View previous topic :: View next topic |
Author |
Message |
RogerC
Joined: 08 Oct 2005 Posts: 14 Location: High Wycombe, Bucks, England
|
Posted: Thu Nov 24, 2005 7:05 am Post subject: November 23 - Very hard |
|
|
Just as I thought I had it cracked, along comes a puzzle that has me stumped. The hint says a 4 at r9c2, but I cannot see why.
Code: |
7 - 5 - 2 - - - 3
6 - 3 5 4 - - - -
4 - - 7 8 3 - - 6
9 - - - 3 - 8 2 5
8 3 - - 5 2 - - -
5 6 2 8 9 - - 3 -
3 - - 2 1 8 - - 9
2 - - 3 6 4 1 - -
1 - - 9 7 5 3 - 2
|
I'm not ashamed of getting hints if I'm really stuck and often learn new tecniques in the process. But I will never just accept the hint without the reason - that's cheating!!
I'd be grateful to anyone who can explain this one.
Roger. |
|
Back to top |
|
|
someone_somewhere
Joined: 07 Aug 2005 Posts: 275 Location: Munich
|
Posted: Thu Nov 24, 2005 8:12 am Post subject: |
|
|
Hi,
Don't trust anybody & anything. Why to trust the "hint"?
Starting from:
Code: | 7 - 5 - 2 - - - 3
6 - 3 5 4 - - - -
4 - - 7 8 3 - - 6
9 - - - 3 - 8 2 5
8 3 - - 5 2 - - -
5 6 2 8 9 - - 3 -
3 - - 2 1 8 - - 9
2 - - 3 6 4 1 - -
1 - - 9 7 5 3 - 2 |
it took me a while to get to the "hints" suggestion:
Code: | 6 not in r5c4, it is in r4c4 or r4c6 (Row on 3x3 Block interaction)
4 not in r5c7, it is in r6c7 or r6c9 (Row on 3x3 Block interaction)
4 not in r5c8, it is in r6c7 or r6c9 (Row on 3x3 Block interaction)
4 not in r5c9, it is in r6c7 or r6c9 (Row on 3x3 Block interaction)
4 in r6c9 - Unique Vertical
7 in r6c7 - Sole Candidate
1 in r5c9 1 in r6c6 - Sole Candidate
9 in r2c6 4 in r5c4 - Sole Candidate
6 in r1c6 2 in r2c7 6 in r4c4 7 in r5c3 - Sole Candidate
1 in r1c4 7 in r4c6 - Sole Candidate
2 in r3c2 - Unique Horizontal
8 not in r8c2, it is in r8c3 or r9c3 (Column on 3x3 Block interaction)
8 not in r9c2, it is in r8c3 or r9c3 (Column on 3x3 Block interaction)
4 in r9c2 - Sole Candidate |
but it looks that the "hint-machine" is very clever ...
see u,
P.S. just to check, for who cares, here the final position:
Code: | 7 9 5 1 2 6 4 8 3
6 8 3 5 4 9 2 1 7
4 2 1 7 8 3 9 5 6
9 1 4 6 3 7 8 2 5
8 3 7 4 5 2 6 9 1
5 6 2 8 9 1 7 3 4
3 7 6 2 1 8 5 4 9
2 5 9 3 6 4 1 7 8
1 4 8 9 7 5 3 6 2 |
|
|
Back to top |
|
|
David Bryant
Joined: 29 Jul 2005 Posts: 559 Location: Denver, Colorado
|
Posted: Thu Nov 24, 2005 12:16 pm Post subject: Why so complicated? |
|
|
Someone_Somewhere wrote: | t took me a while to get to the "hints" suggestion: ... |
It's really a bit simpler than that.
Concentrate on the top left 3x3 box. "8" cannot appear in r3c2 or r3c3 because there's already an "8" at r3c5. Therefore the "8" in column 2 must appear in r1c2 or else in r2c2.
The only values missing in row 9 are {4, 6, 8}. Knowing where the "8" must lie in column 2, we see that "4" is the sole candidate for r9c2. dcb |
|
Back to top |
|
|
someone_somewhere
Joined: 07 Aug 2005 Posts: 275 Location: Munich
|
Posted: Thu Nov 24, 2005 12:37 pm Post subject: |
|
|
Hi David,
Thank you for the solution. It made me think ...
