View previous topic :: View next topic |
Author |
Message |
Philmac
Joined: 04 Aug 2005 Posts: 16 Location: Cheshire, England
|
Posted: Wed Nov 16, 2005 11:25 pm Post subject: nov14 hard |
|
|
Am I being completely stupid or is this tougher than it might appear. I have spent longer on this than I have on many 'very hard' but have come to a complete standstill. If someone tells me it's not that difficult I'll publish where I have got to and ask for a hint .. but am I alone?!
I've not used x-wing or swordfish before so I might be missing something, but I've even tried those without (apparently) any success.
Feeling tired and dumb ...
Phil |
|
Back to top |
|
|
geoff h
Joined: 07 Aug 2005 Posts: 58 Location: Sydney
|
Posted: Wed Nov 16, 2005 11:34 pm Post subject: |
|
|
Hi Phil,
Actually, it's not that hard - you certainly don't need X wing or swordfish to solve it. Why don't you publish where you've got to and we'll give you some hints.
Cheers. |
|
Back to top |
|
|
David Bryant
Joined: 29 Jul 2005 Posts: 559 Location: Denver, Colorado
|
Posted: Wed Nov 16, 2005 11:38 pm Post subject: It wasn't too hard |
|
|
I did get hung up on this one for a little while, Phil. Maybe you're stuck in the same place I was. My big break came when I realized that the "8" in the bottom center 3x3 box had to be in row 8 -- either in r8c4 or r8c5. That allowed me to complete r8c8, and things went more smoothly after that.
If that's not the right hint, maybe you should post your current position. dcb |
|
Back to top |
|
|
Philmac
Joined: 04 Aug 2005 Posts: 16 Location: Cheshire, England
|
Posted: Wed Nov 16, 2005 11:42 pm Post subject: nov14 hard |
|
|
Ok folks here is where I am at:
203 510 000
109 002 305
560 003 012
300 005 246
456 238 001
792 641 538
830 100 029
627 000 103
900 320 000
David. I had located the 8 as you say but that hasn't given me a breakthrough!
Regards
Phil |
|
Back to top |
|
|
David Bryant
Joined: 29 Jul 2005 Posts: 559 Location: Denver, Colorado
|
Posted: Wed Nov 16, 2005 11:54 pm Post subject: That's where I had to scratch my head |
|
|
Hi, Phil!
You need to look at r8c8 more closely. I see {1, 2, 3, 4} in column 8 already. I see {6, 7} in row 8, and I see there's a "9" in the bottom right 3x3 box, at r7c9. So the only apparent possibilities at r8c8 are {5, 8} ... and you can rule out the "8" because it has to be in the bottom center 3x3 box (in row 8). So r8c8 = 5. dcb |
|
Back to top |
|
|
geoff h
Joined: 07 Aug 2005 Posts: 58 Location: Sydney
|
Posted: Thu Nov 17, 2005 12:00 am Post subject: |
|
|
Hi Phil,
Im not sure if you took David's advice on board as your posting shows Cell r8c8 still empty. Using David's hints, you can place Nr 5 in r8c8.
Then you can place Nr 5 in r7c5 since it is the only cell in Column 5 where you can place Nr 5. You can then easily place Nr 5 in r9c3 since it is the only cell in Column 3 where you can place Nr 5.
You should be right from there.
Cheers. |
|
Back to top |
|
|
Philmac
Joined: 04 Aug 2005 Posts: 16 Location: Cheshire, England
|
Posted: Thu Nov 17, 2005 12:03 am Post subject: as i said .. stupid! |
|
|
david
Yes, of course and I'm sure that it will all fall out now.
Unusually for me I panicked and coded up all possibilities in all remaining cells (I usually tend to prefer just looking and seeing .. along the lines that you have explained). There were so many small digits floating around that I missed the simple deduction - serves me right! I was looking everywhere but r8 (I had the 58 in r8c8). I am left wondering whether there was another way forward or is this one of those crunch points that has only one esscape route?
Many thanks for returning my sanity to me anyway.
Phil (it's about -3 C outside, at the moment in Cheshire) |
|
Back to top |
|
|
David Bryant
Joined: 29 Jul 2005 Posts: 559 Location: Denver, Colorado
|
Posted: Thu Nov 17, 2005 12:20 am Post subject: It's getting colder here, too. |
|
|
Hi again, Phil.
I've done that, too -- writing in too many possibilities too soon, I mean, so that the important information is hard to see. That's what practice is for, I guess. I don't know if there was another way forward from that point, or not. That was the one I found.
