The hardest sudokus (new thread)

Everything about Sudoku that doesn't fit in one of the other sections

Re: The hardest sudokus (new thread)

Postby champagne » Mon Aug 23, 2010 5:11 pm

ronk wrote:
champagne wrote:To-day, you get the final result in about 4 hours (on my computer).

Since I'm an engineer, statements like that make my hair stand up. :( Everyone, is it too much to ask that execution times be normalized to ... 'T' Ghz-sec? Then the rest of us could estimate execution times on our computers by dividing 'T' by our CPU clock rate.


*You are surely right, but I guess most of us have a computer with a clock around 3MGH, so the order of magnitude should not change that much.

To be as precise as possible, I have a portable PC under vista with a double core and a clock of 2.4 GHZ (32 bits registers)

The rating was run with others programs, so the overall time could be influenced by other tasks.

Anyway, the 4 hours is a rough estimate. Could be 3 (small chances), could be 5 as well.

champagne
champagne
2017 Supporter
 
Posts: 5675
Joined: 02 August 2007
Location: France Brittany

Re: The hardest sudokus (new thread)

Postby m_b_metcalf » Tue Aug 24, 2010 6:52 am

champagne wrote:
m_b_metcalf wrote: It was 'only' 11.4 back then. I peaked early.

This puzzle is likely the highest known rating for SE.

Gosh, if true, that would make me the reigning World SE Champion! :D
BTW, I tried to find a start for the pattern game using that pattern, but I failed.

I tried that too.

I went back to the original pattern, as an exercise using my current s/w, rather than the antiques back then, and found a diamond and a few others you might like for your collection.
Code: Select all
100000002030400050006000700080203000000070000000508040007000600050009080200000001   ED=10.4/10.4/10.4
100000002030400050006000700040603000000020000000508090007000100080009030200000006   ED=10.5/10.5/10.4
100000002030400050006000700040803000000010000000506090007000600080009030200000001   ED=10.6/10.6/10.4
100000002030400050006000700040603000000020000000508090007000600080009030200000001   ED=10.7/10.7/10.4

Of course, if JPF's brilliant pattern were presented in a Patterns Game today, the 11.8 would never be found and investigated.

Regards,

Mike Metcalf
User avatar
m_b_metcalf
2017 Supporter
 
Posts: 8335
Joined: 15 May 2006
Location: Berlin

Re: The hardest sudokus (new thread)

Postby champagne » Tue Aug 24, 2010 7:59 am

Hi mike,

I did not work on that pattern specifically to find high rating, but looking for a game start, I got also that one


Code: Select all
9.......5.7...4.1...3...8.....2.7.4.....3.....1.6.......8...3...6.1...7.5.......9 ED=10.8/10.8/3.4


I will check if some of your 4 are in the general search I am doing, but I am not ready for a quick answer.

As far as I can see, nearly all puzzles generated with that pattern have the SK loop. All in your lot have it.

champagne
champagne
2017 Supporter
 
Posts: 5675
Joined: 02 August 2007
Location: France Brittany

Re: The hardest sudokus (new thread)

Postby coloin » Tue Aug 24, 2010 9:57 am

Code: Select all
Rating: 11.9  (ER/EP/ED=11.9/11.9/11.3)
Poster: tarek
Label: golden nugget


They all have the SK loop because the 16-clue base [without the central box with 4 clues] - inherently tends to have a low solution count - because some eliminations can be made with just those 16 clues.....

Puzzles which dont have a relatively low solution count with a selection of 16-clues - dont have the SK loop.

But if you dont have the low solution count - there are fewer ways to add a few clues to make a puzzle.

C
coloin
 
Posts: 1637
Joined: 05 May 2005

Re: The hardest sudokus (new thread)

Postby champagne » Tue Aug 24, 2010 11:19 am

coloin wrote:
Code: Select all
Rating: 11.9  (ER/EP/ED=11.9/11.9/11.3)
Poster: tarek
Label: golden nugget


They all have the SK loop because the 16-clue base [without the central box with 4 clues] - inherently tends to have a low solution count - because some eliminations can be made with just those 16 clues.....

Puzzles which dont have a relatively low solution count with a selection of 16-clues - dont have the SK loop.

But if you dont have the low solution count - there are fewer ways to add a few clues to make a puzzle.

C


I agree on Golden Nugget rating. I had it, but I did not keep in ming the results, may-be because I was impressed by the "relative poor rating" for Silver Plate : ED=11.5/11.5/10.1

Regarding the SK loop, I have another way to look at it.

All examples of the SK loop I analyzed have a common piece of pattern :

- four given forming a rectangle in four boxes
- all cells in rows/columns/boxes for this rectangle unknown.

If you have that piece of pattern, then, you have a reasonable chance to find a SK loop.

champagne
champagne
2017 Supporter
 
Posts: 5675
Joined: 02 August 2007
Location: France Brittany

Re: The hardest sudokus (new thread)

Postby m_b_metcalf » Tue Aug 24, 2010 11:24 am

coloin wrote:
Code: Select all
Rating: 11.9  (ER/EP/ED=11.9/11.9/11.3)
Poster: tarek
Label: golden nugget


So, tarek has the Title!

Congrats.

