dailysudoku.com

daj95376 · Joined: 23 Aug 2008 Posts: 3854

Same Rules!

storm_norm · Joined: 18 Oct 2007 Posts: 1741

Asellus · Posted: Thu Nov 06, 2008 6:09 am Post subject:

Norm,

Your AIC uses a 2-cell ALS along the chain. In a bivalue (the smallest possible ALS), it is obvious that there is a strong link (or weak link, since it is conjugate) between the two digits. In an ALS larger than 2 cells, any two (grouped as necessary) digits within the ALS are strongly linked. Here, you are using the 2-cell 148 ALS in r19c2. So, from that, you could, in principle, use (1)r19c2=(4)r9c2 or (1)r19c2=(8)r19c2 or (8)r19c2=(4)r9c2.

In your AIC, you are using the first choice (which is the only useful one since the grouped <8>s do not weakly link to anything). This is how the complete AIC appears:
(4)r9c1 - ALS[(4)r9c2=(1)r19c2] - (1=5)r3c2 - (5)r7c2=(5-2)r7c1=(2-4)r9c1; r9c1<>4

[Edit to fix typo and clarify.]

daj95376 · Joined: 23 Aug 2008 Posts: 3854

Hint: Don't leave [stack 1].

Asellus · Posted: Thu Nov 06, 2008 2:08 pm Post subject:

Norm,

An alternate view that is equivalent to your AIC is the 2 ALS "xz wing" of {1458} in r139c2 and {245} in r79c1. The shared exclusive <4>s give shared common pincer <5>s in r3c2 and r1c7, eliminating <5> from r7c2. As an AIC, it is:
(5)r7c2 - ALSr139c2[(5)r3c2=(4)r9c2] - ALS[(4)r9c1=(5)r7c1] - (5)r7c2; r7c2<>5

I don't know if that's what Danny is hinting at.

ravel · Joined: 21 Apr 2006 Posts: 536

daj95376 · Joined: 23 Aug 2008 Posts: 3854

storm_norm · Joined: 18 Oct 2007 Posts: 1741

can anyone give me a graphical representation of what a ALS-xy, ALS-xz looks like? the reason I ask is because nothing i have read on the subject has brought me closer to finding them. for my money, these set counting techniques are better left for an algorithm in a basic program. Rolling Eyes

what I can guartentee in return is a fat smile and total gratification once this concept is finally integrated in my blind brain. what I don't want to do is read 4 pages of examples. I want candidates circled, I want arrows, I want lines of thought, I want descriptions of why the candidates have a relationship, etc, etc, etc Exclamation

please???

keith · Posted: Thu Nov 06, 2008 11:11 pm Post subject:

> can anyone give me a graphical representation
> of what a ALS-xy, ALS-xz looks like?

And, please put your reply in the "Solution Techniques" thread.

Thank you,

Keith

daj95376 · Joined: 23 Aug 2008 Posts: 3854

Asellus · Posted: Sun Nov 09, 2008 5:05 am Post subject:

I just want to say that I think that the ALS tutorial in the link from the previous post is excellent! I agree with the author that terms such as "xz-wing" are less useful than grasping the power of ALS in general. (And, nowhere is subset counting used, which, I agree, is more for computers than for people.)