The hardest sudokus

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

Postby tarek » Sun Sep 21, 2008 11:26 am

g.r.emlin is the best to do all of this ....

g.r.emlin can verify q1,q2,Sx9 & Sxt ... SE121 rating has to be left for the poster's honesty.

If g.r.emlin is not up for it (for whatever devilish reasons:D ) Then I'm ready to do the same job.

My suggestion:

*Top 10 of each category in both line & graphic format
*categories are q1,q2,Sx9,Sxt,SE121
* verify q1,q2,Sx9,Sxt but not SE121 (verify SE131 if possible)
* SE121 & other q1,q2,Sx9,Sxty have to be provided by poster or that entry is disqualified


there has to be a database of posted puzzles (an updated q2 taxonomy list is the best) ... failing that .... I would use my own database (if organising this falls to me, that include famous puzzle lists to date )

tarek
User avatar
tarek
 
Posts: 2622
Joined: 05 January 2006

Postby gsf » Sun Sep 21, 2008 12:35 pm

tarek wrote:g.r.emlin is the best to do all of this ....

g.r.emlin can verify q1,q2,Sx9 & Sxt ... SE121 rating has to be left for the poster's honesty.

If g.r.emlin is not up for it (for whatever devilish reasons:D ) Then I'm ready to do the same job.

My suggestion:

*Top 10 of each category in both line & graphic format
*categories are q1,q2,Sx9,Sxt,SE121
* verify q1,q2,Sx9,Sxt but not SE121 (verify SE131 if possible)
* SE121 & other q1,q2,Sx9,Sxty have to be provided by poster or that entry is disqualified


there has to be a database of posted puzzles (an updated q2 taxonomy list is the best) ... failing that .... I would use my own database (if organising this falls to me, that include famous puzzle lists to date )

sounds like a good start
might take 2 weeks to get the pieces together
gsf
2014 Supporter
 
Posts: 7306
Joined: 21 September 2005
Location: NJ USA

Postby tarek » Tue Sep 23, 2008 9:30 am

The following is modified from ravel's post on how to rate puzzle.

Puzzle.txt Should have all puzzles needed for rating

Sudoku.exe & Suexratt.exe should be downloaded

a single batch file can be used for rating the puzzles according to q1,q2,Sx9 & Sx9 as follows (windows):
Code: Select all
q1:          sudoku -q1 puzzles.txt >  q1_ratings.txt
q2:          sudoku -q2 puzzles.txt >  q2_ratings.txt
Sx9 & Sxt:   suexratt puzzles.txt 1000 2 > suex_ratings.txt
I tend to increase the number that follows suexratt to 10000 for more accuracy with puzzles that have extreme ratings.

Output Example for PB .......12........3..23..4....18....5.6..7.8.......9.....85.....9...4.5..47...6...:
q1: .......12........3..23..4....18....5.6..7.8.......9.....85.....9...4.5..47...6... # 99485 FNBP C21.m/M3.389.1366
q2: .......12........3..23..4....18....5.6..7.8.......9.....85.....9...4.5..47...6... # 99551 FNBP C21.m/M3.389.1366
Sx9,Sxt (1000 2): rating: 5154 , 2102 , .......12........3..23..4....18....5.6..7.8.......9.....85.....9...4.5..47...6...
Sx9,Sxt (10000 2): rating: 5105 , 2105 , .......12........3..23..4....18....5.6..7.8.......9.....85.....9...4.5..47...6...

tarek
User avatar
tarek
 
Posts: 2622
Joined: 05 January 2006

Postby coloin » Wed Sep 24, 2008 12:02 pm

To assist others in making hard puzzles perhaps a simple command line for gsfs program woulkd be useful

Making puzzles using a {-2+2}
sudoku -go{-2+2} origpuzzles.txt >newpuzzles.txt

at 21 clues 75 puzzles can give 750000 other puzzles [overnight]

This could be initially quickly rated with sudoku -qFN [set to print minimal puzzles with rating over 90000 .....[?command for this]

Isomorph puzzles from a file of previously found puzzles need to be purged and removed at this stage. [command for this ?]

I use boxer text editor to remove #ratings [to be accepted by sxratt] and paste the q2 and sxratt ratings into excel.

I actually use havards rather neat "sudoku architect" software for the making of minimal puzzles and purging for known isomorphs - I will try to contact him to see if there is a general release of this software. [since this is what it was made for !]

