View previous topic :: View next topic 
Author 
Message 
Marty R.
Joined: 12 Feb 2006 Posts: 5466 Location: Rochester, NY, USA

Posted: Sat Mar 12, 2011 5:58 am Post subject: March 12 VH 


Where are the usual thread starters?
I never looked for XWings and Skyscrapers. Just looking for other wings, I didn't notice a onestepper and used two XYWings, 458 and 564. 

Back to top 


cgordon
Joined: 04 May 2007 Posts: 769 Location: ontario, canada

Posted: Sat Mar 12, 2011 6:00 pm Post subject: 


Quote:  I never looked for XWings and Skyscrapers. Just looking for other wings, I didn't notice a onestepper and used two XYWings, 458 and 564. 
Marty: I looked for everything but xy wings cos I'm xywing challenged. But still ended up with the same solution as you. 

Back to top 


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

Posted: Sun Mar 13, 2011 5:09 am Post subject: 


Here is a one step solution using an "almost" xywing.
Code:  **
 35 7 6  248 14 1248  9 35 12 
 35 8 1  256 9 236  2356 7 4 
 2 9 4  567 1356 1367  356 8 135 
++
 479 45 578  3 2 48  1 459 6 
 49 36 2  1 46 5  7 349 8 
 1 36 58 #46=8 7 9  235 345 235 
++
 47 2 57 #45 13 13  8 6 9 
 6 45 9  2457 8 247  35 1 357 
 8 1 3  9 #56 67  4 2 57 
** 
AXYwing (456) vertex (45)r7c4, pincers (46)r6c4 & (56)r9c5 with fin (8)r6c4 (marked with #)
If xywing is true (r6c4=46): r5c5<>6=4
If fin is true then r6c4=8, then r6c3<>8=5, then r7c3<>5, then r7c4=5, then r9c5<>5=6, then r5c5<>6=4
Ted
[A friendly PM pointed out a typo that I edited marked in red: was r7c5, corrected to r7c4.]
Last edited by tlanglet on Sun Mar 13, 2011 12:53 pm; edited 1 time in total 

Back to top 


cgordon
Joined: 04 May 2007 Posts: 769 Location: ontario, canada

Posted: Sun Mar 13, 2011 11:45 am Post subject: 


Ted: I don't really get that "almost xy wing". Why does R6C4 have to be a 46 or an 8  why not a 48 or a 68. And if one has to follow on with all those "ifs"  is it really a one stepper? 

Back to top 


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

Posted: Sun Mar 13, 2011 1:19 pm Post subject: 


cgordon wrote:  Ted: I don't really get that "almost xy wing". Why does R6C4 have to be a 46 or an 8  why not a 48 or a 68. And if one has to follow on with all those "ifs"  is it really a one stepper? 
I am sure that others can provide the fundamental math/logic explanation, but here is my "working" view.
In fact, r6c4 does NOT have to be a 46 or an 8, but it MAY be treated as a 46 or an 8! If you prefer, you could also view it as a 48 and a 6 or 68 and a 4. Any grouping of digits within a cell may be formed including each digit individually. Grouping r6c4=46 or r6c4=8 is useful in this puzzle since it then provides part of a xywing pattern.
Grouping of cells in an integral part of handling Type n URs. Consider the following Type 3 UR pattern.
Code:  ++
    
 12  123 
    
++
 12  124 
    
    
++
    
    
    
++ 
In this case, we treat the cells as
Code:  ++
    
 12  12=3
    
++
 12  12=4
    
    
++
    
    
    
++ 
which allows us to form a (34) pseudocell.
Again, I can't provide the gory details as to why you can group digits in any arbitrary manner, but you may. Simply group them in the way that is beneficial to you; just remember that the implication of using all the digits must be considered.
Hope this helps a little bit............
Ted 

Back to top 


peterj
Joined: 26 Mar 2010 Posts: 974 Location: London, UK

Posted: Sun Mar 13, 2011 9:35 pm Post subject: 


cgordon wrote:  Why does R6C4 have to be a 46 or an 8  why not a 48 or a 68. 
r6c4 is never going to "be" 46 ... or 48 or 68 for that matter. It is going to be exactly one of the candidate set {468}. It is not neccesary to consider every permutation of pairs in the candidate set, only to make sure each candidate occurs at least once in the streams you consider.
In "almost pattern" logic one of the digits (the "fin") is usually the start of some chain or other.. this leaves a reduced candidate set that matches a shortcut we think of as a "pattern" e.g. xywing in this case.
In "kraken" logic each digit is usually considered one at a time  which you could do in this case by following each digit after the fin down to each pincer of the wing.
(I saw an unusual case recently where {ABC} could be treated as {AB} to make an xywing and {BC} to make a wwing. This is fine even though B is used twice  it just means if you followed the kraken logic you have two chains to make the elimination starting with B.)
Enough....! 

Back to top 


cgordon
Joined: 04 May 2007 Posts: 769 Location: ontario, canada

Posted: Sun Mar 13, 2011 10:13 pm Post subject: 


Thanks Ted and Peterj  I appreciate the detailed responses  but to be honest, for me, these solutions go from fun to painful. In fact after 4 years I’m even wondering whether the Daily Sudukos even need a forum – since the daily solutions are invariably basic "wings and things"  and the more complex solutions generate a very minimal and esoteric response.
Sorry  just whingin' 

Back to top 


peterj
Joined: 26 Mar 2010 Posts: 974 Location: London, UK

Posted: Sun Mar 13, 2011 10:25 pm Post subject: 


cgordon, I agree with you. The Daily VHs really lack variety  and finding a "different" solution often involves "unnatural acts"!
I think Danny's moderates and oldVHs are a well judged stepup. 

Back to top 


cgordon
Joined: 04 May 2007 Posts: 769 Location: ontario, canada

Posted: Sun Mar 13, 2011 10:49 pm Post subject: 


[ Quote:  cgordon, I agree with you. The Daily VHs really lack variety  and finding a "different" solution often involves "unnatural acts"! 
I agree with you too Peter J  though it's been a while since I was involved in unnatural acts  at least a week!
I think I am gonna shift my focus to daj or others.
Cheers, Craig 

Back to top 