Mike Metcalf
Last edited by m_b_metcalf on Tue Aug 24, 2010 11:33 am, edited 1 time in total.
User avatar
m_b_metcalf
2017 Supporter
 
Posts: 8335
Joined: 15 May 2006
Location: Berlin

Postby Pat » Tue Aug 24, 2010 11:31 am

champagne wrote:I did not work on that pattern specifically to find high rating, but looking for a game start, I got also --
Code: Select all
9.......5.7...4.1...3...8.....2.7.4.....3.....1.6.......8...3...6.1...7.5.......9 ED=10.8/10.8/3.4
As far as I can see, nearly all puzzles generated with that pattern have the SK loop.

    looking for a game start ?
    so now, the challenge is to find an easy puzzle with this pattern ?
User avatar
Pat
 
Posts: 3438
Joined: 18 July 2005

Re: The hardest sudokus (new thread)

Postby m_b_metcalf » Tue Aug 24, 2010 2:15 pm

champagne wrote:... looking for a game start, I got also that one
Code: Select all
9.......5.7...4.1...3...8.....2.7.4.....3.....1.6.......8...3...6.1...7.5.......9 ED=10.8/10.8/3.4


And there are these too:
Code: Select all
1.......2.3...4.5...6...7.....8.3.4.....6.....5.9.......7...6...8.5...3.2.......1   ED=10.7/10.7/3.4
1.......2.3...4.5...6...7.....5.3.4.....1.....8.9.......7...6...9.8...3.2.......1   ED=11.6/11.6/3.4


Regards,

Mike Metcalf
User avatar
m_b_metcalf
2017 Supporter
 
Posts: 8335
Joined: 15 May 2006
Location: Berlin

Re: The hardest sudokus (new thread)

Postby ab » Sun Aug 29, 2010 12:25 pm

The Guardian published this puzzle this week:
http://www.guardian.co.uk/media/2010/au ... ed-picture
I don't know if it's one of yours, or how it compares!
ab
 
Posts: 451
Joined: 06 September 2005

Re: The hardest sudokus (new thread)

Postby champagne » Sun Aug 29, 2010 1:16 pm

ab wrote:The Guardian published this puzzle this week:
http://www.guardian.co.uk/media/2010/au ... ed-picture
I don't know if it's one of yours, or how it compares!


1) that puzzle was not in my data base.

2) In a pattern game (that could not happen due to the fact that the puzzle has not the symmetry requirements), i would likely have rated it.

3) The SE rating for that puzzle is 10.6 1.2 1.2 which is not bad

4) it has no special pattern (SK loop, EXOCET) and I found no rank0 logic

5) seen by my solver
it enters level 3 of tagging
both run time and print size put it limit down to enter the family of hardest.
The most difficult step uses 3 cycles of derived links

Surely an interesting puzzle for skill players as "ttt" or stevek

champagne
champagne
2017 Supporter
 
Posts: 5675
Joined: 02 August 2007
Location: France Brittany

Re: The hardest sudokus (new thread)

Postby ronk » Sun Aug 29, 2010 1:36 pm

ab wrote:The Guardian published this puzzle this week:
http://www.guardian.co.uk/media/2010/au ... ed-picture
I don't know if it's one of yours, or how it compares!

Not in my database either, so it's probably not a puzzle previously known to this forum.
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Re: The hardest sudokus (new thread)

Postby gsf » Sun Aug 29, 2010 2:34 pm

ab wrote:The Guardian published this puzzle this week:
http://www.guardian.co.uk/media/2010/au ... ed-picture
I don't know if it's one of yours, or how it compares!

google Arto Inkala hardest sudoku
looks like Arto has a 15-minutes-of-fame generator
and the media bites every time
gsf
2014 Supporter
 
Posts: 7306
Joined: 21 September 2005
Location: NJ USA

Re: The hardest sudokus (new thread)

Postby 999_Springs » Mon Aug 30, 2010 7:39 pm

Strange how Arto Inkala's puzzles get a lot of attention from the media whereas none of the other puzzles in this thread do. I wonder how much the media paid for it.
Hidden Text: Show
Once upon a time I was a teenager who was active on here 2007-2011
999_Springs
 
Posts: 367
Joined: 27 January 2007
Location: In the toilet, flushing down springs, one by one.

Re: The hardest sudokus (new thread)

Postby champagne » Mon Aug 30, 2010 8:47 pm

999_Springs wrote:Strange how Arto Inkala's puzzles get a lot of attention from the media whereas none of the other puzzles in this thread do. I wonder how much the media paid for it.


most surprising is that somebody can claim to have found "the hardest puzzle" publishing something similar to what each of the pattern game players generates every week.

The "hardest puzzles" published in that forum have been found 2 or 3 years ago !!!!
They are about ten times harder than that one.


champagne
champagne
2017 Supporter
 
Posts: 5675
Joined: 02 August 2007
Location: France Brittany

Re:

Postby Mauricio » Wed Sep 15, 2010 3:00 am

Pat wrote:
champagne wrote:I did not work on that pattern specifically to find high rating, but looking for a game start, I got also --
Code: Select all
9.......5.7...4.1...3...8.....2.7.4.....3.....1.6.......8...3...6.1...7.5.......9 ED=10.8/10.8/3.4
As far as I can see, nearly all puzzles generated with that pattern have the SK loop.

    looking for a game start ?
    so now, the challenge is to find an easy puzzle with this pattern ?

Here are two possible starts, though it would not be an interesting game.
Code: Select all
1.......2.3...4.5...2...6.....6.2.4.....7.....5.8.......6...4...8.3...7.2.......1
1.......2.3...4.5...2...6.....6.7.8.....3.....9.2.......6...9...4.9...3.7.......1
Mauricio
 
Posts: 1174
Joined: 22 March 2006

PreviousNext

Return to General