Last night it hit about 15 degrees F (-9.5 C) here in Denver. Winter is coming! dcb |
|
Back to top |
|
|
Philmac
Joined: 04 Aug 2005 Posts: 16 Location: Cheshire, England
|
Posted: Thu Nov 17, 2005 12:31 am Post subject: across the world |
|
|
... but not in Sydney!
Thanks again everyone.
phil |
|
Back to top |
|
|
geoff h
Joined: 07 Aug 2005 Posts: 58 Location: Sydney
|
Posted: Thu Nov 17, 2005 3:21 am Post subject: |
|
|
Sunny 25 degrees Celsius here in Sydney and we've all got hangovers this morning after celebrating Australia's win over Uruguay last night to put us in the Soccer World Cup Next year!!!
First time since 1974 - let's go Australia!!!!! |
|
Back to top |
|
|
thesaint
Joined: 17 Nov 2005 Posts: 4 Location: Cambridgeshire
|
Posted: Thu Nov 17, 2005 12:03 pm Post subject: |
|
|
I found this one quite straight forward in that I did not have to go beyond the simple solving methods. I did not have to look for any sets (naked or concealed). It took me 64 minutes; this was beause when I do a puzzle labelled "Hard"; I only spend a few minutes finding a few easy numbers; I then go straight into writing, in every empty cell, all the possible candidates. I know this is boring and time consuming, but, (if done carefully without making errors) it makes looking for sets and other patterns much easier. |
|
Back to top |
|
|
Guest
|
Posted: Thu Nov 17, 2005 6:05 pm Post subject: |
|
|
Still at it on this one... I don't see how Philmac got the 2s and 3s in boxes 1,4, 5, and 6. I see the 8/1 twins in row 4, col 2&3, and mapped out the 2's and 3's.
000 510 000
109 002 305
560 003 012
000 005 046
456 238 001
790 641 008
830 100 029
627 000 103
900 320 000
Hint? Thx. |
|
Back to top |
|
|
thesaint
Joined: 17 Nov 2005 Posts: 4 Location: Cambridgeshire
|
Posted: Thu Nov 17, 2005 7:25 pm Post subject: The "Hard" one |
|
|
Anonymous wrote: | Still at it on this one... I don't see how Philmac got the 2s and 3s in boxes 1,4, 5, and 6. I see the 8/1 twins in row 4, col 2&3, and mapped out the 2's and 3's.
000 510 000
109 002 305
560 003 012
000 005 046
456 238 001
790 641 008
830 100 029
627 000 103
900 320 000
Hint? Thx. | Hello Guest, From your stated position I note that you have overlooked the fact that in the 3 x 1 set of boxes 3/6/9; you have a 3 at r2 c7 and another 3 at r8 c9. Therefore the 3 required in box 6 either can go in r5 c8 or in r6 c8. However; there is already a 3 in row 5 at c5 therefore the missing 3 must go in r6 c8. Now look at the 5s in boxes 4/5/6 and you will get another "solve". These two items should get you going again Regards from Simon Templer |
|
Back to top |
|
|
Guest
|
Posted: Thu Nov 17, 2005 8:04 pm Post subject: |
|
|
Simon- Thanks for the quick reply. Unfortunately, in haste I copied my position wrong - I don't have the '3' filled at R5C5.
Corrected position-
000 510 000
109 002 305
560 003 012
000 005 046
456 208 001
790 641 008
830 100 029
627 000 103
900 320 000 |
|
Back to top |
|
|
David Bryant
Joined: 29 Jul 2005 Posts: 559 Location: Denver, Colorado
|
Posted: Thu Nov 17, 2005 8:54 pm Post subject: Try a "5" at r8c8 |
|
|
Hello, Guest!
You can put a "5" at r8c8. Column 8 already contains {1, 2, 4}. Row 8 contains {6, 7}. And the bottom right 3x3 box contains {3, 9}. So the only possible values at r8c8 are {5, 8}.
But you can rule out an "8" because of the "8"s at r7c1 and r5c6 -- the only way an "8" can fit in the bottom center 3x3 box is if it goes in the eighth row, at r8c4 or r8c5. So r8c8 = 5.
After that you can quickly place a "5" at r6c7, a "3" at r6c8, and a "2" at r6c3. Then the "3" at r5c5 pops out, because the only way left to fit a "3" in the middle left 3x3 box is in row 4.
