| View previous topic :: View next topic   | 
	
	
	
		| Author | 
		Message | 
	
	
		Captain Pete
 
 
  Joined: 09 Jun 2007 Posts: 55 Location: Oley, PA
  | 
		
			
				 Posted: Tue May 05, 2009 12:46 pm    Post subject: Menneske SH | 
				     | 
			 
			
				
  | 
			 
			
				I'm stuck.  What technique gets me to the next number?
 
 
 	  | Code: | 	 		  
 
+--------------+-----------+------------+
 
| 358  356 1   | 4 9    78 | 358 2   67 |
 
| 2    4   368 | 5 1    78 | 38  9   67 |
 
| 578  57  9   | 6 2    3  | 4   58  1  |
 
+--------------+-----------+------------+
 
| 9    136 368 | 2 3568 56 | 7   158 4  |
 
| 458  2   468 | 7 568  1  | 58  3   9  |
 
| 1378 137 5   | 9 38   4  | 6   18  2  |
 
+--------------+-----------+------------+
 
| 145  15  7   | 8 56   2  | 9   46  3  |
 
| 6    8   2   | 3 4    9  | 1   7   5  |
 
| 345  9   34  | 1 7    56 | 2   46  8  |
 
+--------------+-----------+------------+
 
 | 	  
 
Play this puzzle online at the Daily Sudoku site | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		keith
 
 
  Joined: 19 Sep 2005 Posts: 3355 Location: near Detroit, Michigan, USA
  | 
		
			
				 Posted: Tue May 05, 2009 1:56 pm    Post subject:  | 
				     | 
			 
			
				
  | 
			 
			
				The candidate <5> in R5C1 does not belong.
 
 
After that, I think you need a chain.  I do not see any simple named pattern.
 
 
Keith | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		arkietech
 
 
  Joined: 31 Jul 2008 Posts: 1834 Location: Northwest Arkansas USA
  | 
		
			
				 Posted: Tue May 05, 2009 3:17 pm    Post subject:  | 
				     | 
			 
			
				
  | 
			 
			
				 	  | Code: | 	 		   *-----------------------------------------------------------*
 
 | 358   356   1     | 4     9     78    | 358   2     67    |
 
 | 2     4     368   | 5     1     78    | 38    9     67    |
 
 |*578   57    9     | 6     2     3     | 4    *58    1     |
 
 |-------------------+-------------------+-------------------|
 
 | 9     136   368   | 2     3568  56    | 7    #158   4     |
 
 | 4-8   2     468   | 7     568   1     |*58    3     9     |
 
 | 1378  137   5     | 9     38    4     | 6    #18    2     |
 
 |-------------------+-------------------+-------------------|
 
 | 145   15    7     | 8     56    2     | 9     46    3     |
 
 | 6     8     2     | 3     4     9     | 1     7     5     |
 
 | 345   9     34    | 1     7     56    | 2     46    8     |
 
 *-----------------------------------------------------------*
 
er on 8 (empty rectangle) either r3c1 is an 8 or r5c7 is 8 making r5c1<>8 | 	  Hope this helps. | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		Marty R.
 
 
  Joined: 12 Feb 2006 Posts: 5770 Location: Rochester, NY, USA
  | 
		
			
				 Posted: Tue May 05, 2009 3:51 pm    Post subject:  | 
				     | 
			 
			
				
  | 
			 
			
				| Alternatively, an M-Wing on the 58s in boxes 36 eliminates the same 8 as the ER. | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		Captain Pete
 
 
  Joined: 09 Jun 2007 Posts: 55 Location: Oley, PA
  | 
		
			
				 Posted: Tue May 05, 2009 4:03 pm    Post subject:  | 
				     | 
			 
			
				
  | 
			 
			
				| Architech,  Thanks for the reply.  I've just been reading up on ERs, and I'm trying to use this puzzle as a tutorial.  I see the strong link in row 3, but I can't see the ER.  In your response, you identified two cells with #.  What do those cells represent?  In using the ER technique, is it necessary to screen each box for each number? | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		Marty R.
 
