Tough Sudoku

Another good photo Keyan

New day, new game.

12:47 Way too easy for a 'Tough'

1) Start at 23 filled - the given puzzle. UP to 28. (Unique Possibilities).
2) Pair 89 at hi2 forbids bdf2,h3=8 and abd2,h3=9 UP 29
3a) Pair 57 at ab2 forbids b3,df2=5
3b) Pair 26 at d27 forbids d8=6
3c) Pair 58 at d8e9 forbids f89=58
3d) fc on 8's: f5 == f3 -- b3 == c1 forbids c5=8 UP 81

Sets: 5(2) =10. Max depth 2. Rating 5(.03) = .15 - almost trivial

Good Maen to all! everyone seeing all the sights, but where are the camel rides??

15.23 What desert?
'Politics: A combination of 'poli' meaning many, and 'tics' meaning irritating little bugs.

Alternative proof - one elimination only:
1) Start at 23 filled - the given puzzle. UP to 28. (Unique Possibilities).
2)
f5=8 f5=4 f5=5
0000 f4=4 f4=5
f3=8 f3=4 f3=5
establishes f5=8 == f3=8, but f3=8 -- e1=8 == c1=8 forbids c5=8
UP 81
Sets: 4. depth 4. rating .15

Andrei or anyone - any help on how to express this as a sigle matrix or single chain would be appreciated.

I ask this question not because of today's puzzle, but rather in some of the more complicated puzzles, an implied strong link sometimes needs to be used, and the representation of such a link can lose transparency.

Perhaps: ??
2)
f5=8 f5=4 f5=4 *
0000 f4=4 f4=5 * => f5=8 f3=8
f3=8 f3=4 f3=5 * c1=8 e1=8
*****************
Forbids c5=8

Perhaps: ?? (darn spacing!)
2)
f5=8 f5=4 f5=4 *
0000 f4=4 f4=5 * => f5=8 f3=8
f3=8 f3=4 f3=5 * c1=8 e1=8
*****************
Forbids c5=8 UP 81

Perhaps: ?? (darn spacing again!)
2)
f5=8 f5=4 f5=4 *
0000 f4=4 f4=5 * =>f5=8 f3=8
f3=8 f3=4 f3=5 * =>c1=8 e1=8
*****************
Forbids c5=8 UP 81

6:25

Sorry about my density:
step 2) representation with clarity:
c1=8 e1=8
0000 f3=8 f3=4 f3=5
0000 0000 f4=4 f4=5
f5=8 0000 f5=4 f5=5
forbids c5=8. UP 81.

14:05 got a bit stuck and had to resort to guesses (calculated, of course!). Like the definition of politics Suz!.

11:43

poli in politics does not originate from poli,meaning in Greek much,but from polis,meaning initially town,and later implying govermental structure.As for the puzzle it was easier than others!

10:05 but had 1 guess. We have a small respite from the heat today. It's only 84, but very humid. Afternoon/evening thunderstorms predicted. Hope there's some rain in the future.

Very interesting picture. Wonder what's on the other side of the dune!

mAen all. I like SUZ's definition of politics the best. Tee Hee.

1) Start 23, unique possibilites to (UP) 28.
2) Naked pair 89 at hi2 forbids h3=89, a2=9, b2=89, d2=89, f2=8. UP 29.
3) Naked triplet f3=458, f4=45, f5=458 forbids f2=5, f8=58, f9=58.
4) f5=8 == f3=8 -- b3=8 == c1=8 forbids c5=8. UP 81.

Nice pictures today!

To Steve from OH

I would have simply cited multi-coloring as the
reason c5 (and b3, e1) can't be 8 (one side of
color a sees both sides of color b and is thus
false).

Still having problems finding FC's. The difficulty -
Often as not, after grueling mental effort, some
innocent bystander gets eliminated, and little or
no progress is made. What motivates you in your
search for chains ? For myself, at least even odds
of a win must be in sight.

Hi Pat!

Today's puzzle the coloring elimination in 8's is fairly easy - and really there is no search for chains involved - coloring eliminations are part of my standard search (like pairs, triplets, etc) - and this one just sort of jumps out - it is hard to describe the thought process on one like today becuase it is obvious upon inspection.

