I'm STUCK!! (img included)

Everything about Sudoku that doesn't fit in one of the other sections

I'm STUCK!! (img included)

Postby Spazactaz » Fri Sep 30, 2005 6:56 am

I've been trying this for hours, and I cannot seem to go any further!! All my moves are correct so far. Any help would be much appreciated:

Image
Spazactaz
 
Posts: 6
Joined: 29 September 2005

Postby PaulIQ164 » Fri Sep 30, 2005 7:39 am

There's a quite involved tactic you can use that will fix r9c7 (9th row down; 7th column across). It's all about the 4 cells at r3c7, r3c8, r9c7, r9c8. If they were all 4s and 5s, then, once you've finished the puzzle, you'd be able to swap round the 4s and 5s in those four cells and you'd get another valid solution (Think about it). So one of those cells has to be something else. The only one that can be is r9c7, which can be an 8. So it has to be that.

I'm sure there's probably something more abvious, though.

Edit: Look how early it is! And I'm all up and posting!
PaulIQ164
 
Posts: 533
Joined: 16 July 2005

Postby Spazactaz » Fri Sep 30, 2005 8:37 am

Wow - I don't know if I'd have ever came up with that!! :o Thank you so much!!:)

Hopefully some day I'll be able to complete a "tough" one. :P
Spazactaz
 
Posts: 6
Joined: 29 September 2005

Postby Shazbot » Fri Sep 30, 2005 8:42 am

That's very clever - I'd have struggled long and hard with it! Is there a name for this technique (I'm still grasping hidden and naked singles and pairs)?

Is this the only option for a next step, or another that might be easier to pick up but still assist towards a solution?

Spazactaz, did the puzzle have a difficulty rating?
Shazbot
 
Posts: 220
Joined: 24 September 2005

Postby Spazactaz » Fri Sep 30, 2005 9:48 am

Yeah, it was rated as "tough" (the most difficult..)
Spazactaz
 
Posts: 6
Joined: 29 September 2005

Postby simes » Fri Sep 30, 2005 10:57 am

There's also an elimination to be made from colouring 7s

Code: Select all
Colouring 7: double exclusion found, eliminating 7 from r6c5, (r1c5 => r1c2 => r4c2 => r6c1)


S
simes
 
Posts: 324
Joined: 11 March 2005
Location: UK

Postby saraalba » Fri Sep 30, 2005 11:24 am

como cyhe vai caralliño¿?somos as super nenas (só que solo somos 2)me cagho na cona!!!!!!!!!!!!!! VIBA O PP!!!!!arriba aznar
saraalba
 
Posts: 1
Joined: 30 September 2005

Postby PaulIQ164 » Fri Sep 30, 2005 2:29 pm

I'm not sure there's a name for the technique I used. Several people have independently come up with it and posted it here, but I don't think anyone's got a catchy name for it.
PaulIQ164
 
Posts: 533
Joined: 16 July 2005

Postby nj3h » Fri Sep 30, 2005 3:28 pm

I save different messages and treads from this and other forums as reference. I was stuck in what to name the file.

So I called it 45 45 45 458.doc.

Therefore, should it be the "45 45 45 458 Rule" or the XY XY XY XYZ Rule"?

Just kidding, of course.
nj3h
 
Posts: 47
Joined: 07 July 2005

Postby Karyobin » Fri Sep 30, 2005 3:54 pm

It tends to go under the epithet of the Uniqueness Rule. At least, that's how most people who use it seem to think about it.
Karyobin
 
Posts: 396
Joined: 18 June 2005

Postby ab » Fri Sep 30, 2005 8:28 pm

I'm not happy about using that technique! Suppose the puzzle composer was careless and posted a puzzle that doesn't have a unique solution.

You're supposed to be able to solve sudoku using logic, but this technique makes an extra assumption, that there is a unique solution. Logic alone does not lead to that conclusion.
ab
 
Posts: 451
Joined: 06 September 2005

Postby Karyobin » Fri Sep 30, 2005 8:45 pm

This has been discussed before but not by me, so sadly I can't point you to the relevant threads, Paul probably can. If I remember correctly, the argument for the technique is the stonger, on the reasoning that sudoku have only one solution by definition, and can be solved logically. As a result of this fact, any puzzle which has more than one solution must involve a placement at some point which Sadman's software excitedly describes as "A Guess!" So: were you to stumble upon an arrangment such as
Code: Select all
{1,2}{1,2}
{1,2}{1,2}

you would have arrived at a point where a guess was the only way forward, so the puzzle does not have one solution. In the reverse, therefore:
Code: Select all
{1,2}{1,2}
{1,2}{1,2,3}

the fourth cell must be a '3', as if it were not you would have arrived back at the earlier position, and logic could take you no further. I know it seems a bit like saying 2+2=4 because 4=2+2, but it isn't, as the Uniqueness Rule is essential and inherent to the puzzle you are solving.

I think.


Paul?
Karyobin
 
Posts: 396
Joined: 18 June 2005

Postby PaulIQ164 » Fri Sep 30, 2005 9:07 pm

Hello! There's a thread about this tactic here (as well as the topic that you've evidently already found, having posted in it). I think Karyobin got the arguments for it pretty much right. There's no getting around the fact that the tactic does require you to 'assume' the puzzle has a unique solution, but that's surely something you should assume anyway. If there are multiple solutions it would be a guess, but you'd have to guess anyway!

A question that, to my knowledge, remains open is whether you can construct a multiple-solution puzzle that, if you do use this tactic, fails to work. If you can't, that's surely the last objection done away with, as then if you used the tactic on a multiple solution puzzle you could just consider it the guess that you'd have to make anyway, safe in the knowledge that you'll be alright in the end.

What distinguishes this puzzle from others I've seen is that, putting it into the Susser, I see that without this tactic you in fact can't actually solve it without dead clever advanced tactics (forcing chains in this case).[/url]
PaulIQ164
 
Posts: 533
Joined: 16 July 2005

Postby Spazactaz » Fri Sep 30, 2005 10:49 pm

Today's puzzle seemed to require that technique as well, which led me to completing my first ever "tough" sudoku!:D yay. Thanks you all!:)
Spazactaz
 
Posts: 6
Joined: 29 September 2005


Return to General

cron