One should find out what are ALL the numbers that he can set in a given Sudoku position, that are independent of each other. Meaning all the "NEXT" numbers that could be found out by some technique.
Humans always tend to implement the first best solution they are finding, sometimes thinking that it is also the only one (because they are happy to found one ;-)
I remembered to have developed a test, asking to "find why something was not working". Almost every candidate stopped, when he found the first mistake. There where 3 mistakes hidden. Only a small % found 2 and even a smaller % all of them.
Now I was the rabbit ... of the same experiment.
see u, |
|
Back to top |
|
|
RogerC
Joined: 08 Oct 2005 Posts: 14 Location: High Wycombe, Bucks, England
|
Posted: Thu Nov 24, 2005 1:18 pm Post subject: It's really a bit simpler than that |
|
|
Thanks David.
I knew it would be simple - I always look for a number occuring two or three times in a row or column of a box (whatever that is called) and eliminate that number from the rest of the row/column. I guess I was just too tired last night and the 'fresh' mind this morning was no good either.
Roger C. |
|
Back to top |
|
|
David Bryant
Joined: 29 Jul 2005 Posts: 559 Location: Denver, Colorado
|
Posted: Thu Nov 24, 2005 2:41 pm Post subject: Re: It's really a bit simpler than that |
|
|
You're welcome, Roger. This puzzle illustrates a setup that occurs quite frequently, and that is always easy to recognize.
Code: | x . x
x . x
x . . . . y . . . |
The value "y" MUST appear at r1c2 or at r2c2. I think of this as a sort of mirror, or a tunnel -- certain missing values appearing in a row (or in a column, if we rotate everything) sort of get focused, or bent, and "shoot out" in the orthogonal direction.
Glassman called this kind of formation -- or a very similar formation -- a "double pan." You can read that discussion here. |
|
Back to top |
|
|
Tigerphant Guest
|
Posted: Fri Nov 25, 2005 3:54 am Post subject: hints |
|
|
You were talking about hints...where do u get them? |
|
Back to top |
|
|
PoppaPoppa
Joined: 06 Nov 2005 Posts: 21 Location: Arkansas USA
|
Posted: Fri Nov 25, 2005 4:58 am Post subject: Hints |
|
|
I'm very slow this evening - worked on a difficult Sudoku puzzle in the local newspaper after a big Thanksgiving meal!
Hints are obtained after filling in the draw program and running it the first time - there's an option button next to the "show solution" button. I wish the author would modify the program to show a hint option each time it is used.
Cheers from Colorado
Chris (PoppaPoppa) |
|
Back to top |
|
|
zaks
Joined: 25 Nov 2005 Posts: 13
|
Posted: Fri Nov 25, 2005 8:55 am Post subject: Chess-like notation |
|
|
Dear sudoku fans,
While solving the sudoku i use chess-like notation:
columns l->r : a,b,c,d,e,f,g,h,i
rows d->u: 1,2,3,4,5,6,7,8,9
so d4 stands for cell r4c4, and 4a1 stands for move "put 4 in a1 (aka r1c1) cell "
Here is my solution (without comments and not in partucularly logical order):
Starting position:
7 - 5 - 2 - - - 3
6 - 3 5 4 - - - -
4 - - 7 8 3 - - 6
9 - - - 3 - 8 2 5
8 3 - - 5 2 - - -
5 6 2 8 9 - - 3 -
3 - - 2 1 8 - - 9
2 - - 3 6 4 1 - -
1 - - 9 7 5 3 - 2
1. 4b1! ("!" means good move as in chess)
2. 4i4 ! other moves are more or less simple:
3. 1f4 4. 7g4 5. 1d9 6. 6f9 7. 9f8 8. 7f6 9. 6d6
10. 4d5 11. 4c6 12. 1b6 13. 7c5 14. 1c7 15. 9c2 16. 8c1
17. 6c3 18. 6h1 19. 8g5 20. 9h5 21. 1i5 22. 5h7 23. 7b3
24. 5g3 25. 4h3 26. 4g9 27. 9g7 28. 2g8 29. 9b9 30. 2b7
31. 8b8 32. 1h8 33. 7i8 34. 8h9 35. 8h2 36. 7g2.