On more difficult puzzles, however - the choice of a search starting point is more problematic. The typical idea, after hitting the 'stuck' point, is print out the puzzle, circle the strong links - all of the two party ones - and look where there is a critical mass of strong link convergence. This helps to insure that the elimination is not just an 'innocent bystander'. It also points up the most obvious places to look for simple fc's. If that investigation fails, then the bigger guns are pulled out - ALS, condidtional coloring, etc. (I have a very long list of things that really have obscure names - in my mind at least.) Same idea though for places to look - convergence of conflicts. Generally this results in at least some progress. Then, it is simply a matter of defining the bare bones contradiction, writing it out, and remembering that most puzzles then follow a pattern - same idea is used more than once(at least part of it). Of course, these are just generalizations - and like all generalizations, one must know when to abandon them too. Experience seems to make most of this process automatic, and the evaluation of likely locations to search becomes more intuitive then strictly reasoned - I think it is not really intuition, just a lot of processing that goes on without really thinking too hard about it - previous people at this site referred to this as 'pattern recognition'.

Hi Pat (again)
If your goal is speed, then the process above is not efficient. Guessing is a far more efficient use of time to solve most very difficult puzzles. If your goal is a reasoned approach and learning about various interesting elimination patterns, then the process above may be helpful. I make few representations of my time, becuase it is generally not fast - unless the puzzle is easy like today's. Even with a puzzle like today's, the time from looking at the puzzle to first published proof was about 15 minutes. On a more difficult puzzle, like yesterday's - probably 30 to 60 minutes from start to published proof. On a very difficult puzzle like that of 7/22, probably I spend 90 minutes or so. On monsters - rating over 2 - 2 hours plus. If I just wanted to guess and not have a reasoned approach - most of the times probably would be much, much faster.

To Steve
thanx for the input. I fully agree that guessing
is an ignoble end. 'critical mass of strong link
convergence' is a formidable and mildly alarming
concept which I shall ponder in a well lit room.
Just about every technique other than FC's can be
coded into a compact algorithm. I guess my next step
will be to code a neural-net (perhaps using a GA
instead of back-propogation) to find these patterns.

8.29 bit easy
another great picture from Keyan
have a great day/night one and all

4:55 after setup. Definitely personal best but I think not deserved since very easy puzzle.

Wow! The red of that sand is amazing!

Christine: What part of the state are you in? We're in the Midlands and have only seen clouds (but did get a short storm the other night) My brother's in the upstate and swears they all pass him by. Glad we're not farmers that need the rain. Keyan,from a distance the desert looks like Carolina red clay. meaN everyone.

Wow Jill - are you talking about the tough puzzle or one of the others? Took forever and I had to guess. I don't know how to understand Steve's solutions. Is there a place to go to learn this?

Hi Suz - its about an hour from Dubai towards Sharjah. A great 4WD tour.

Steve,
Using my own un-orthodoxed techniques, I got to 28. I had found all the pairs you did in steps 2-3c. Then I got stuck and followed your 3d chain. In my solution, it logically forbid the 8 from c5, but I could not UP from it...

I still had two 8's in col d, row 6 and in the middle block. I had 3 8's in row 5. What did I miss?

I am really trying to understand your techniques(and others) for solving(not guessing) the toughs and am getting pretty close.

Thanks again to all of you that post proofs.

Hi Jeff!
AFter c5<>8, you should have only one 8 in column c at c1. Thus:
c1=8%Col, e1=4%Cell, g1=1%Cell, a1=1%Cell,h3=4%Box, i8=4%Box, b8=4%Box, i6=1%Cell, i9=8%Cell, etc. c1=8%Col means c1=8 because it is the only 8 left in the Column, for example. Hopefully this helps.

Hi Toni!
Check proofs link at bottom of web page.

Steve,
Yes you are correct. As usual. I see my mistake now. I forbid 8 from d5, not c5(dang off by 1 error). Which may have been correct, but it is too late for me to verify. Maybe if I ran the fc again it would have forbid that too.

Almost. Again, thanks.

Maybe if the columns were labeled a-i and the rows 1-9 it would be slightly easier...

Stunning dune photo - reminds me of Sossusvlei in Namibia (South West Africa)- the highest dunes in the world. After rains - which are usually heavy and infrequent - the ancient river beds at the base of the dunes turn green from new shoots. Quite spectacular.

7:22 Maen

13:37 Fairly easy tough. Just wish I could get under 4:00 on easy!!!

Thanks for all the great photos Keyan! Where are you going next?

Maen

yellow river bu i.p daily

Aug. 25, 2007.

50.45