dailysudoku.com :: View topic - Vanhegan extreme 4/7/12

dailysudoku.com
Discussion of Daily Sudoku puzzles

FAQ

Memberlist

Usergroups

Profile

Vanhegan extreme 4/7/12

dailysudoku.com Forum Index -> Other puzzles

View previous topic :: View next topic

Author

Message

arkietech

Joined: 31 Jul 2008
Posts: 1834
Location: Northwest Arkansas USA

Posted: Mon May 07, 2012 6:34 am Post subject: Vanhegan extreme 4/7/12

Code:

*-----------*
|...|7.4|...|
|17.|.9.|.52|
|.3.|.1.|.9.|
|---+---+---|
|36.|.8.|.74|
|..9|...|6..|
|28.|.6.|.19|
|---+---+---|
|.1.|.4.|.3.|
|54.|.7.|.26|
|...|6.8|...|
*-----------*

Play/Print online

aran

Joined: 19 Apr 2010
Posts: 70

Posted: Mon May 07, 2012 10:25 am Post subject:

extreme 7.5.12 after ssts

Code:

*-----------------------------------------------------------*
| 89 29 258 | 7 35 4 | 138 6 18 |
| 1 7 468 | 38 9 36 | 348 5 2 |
| 468 3 4568 | 258 1 256 | 478 9 78 |
|-------------------+-------------------+-------------------|
| 3 6 1 | 59 8 59 | 2 7 4 |
| 47 5 9 | 14 2 17 | 6 8 3 |
| 2 8 47 | 34 6 37 | 5 1 9 |
|-------------------+-------------------+-------------------|
| 6789 1 678 | 259 4 259 | 789 3 578 |
| 5 4 38 | 139 7 139 | 89 2 6 |
| 79 29 237 | 6 35 8 | 179 4 157 |
*-----------------------------------------------------------*

UP259r347c46=(368r23c46=5r7c46-(5=3)r9c5) : =><3>r1c5
sstste

arkietech

Joined: 31 Jul 2008
Posts: 1834
Location: Northwest Arkansas USA

Posted: Mon May 07, 2012 12:17 pm Post subject:

aran wrote:

extreme 7.5.12 after ssts

Code:

UP259r347c46=(368r23c46=5r7c46-(5=3)r9c5) : =><3>r1c5
sstste

Nice-- I like Very Happy

I assume UP means "Unique Positions" This puzzle certainly is that! I found three wings, each will solve the puzzle and each using the the 35's in r19c5.

H3-wing
(3=5)r9c5-r1c5=(5-2)r1c3=(2)r9c3 => r9c3<>3; stte

(5=3)r1c5-r9c5=(3-2)r9c3=(2)r1c3 => r1c3<>5; stte

or
L3-wing
(5)r1c5=(5-2)r1c3=(2-3)r9c3=(3) => r1c5<>3, r9c5<>5; stte

There is more to this than meets the eye!

Last edited by arkietech on Mon May 07, 2012 11:47 pm; edited 1 time in total

tlanglet

Joined: 17 Oct 2007
Posts: 2468
Location: Northern California Foothills

Posted: Mon May 07, 2012 1:25 pm Post subject:

I also found the DP(259) pattern, which I understood to be a BUG-Lite+4, but I used the external SIS conditions: (5)r3c3,(9)r8c46 (3)r9c5=(3)r9c3-(3=8=9)r8c37-BUG-Lite(259)r237c46[(9)r8c46=(5)r3c3]-r1c3=r1c5; r1c5<>3 Ted

ronk

Joined: 07 May 2006
Posts: 398

Posted: Mon May 07, 2012 2:21 pm Post subject:

tlanglet wrote:

I also found the DP(259) pattern, which I understood to be a BUG-Lite+4, but I used the external SIS conditions: (5)r3c3,(9)r8c46

(3)r9c5=(3)r9c3-(3=8=9)r8c37-BUG-Lite(259)r237c46[(9)r8c46=(5)r3c3]-r1c3=r1c5; r1c5<>3

Viewed as a (259)MUG:r347c46, the only extra candidates are (68)r3c46 for a quantum naked triple ("qnt"), quantum because the two cells of the MUG behave as one. Then (368)qnt:r23c46 => r1c5<>3 directly.