 
  Joined: 12 Feb 2006 Posts: 5770 Location: Rochester, NY, USA
  | 
		
			
				 Posted: Tue May 05, 2009 5:08 pm    Post subject:  | 
				     | 
			 
			
				
  | 
			 
			
				| An ER is where a number exists in just two lines within a box. Where the lines intersect is the ERI (Empty Rectangle Intersection). In this case the ERI is r5c8. The ERI needs to see one number of a strong link, the 8s in row 3 in this case. Any cell seeing both the ERI and the second half of the strong link (r5c1 in this case) can't contain the number. | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		arkietech
 
 
  Joined: 31 Jul 2008 Posts: 1834 Location: Northwest Arkansas USA
  | 
		
			
				 Posted: Tue May 05, 2009 7:31 pm    Post subject:  | 
				     | 
			 
			
				
  | 
			 
			
				Captain Pete,
 
 
Marty's answer was better than mine. I find er'[s by the logic of -- if this is on then this is off. The #'s  are off if either r3c8 or r5c7 is on.. This is how I spot er's. I hope I have not confused you. | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		Captain Pete
 
 
  Joined: 09 Jun 2007 Posts: 55 Location: Oley, PA
  | 
		
			
				 Posted: Tue May 05, 2009 10:03 pm    Post subject:  | 
				     | 
			 
			
				
  | 
			 
			
				Thanks to all.  Can the ERI contain one of the numbers in the comlementary strong link?  For example, in this puzzle if the strong link contained a 4, would the 4s in box 7 constitute a ER with 345 being the ERI?
 
 
And, what prompts you to look for a specific number. in this case 8.  Is there a tipoff of some kind?
 
 
M-wing, Oh no.... | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		storm_norm
 
 
  Joined: 18 Oct 2007 Posts: 1741
 
  | 
		
			
				 Posted: Wed May 06, 2009 12:10 am    Post subject:  | 
				     | 
			 
			
				
  | 
			 
			
				ER is a HINGE
 
as it relates to the candidates in a particular Box
 
 
 
 
 
notice how the 8's in row 3 are conjugate. both can't be false.
 
if the 8 in r3c1 is true, then the 8 in r5c1 can't be true.
 
 
now, the tricky part.
 
 
if 8 in r3c1 isn't true, then the 8 in r3c8 is true
 
then follow the hinge around
 
8 in r46c8 not true
 
8 in r5c7 true
 
 and it points at the 8 in r5c1.  that 8 can't be true.
 
 
so in both situations, the 8 in r5c1 can't be true.
 
 
another one...
 
this is Danny's "puzzle NR-005"
 
 
 
if you want to be imaginative with the hinge,
 
imagine using colors to label the 8's in row 4, red and green.
 
the HINGE allows the 8 in r4c6 to SEE AROUND THE CORNER.    
 
 
now that the 8 in r4c6 can see around the corner at the 8 in r1c1
 
the 8 in r1c1 can now see two colors.
 
in the world of coloring, if a candidate can see two colors, its toast.
  Last edited by storm_norm on Wed May 06, 2009 7:01 am; edited 3 times in total | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		Marty R.
 
 
  Joined: 12 Feb 2006 Posts: 5770 Location: Rochester, NY, USA
  | 
		
			
				 Posted: Wed May 06, 2009 12:27 am    Post subject:  | 
				     | 
			 
			
				
  | 
			 
			
				 	  | Quote: | 	 		  | And, what prompts you to look for a specific number. in this case 8. Is there a tipoff of some kind?  | 	  
 
I don't know what others do, but I just look at every box and scan the numbers to see which meet the criterion of being in two lines, then check for a strong link. 
 
 
If there's a more efficient way, I'd be interested in knowing it. | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		Captain Pete
 
 
  Joined: 09 Jun 2007 Posts: 55 Location: Oley, PA
  | 
		
			
				 Posted: Wed May 06, 2009 12:13 pm    Post subject:  | 
				     | 
			 
			
				
  | 
			 
			
				| I think I have it.  Thanks to all! | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		tlanglet
 
 
  Joined: 17 Oct 2007 Posts: 2468 Location: Northern California Foothills
  | 
		
			
				 Posted: Wed May 06, 2009 12:17 pm    Post subject:  | 
				     | 
			 
			
				
  | 
			 
			
				 	  | Captain Pete wrote: | 	 		  | And, what prompts you to look for a specific number. in this case 8. Is there a tipoff of some kind?  | 	  
 
 
I look for strong links (conjugate pairs) first because they are the source of multiple patterns: skyscrapers, kites, ERs, etc. Thus when I find a conjugate pair, I scan for all patterns that could use it. I find this approach more efficient since I only handle the conjugate pair one time.  
 
 
But in reality, it is whatever one is most comfortable with; try different approaches and form your own style.
 
 
Ted | 
			 
		  | 
	
	
		| Back to top | 
		 | 
	
	
		  | 
	
	
		 |