dailysudoku.com

Earl · Joined: 30 May 2007 Posts: 677 Location: Victoria, KS

Here is a thorough exercise in "advanced" techniques from Decabit. Thanks to Keith, Marty and Nataraj I now have the correct format.

Earl

Marty R. · Posted: Fri Jun 05, 2009 4:31 pm Post subject:

Earl,

Highlight the whole Decabit grid, copy it and paste it into Nataraj's format tool. You might want to check the Site Help sub-forum and look at Nataraj's thread about formatting puzzles from other sources and my thread about aligned grids.

P.S. Earl, that Decabit site is new to me. What level of puzzle there do you recommend?

storm_norm · Joined: 18 Oct 2007 Posts: 1741

arkietech · Posted: Sat Jun 06, 2009 2:52 pm Post subject:

storm_norm · Joined: 18 Oct 2007 Posts: 1741

Dan,
I am uncertain of the names and configurations of all the fish. In my mind's eye, I see non x-wing/finned x-wing patterns as coloring or multicoloring chains w/wo groups. its easier for me to pay more attention to the progression of links in a singles chain rather than trying to associate the pattern to a specific fishy name.

arkietech · Posted: Sat Jun 06, 2009 8:38 pm Post subject:

Norm,

Thanks for the reply. I guess it is how you get started. I could not handle coloring. So I started looking for x-wing and fish patterns.

You find a lot more solutions than I do. How do you step through the coloring?

tlanglet · Posted: Sat Jun 06, 2009 8:54 pm Post subject:

I found a skyscraper on 6 which uses the some of the same cells as Norm's grouped coloring chain. The skyscraper is in r3c57 with a grouped cell in r89c5. Using the normal logic for a skyscraper,
if r3c5=6 then r3c7<>6 and r7c7=6; r7c46<>6
if r3c7=6 then r3c5 <> 6 and r89c5=6; r7c46<>6
Thus, r7c7=6

Ted

tlanglet · Posted: Sat Jun 06, 2009 9:12 pm Post subject:

storm_norm · Joined: 18 Oct 2007 Posts: 1741

Dan,
that old cliche comes to mind that there are two sides to every argument.

when you are dealing with single candidate patterns, there are two views..

#1...
Danny explains the set cover theory here....
http://www.dailysudoku.com/sudoku/forums/viewtopic.php?t=3111&sid=a811a6b75ad63d10268931ac1e880455

where he also links to the a fish guide here...
http://www.sudoku.com/boards/viewtopic.php?t=4993

#2...
coloring, A.K.A. singles chains.
the explanation that seems to be given the most attention is the one by Andrew Stuart here...
http://www.scanraid.com/Singles_Chains
and...
http://www.scanraid.com/Multi_Colouring_Strategy

BUT!
coloring has since been upgraded to what is known as Medusa coloring. This not only deals with a single candidate chain but goes on to deal with all chains that are formed from conjugate links of candidates including within bi-value cells.

you can find a discussion about Medusa here on this site
http://www.dailysudoku.com/sudoku/forums/viewtopic.php?t=2395&sid=a811a6b75ad63d10268931ac1e880455

and...
http://www.sudopedia.org/wiki/3D_Medusa

and if you scroll down the sudocue solving guide page to near the bottom...
http://www.sudocue.net/guide.php...
---------

what I find fascninating is that discussions on groups or fins (as seen in the above puzzle) is found quite consistantly in the set cover arguments and not in the medusa coloring arguments.
but there is a catch... Surprised

again, in the Andrew Stuart technique guide he designates these chains and grouped chains as x-cycles... funny how the name xx like that Rolling Eyes

... sometimes you have to look in between the cracks and floor boards to catch them.
http://www.scanraid.com/X_Cycles
http://www.scanraid.com/X_Cycles_Part_2
http://www.scanraid.com/Grouped_X_Cycles

what is important to note is that x-cycles are already dealt with in medusa coloring and in the fishy guides.

Andrew Stuart's definition of a grouped x-cycle is not dealt with in medusa coloring but is dealt with in the fishy guide.
fwiw, I see grouped x-cycles the way its defined in Andrew's guide and not how its defined in the fishy guide and I have combined that with my knowledge of medusa to find grouped chains and coloring patterns. the puzzle in this post is a great example of a grouped x-cycle.

arkietech · Posted: Sat Jun 06, 2009 11:24 pm Post subject:

Norm,

Thanks!!! Very Happy

-- should keep me out of trouble for a while. Cool