Posts: 3355 Location: near Detroit, Michigan, USA
|
Posted: Thu Jul 10, 2008 3:06 am Post subject: |
|
|
If you go back through the history of this list, you will see that it has always involved a discussion of other puzzles, and of techniques much more complicated than those required to solve the Daily Sudoku.
Keith |
|
Back to top |
|
|
Marty R.
Joined: 12 Feb 2006 Posts: 5770 Location: Rochester, NY, USA
|
Posted: Thu Jul 10, 2008 5:20 am Post subject: |
|
|
Craig, you can't alienate me. But I love W-Wings. They're one of the new techniques I've learned over the last year, along with Medusa, pincer coloring and ERs, that have significantly expanded my puzzle-solving capabilities. (Although I have a long way to go). |
|
Back to top |
|
|
cgordon
Joined: 04 May 2007 Posts: 769 Location: ontario, canada
|
Posted: Thu Jul 10, 2008 7:52 pm Post subject: |
|
|
Let me reiterate the situation.
A very nice person (Deb) wanted to learn the Very Hard techniques. And another very nice person - (Me) - directed her to the obvious basic techniques – x and xy wings. I am not denying that w-wings are a valid solution – I just believe that w-wings should not be recommended to novices.
Stay by me Deb – I will protect you from these elitists. |
|
Back to top |
|
|
keith
Joined: 19 Sep 2005 Posts: 3355 Location: near Detroit, Michigan, USA
|
Posted: Thu Jul 10, 2008 11:49 pm Post subject: |
|
|
Quote: | I will protect you from these elitists. | The "elitists" who want you to know about methods they find useful?
A storm in a teacup.
Best wishes,
Keith |
|
Back to top |
|
|
|