Hopefully that's enough to get you rolling again! dcb |
|
Back to top |
|
|
Guest
|
Posted: Thu Nov 17, 2005 9:11 pm Post subject: |
|
|
Thanks David-
Thanks for spelling it out for me- I should have read your first comment more carefully. |
|
Back to top |
|
|
David Bryant
Joined: 29 Jul 2005 Posts: 559 Location: Denver, Colorado
|
Posted: Thu Nov 17, 2005 9:56 pm Post subject: That's OK |
|
|
Oh, that's OK. I wouldn't be spending so much time over here if I didn't get a charge out of it! dcb |
|
Back to top |
|
|
thesaint
Joined: 17 Nov 2005 Posts: 4 Location: Cambridgeshire
|
Posted: Thu Nov 17, 2005 10:24 pm Post subject: The "Hard" one |
|
|
Hello again Guest, From your corrected position; write in all the candidates for every emply cell. You will find that r8 c8 has only one possible candidate - a 5 - then look at the other the 5s now contained in boxes 3/6/9 - you can now place another 5. You should be okay now. Regards from Simon Templer |
|
Back to top |
|
|
alanr555
Joined: 01 Aug 2005 Posts: 198 Location: Bideford Devon EX39
|
Posted: Fri Nov 18, 2005 1:55 am Post subject: Re: nov14 hard |
|
|
Code: |
> Is this tougher than it might appear. I have spent longer on this than I
> have on many 'very hard' but have come to a complete standstill.
It is quite a difficult one. Although I usually try to keep to using
just Mandatory Pairs, I did xx to full candidate profiles on this one.
The solution appears to depend upon identifying the subsets on rows
1 and 4 plus columns 3 and 6.
Row 1 has three pairs - 23/47/68.
Row 4 has 18 as a pair leaving 2379 to be resolved.
Col 3 has 23 as a pair leaving 1458 to be resolved.
Col 6 has two pairs 49 and 67
The challenge is then to break into the pairs.
I do not have a full record of my solution trail but I did get there in
the end. The intermediate point was
000 510 000
109 002 305
560 003 012
000 005 046
456 208 001
790 641 008
830 100 029
627 000 103
910 320 000
I was perhaps a little premature on moving to candidate profiles
at that point and I might redo the puzzle sometime to see if it can
be fully resolved using Mandatory Pairs.
Alan Rayner BS23 2QT
|
|
|
Back to top |
|
|
alanr555
Joined: 01 Aug 2005 Posts: 198 Location: Bideford Devon EX39
|
Posted: Fri Nov 18, 2005 2:22 am Post subject: Re: It's getting colder here, too. |
|
|
Code: |
> I have written in too many possibilities too soon - so that the
> important information is hard to see.
This is wise counsel.
When I first started, doing a candidate profile (called "tiny writing" in
a "Guardian" article) was the only way that I knew. I had not looked
up "sudoku" in any search engine at that time. I coded up a solver in
dBase for what I understood to be the methods in the Guardian article
- and then found that it did not solve all the puzzles put to it!! After I
became acquainted with the more complex solver rules and methods
I gave up on the dBase project and gradually have concentrated on
"manual" methods - attempting to avoid the "chore" of coding up the
candidate profiles!
So, yes, patience is a virtue. Doing the candidate profiles is much
easier if one has resolved a lot of cells first by simple methods. It
does complicate the page but makes setting the candidate profiles
easier if one first prepares the "Missing" profile for each row/column
and records that at the edge of the grid by the relevant row/column.
With this 14th Nov puzzle, I was impatient and set the candidate profiles
at probably too early a stage. I paid the price by having a very "messy"
page - two lines have seven entries in the "missing" profile and at
least four others had six. This leads to overly long profiles for the
individual cells (eg 46789 appeared for r1c7) and thus makes scanning
for patterns somewhat harder than it should be. The row actually
resolved to three pairs and so having 68 in place of 46789 did make
progress much easier once that state had been reached!
Perhaps the greatest difficulty is achieving a balance between time
and detail. If "Look and See" enables a cell to be resolved, it is
generally quicker than deriving all the profiles and, also, simplifies
any future derivations. On the other hand, poring over the grid for
twenty minutes and seeing nothing is not very encouraging or, indeed,
inspiring to progress - the time could perhaps be better spent on
deriving the profiles!
Does anyone have any hints as to WHEN in a puzzle to take the decision
to develop the profiles rather than to continue looking for the "simple"
patterns?
I find that applying a logical approach (eg checking everything for 1,
then everything for 2 etc) can take an inordinate time and then, during
that process, I find that attention to an "evident" pattern could have
shortened the timespan considerably. We all know that searching for
a swordfish or turbot will take time and so my question is really for
any guidance on time management.
Alan Rayner BS23 2QT
|
|
|
Back to top |
|
|
|