forcing chains

Advanced methods and approaches for solving Sudoku puzzles

forcing chains

Postby michgirl » Tue Jan 24, 2006 11:05 pm

I am unable to figure out what cell to start in or how far to go with it so when I attempt a fiendish puzzle I get stuck every time. Although I go to the step by step solutions I only get the answer not the steps to take to use this technique successfully. I also would like to get a book explaining and giving examples of advanced solving techniques. Barnes and Noble only seems to have books with puzzles. Any help would be appreciated!
michgirl
 
Posts: 4
Joined: 05 December 2005

Postby Jeff » Wed Jan 25, 2006 3:16 am

Welcome to the forum, michgirl.

In this forum, you will find a lot of discussions on forcing chains, but very little on how to identify them. Here is a thread which may help you out.
Jeff
 
Posts: 708
Joined: 01 August 2005

Postby Myth Jellies » Wed Jan 25, 2006 3:20 am

[Edit: unfortunately, Jeff and I posted at almost the same time, so it looks like I am responding to his post and implying that it is too complicated for beginners, which I most assuredly am not:) .]

I would suggest learning to walk before you run, and start with simple bivalue xy-chains. To find these, you can choose any bivalued cell (lets call the contents 'ab'). Then check if that cell "sees" (shares a group) another bivalue cell containing 'b', like bc. Then check if that cell sees another bivalue cell containing c, etc. etc. until you either find nothing else, or you find a cell that gets you back to 'a'. Then you have found a chain of cells which looks something like ab - bc - cd - ... - ef - fa. Now, you can remove all the a's from any cell which sees both the ab-cell and the fa-cell.

Candidate BUG grids usually make good practice grids for these simple bivalue chains. You can find a lot of them in this thread.
Myth Jellies
 
Posts: 593
Joined: 19 September 2005

Postby sweetbix » Wed Jan 25, 2006 11:16 am

I would be interested to know if most new players are really interested in bilcation/bivalue plots as described in your thread, Jeff.

As a newish player I would have to say that this method of identifying forcing chains seems pretty advanced. With or without a computer solver I identify chains only by trial and error which is I how I would describe Myth Jellies approach i.e. select a bivalue cell and follow it to see where it goes.

I would add that my idea of forcing chains is not just the loop but also the chain that leads to a contradiction and the double implication chain which I would say is easier to follow.

I would be interested to know if other people really use a method other than trial and eror to locate a chain.
sweetbix
 
Posts: 58
Joined: 10 December 2005

Postby Jeff » Wed Jan 25, 2006 12:00 pm

sweetbix wrote:I would be interested to know if most new players are really interested in bilcation/bivalue plots as described in your thread, Jeff.

As a newish player I would have to say that this method of identifying forcing chains seems pretty advanced.

Hi Sweetbix, thanks for sharing your thought.

I know that most players don't have the patience to carry out the B/B plot. That is why I have also introduced the x-cycle and xy-chain for elementary level players. These 2 types of nice loops do not require the construction of a B/B plot.

B/B plot is to be used as a general means for identification of forcing chains including combination loops. So, there is something for everyone.
Jeff
 
Posts: 708
Joined: 01 August 2005

Postby CathyW » Wed Jan 25, 2006 4:00 pm

I've always struggled with forcing chains too - particularly on identifying them. However, the bivalue xy-chains, which I think I've now got the hang of, are very useful when other techniques e.g. up to x-wing haven't solved a puzzle.
CathyW
 
Posts: 316
Joined: 20 June 2005


Return to Advanced solving techniques