Final position coincides of course with given one
by someone-someone(?)
The notation is good for easy and middle problems, while for hard cases the "reasoning" includes sometimes very long sequences of "if - then"s and it is not clear how to write down the solution.
Enjoy, zaks |
|
Back to top |
|
|
zaks
Joined: 25 Nov 2005 Posts: 13
|
Posted: Fri Nov 25, 2005 12:25 pm Post subject: Nov 25: scoresheet of solution |
|
|
Solution of Nov 25 sudoku
1. 2h3 2. 7d5 3. 6b1 4. 2f5 5. 2a6 6, 2c7 7. 2i9 8. 8c5
9. 8i4 10. 4e5 11. 6g5 12. 6h2 13. 7i7 ! 14. 7g3 15. 7a1 16. 7b8
Here i asked hint and get 8d8 which is really difficult to find wthout "pencilling", so i proceed;
17. 8d8 18. 4f7 !(is it OK? it was again hard to me)
19. 3a3 ! 20. 9c3 ! 21. 8e3 22.4c2 23.1b3 24. 4d3
25. 4b4 26. 1a4 27. 3c8 28. 4a8 29. 5a9
30. 6d9 (i'm not sure that it's evident because since 16th move i use extensive pencillling)
31. 9e9 32. 8b9 33. 9b7 34. 3h9 35. 4h6 36. 1h1
37. 8h7 38. 3i6 39. 9i2 40. 4i1 41. 3g2 42. 3e1 43. 9f1 44. 5e4
45. 3f4 46. 1e7 47. 5f8 48. 1g8 49. 5g7 50. 1f2 51. 5e2 52. 6e6
53. 1d6 end!
And only now i see that today's sudoku labelled "hard", and at least for me it was such!
Is this notation useful/instructive?
Enjoy, zaks |
|
Back to top |
|
|
zaks
Joined: 25 Nov 2005 Posts: 13
|
Posted: Sat Nov 26, 2005 10:05 am Post subject: Re: Chess-like notation |
|
|
Here's Nov 23 2005 (v hard) sdk's soln from the beginning:
01 3b5 02 3f7 03 3d2 04 3h4 05 6a8 06 2a2 07 2f5 08 2d3
09 1e3 10 8e7 11 4e8 12 5e5 ! 13 6e2 14 9d1 15 5f1 16 4b1
17 8a5 ! 18 5a4 19 9a6 20 5i6 ! 21 8d4 ! 22 7g4 ! 23 1f4 24 4i4
25 1i5 26 4d5 27 6d6 28 7f6 29 9f8 30 6f9 31 1d9 32 1b6
33 4c6 34 7c5 35 6c3 36 8c1 37 9c2 38 1c7 39 2g8 40 8b8
41 7i8 42 1h8 43 6h1 44 9h5 45 5h7 46 9g7 47 4g9 48 8h9
49 9b9 50 2b7 51 6g5 52 5g3 53 7b3 54 5b2 55 7h2 56 4h3
57 8i8 x!
it's really v. hard - at least for me.
enjoy, zaks |
|
Back to top |
|
|
Guest
|
Posted: Mon Nov 28, 2005 2:58 am Post subject: Re: Why so complicated? |
|
|
David Bryant wrote: | ... The only values missing in row 9 are {4, 6, 8}. Knowing where the "8" must lie in column 2, we see that "4" is the sole candidate for r9c2. dcb |
I found this puzzle both interesting and elegant, and my congratulations to the compiler.
My thought processes were quite different to David's, but with the same outcome. I looked at what options were available for cell r9c3. If it was a 4, this would require both 6 and 8 to fit into r9c8, clearly impossible. So r9c3 had to be either 6 or 8. It was then a short step to realising that r9c3 and r9c8 were a {6 8} pair, forcing the 4 into r9c2.
Glassman |
|
Back to top |
|
|
Glassman
Joined: 21 Oct 2005 Posts: 50 Location: England
|
Posted: Mon Nov 28, 2005 7:42 am Post subject: |
|
|
... and that really was me! I seemed to have got myself logged out.
The point I was making was that there are often several ways of looking at the same problem. Here David concentrated on the 8's, whereas I concentrated on the 4's. Both of us looked at all possible permutations in our own individual way and ended up with the same result.
Glassman |
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|