DonM

Joined: 15 Sep 2009
Posts: 51

Posted: Mon May 07, 2012 10:25 pm Post subject:

ronk wrote:

Do my eyes deceive me? Ronk is using the term 'quantum'? Awaiting 'dual-linked ALS'! Very Happy

ronk

Joined: 07 May 2006
Posts: 398

Posted: Mon May 07, 2012 11:34 pm Post subject:

DonM wrote:

ronk wrote:

Do my eyes deceive me? Ronk is using the term 'quantum'? Awaiting 'dual-linked ALS'! Very Happy

Your recollection is incorrect. Following MadOverlord's lead, I've been using the quantum term for the Type 3 UR cells with extra candidates for many years, probably since 2006.

tlanglet

Joined: 17 Oct 2007
Posts: 2468
Location: Northern California Foothills

Posted: Tue May 08, 2012 1:48 pm Post subject:

ronk wrote:

tlanglet wrote:

Ron,

How do you test if a pattern is a potential MUG? For potential URs and BUG-Lites, I understand the concept of each bivalue occurring exactly twice in its row, column and box. I am not aware of a "practical" technique to check for MUGs.

Ted

aran

Joined: 19 Apr 2010
Posts: 70

Posted: Tue May 08, 2012 2:02 pm Post subject:

arkietech

Quote:

I assume UP means "Unique Positions"

Yes...or even "Unique Position" Smile

... since it is only arises (in a chain) in the singular form : "not Unique position implies"

tlanglet

Quote:

I also found the DP(259) pattern, which I understood to be a BUG-Lite+4, but I used the external SIS conditions: (5)r3c3,(9)r8c46

Quite right
UP is just a personal "generic" term for BUGs, MUGs and any other UGs not yet invented !

ronk

Quote:

Yes indeed, the quantum single 5 in r7c46 was entirely surplus to requirements

Don

Quote:

Do my eyes deceive me? Ronk is using the term 'quantum'? Awaiting 'dual-linked ALS'!

Has there been a "quantum issue"...?
To me, quantum is the relevant term when n candidates in n+x cells behave as an n-locked set with a pertinent consequence...(where x=1 usually)
(without the pertinence any bilocal candidate would be a quantum single).

Marty R.

Joined: 12 Feb 2006
Posts: 5770
Location: Rochester, NY, USA

Posted: Tue May 08, 2012 6:14 pm Post subject:

Quote:

Ron,

Why are the only extra candidates in box 2? What about r7c46? Maybe I'm unclear as to what DP we're trying to avoid.

Luke451

Joined: 20 Apr 2008
Posts: 310
Location: Southern Northern California

Posted: Tue May 08, 2012 6:19 pm Post subject:

tlanglet wrote:

How do you test if a pattern is a potential MUG? For potential URs and BUG-Lites, I understand the concept of each bivalue occurring exactly twice in its row, column and box. I am not aware of a "practical" technique to check for MUGs.

Hi, Ted.

It's been said that a MUG should be reducible to overlapping
BUG-Lites. Myth Jellies had a method that has been recently
updated by Julp on the Eureka forum. Keeping in mind that a UR
is the simplest of all BUG-Lites, here's a way to break down this
pattern.

Code:

*-----------------------------------------------------------*
| 89 29 258 | 7 35 4 | 138 6 18 |
| 1 7 468 | 38 9 36 | 348 5 2 |
| 468 3 4568 |*25+8 1 *25+6 | 478 9 78 |
|-------------------+-------------------+-------------------|
| 3 6 1 |*59 8 *59 | 2 7 4 |
| 47 5 9 | 14 2 17 | 6 8 3 |
| 2 8 47 | 34 6 37 | 5 1 9 |
|-------------------+-------------------+-------------------|
| 6789 1 678 |*259 4 *259 | 789 3 578 |
| 5 4 38 | 139 7 139 | 89 2 6 |
| 79 29 237 | 6 35 8 | 179 4 157 |
*-----------------------------------------------------------*

r7c46 has 3 candidates in 2 cells. When the puzzle is solved
there can only be 2 (different) candidates in those 2 cells.
So one of 2,5,9 has to go from both cells.

If r7c46<>2 then we have a UR(59)r47c46
If r7c46<>9 then we have a UR(25)r37c46
If r7c46<>5 then we have a BUG-Lite(259)r347c46

All possible reductions lead to a BUG-Lite, so there's your MUG.

Incidentally, Myth Jellies' seminal thread has been recently resurrected at Players' to great effect,
starting with ronk's post Feb 25, 2012. Not all MUGs are as conveniently laid out as the one above!

DonM

Joined: 15 Sep 2009
Posts: 51

Posted: Wed May 09, 2012 6:16 am Post subject:

aran wrote:

Don

Quote:

Do my eyes deceive me? Ronk is using the term 'quantum'? Awaiting 'dual-linked ALS'!

Oh, I was just having a little joke with ronk given his well-known sense of humor. I was always happy with the term 'quantum' as used broadly by Steve K. Ronk apparently preferred a more restricted use as per a poster from yesteryear on the original Players forum. Some discussion on the subject occurred here:

http://forum.enjoysudoku.com/quantums-t30088.html

ronk

Joined: 07 May 2006
Posts: 398

Posted: Wed May 09, 2012 11:23 am Post subject:

DonM wrote:

I was just having a little joke with ronk given his well-known sense of humor.

Your jokes remind me of this oldie. Smile

Quote:

A man is sent to prison for the first time. At night, the lights in the cell block are turned off, and his cellmate goes over to the bars and yells, "Number twelve!" The whole cell block breaks out laughing. A few minutes later, somebody else in the cell block yells, "Number four!" Again, the whole cell block breaks out laughing.

The new guy asks his cellmate what's going on. "Well," says the older prisoner, "we've all been in this here prison for so long, we all know the same jokes. So we just yell out the number instead of saying the whole joke."

So the new guy walks up to the bars and yells, "Number six!" There was dead silence in the cell block. He asks the older prisoner, "What's wrong? Why didn't I get any laughs?"

"Well," said the older man, "sometimes it's not the joke, but how you tell it."

ronk

Joined: 07 May 2006
Posts: 398

Posted: Wed May 09, 2012 12:05 pm Post subject:

tlanglet wrote:

Not as easy for MUGs as for BUG-Lites. As Luke451 wrote in response to your question, one needs to show that every valid reduction of a prospective MUG leads to a BUG-Lite, [edit2: a UR, or an overlay of BUG-Lites and URs.]

Consider these two simple MUGs, more precisely MUG+0 patterns, which IMO the manual solver should commit to memory.

Code:

. abc . | . abc . | . . . . abc . | . abc . | . abc .
. abc . | . abc . | . . . . . . | . . . | . . .
. abc . | . abc . | . . . . abc . | . abc . | . abc .
----------+----------+---------- -----------+----------+----------
. . . | . . . | . . . . . . | . . . | . . .

For the left: Due to the 'abc' locked sets, reduction placements '(x)' are not possible in b1 and b2. Furthermore, no two placements may occur in one mini-row of b3. That leaves ...

Code:

. bc . | . bc . |(a) . .
. ac . | . ac . | . (b) .
. ab . | . ab . |(c) . .
----------+----------+----------
. . . | . . . | . . .

For the right: Due to 'abc' locked sets, reduction placements are not possible in r1 and r3. Furthermore, no two placements may occur in one mini-row of r2. That leaves ...

Code:

. bc . | . ac . | . ab .
. . (a) | . (b) . |(c) . .
. bc . | . ac . | . ab .
----------+----------+----------
. . . | . . . | . . .

[edit1: "reduction placement" was incorrectly termed "MUG-buster placement"]

Last edited by ronk on Sun May 13, 2012 9:26 pm; edited 1 time in total

aran

Joined: 19 Apr 2010
Posts: 70

Posted: Wed May 09, 2012 3:55 pm Post subject:

Just to sum the approaches to a position P which looks potentially UP, as illustrated by Luke and Ronk - assign a candidate to one cell, see how P dissolves to a position which is recognisably UP (or not), and if so, observe (most likely) that this logic was candidate neutral, and so conclude P is UP - assume P is not UP, work with the constraints (eg in example 1, that no two of abc can appear on the same line in b3), generate a recognisable UP (or not) and if so conclude : UP=UP =>UP

Display posts from previous:

	dailysudoku.com Forum Index -> Other puzzles	All times are GMT
Page 1 of 1

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