C
coloin
 
Posts: 1637
Joined: 05 May 2005

Postby vrencianz » Fri Sep 26, 2008 9:38 am

I have a generator program here http://code.google.com/p/dance-sudoku/ and generated few hard ones (rating 11, few of yours have 14)

..7.2.3......14...1.5........629..73....6....92..786........9.1...68......2.5.7..
..2.....8...7..32.8....9...72.6.....6..247..1.....1.76...1....7.89..6...2.....5..

I am interested about your opinion about these puzzles.
vrencianz
 
Posts: 2
Joined: 26 September 2008

Postby champagne » Fri Sep 26, 2008 10:25 am

vrencianz wrote:I have a generator program here [url]
I am interested about your opinion about these puzzles.


unhappily, no good news

the first one is good for beginners, the second one could be proposed to skilled players.

Both are at thousand miles of the target.

champagne
champagne
2017 Supporter
 
Posts: 5678
Joined: 02 August 2007
Location: France Brittany

Postby StrmCkr » Fri Sep 26, 2008 8:51 pm

Code: Select all
rating 11, few of yours have 14


how are you rating them compared to ours? to say few are 14s?

reading your page
are you analizing vers a brute force algorithem of Dancing links (algorithm x.) ? and justifing ease with that.

vrs a constraint logic based solver.
which is where our ratings are based.

we use a colage of programs Soduko Explainer, GSF, and others to list a few. SE is a fairly skilled at rating program given what is currently avialable.

both the puzzles have reletivly no hard steps required to solve?
which makes them not difficult or rated with a high SE index.

they are no where near the scale or caliber to the puzzles many of us are displaying.
Some do, some teach, the rest look it up.
User avatar
StrmCkr
 
Posts: 647
Joined: 05 September 2006

Postby vrencianz » Mon Sep 29, 2008 9:19 am

Thank you for your answers and tips.

(I realized that this isn't the right forum topic to ask these kind of questions. Next time I'll find a better one.)
vrencianz
 
Posts: 2
Joined: 26 September 2008

Postby Obi-Wahn » Mon Oct 06, 2008 2:51 pm

This weekend I felt like looking into Sudokus again and checked coloin's latest list.
There were a bunch of toughies that scored 1 Jelly but colx006 was outstanding with 6 Jellies and 19 Swords, which gets it the fourth place in my list behind the three Star scoring puzzles.

Code: Select all
colx006
1.......6.5.7...8...3...4.....5.8.9.....3.....8.92....6.....3...7...5.2...4.....1
SE121 = 11.8

IIRC this is the second highest SE121 rating so far, only one tenth short of the Golden Nugget.

By altering the central box I managed to generate another hard one with 5 Jellies and 14 Swords for fifth place.

Code: Select all
OW081006
1 . .|. . .|. . 6
. 5 .|7 . .|. 8 .
. . 3|. . .|4 . .
-----+-----+-----
. . .|2 . 8|. 9 .
. . .|. 1 .|. . .
. 8 .|9 5 .|. . .
-----+-----+-----
6 . .|. . .|3 . .
. 7 .|. . 5|. 2 .
. . 4|. . .|. . 1
SE121 = 11.7


Did I miss any other SE121 ratings between the Easter Monster's 11.6 and the Golden Nugget's 11.9?

[edit] Actually I did miss my own post, where I stated that coloin_04/13-1600 and coloin_04/13-1414 reached SE121 = 11.8

I found another 2 puzzles with multiple Jellies:
Code: Select all
OW081007
1 . .|. . .|. . 7
. 4 .|6 . .|. 8 .
. . 3|. . .|5 . .
-----+-----+-----
. 8 .|9 . .|. . .
. . .|. 1 6|. . .
. . .|4 . 2|. 9 .
-----+-----+-----
. . 5|. . .|3 . .
. 6 .|. . 4|. 2 .
7 . .|. . .|. . 1
2 Jellies & 11 Swords, SE121 = 11.4

OW081008
1 . .|. . .|. . 5
. 9 .|2 . .|. 4 .
. . 3|. . .|7 . .
-----+-----+-----
. 8 .|9 . 6|. . .
. . .|. 5 .|. . .
. . .|8 4 .|. 9 .
-----+-----+-----
. . 5|. . .|3 . .
. 6 .|. . 4|. 2 .
7 . .|. . .|. . 1
7 Jellies & 23 Swords! SE121 = 11.5


Searching through tarek's Pearly 6000 I found another one:
Code: Select all
tarek_Pearly_#3298
. . .|. . .|. . 4
. 1 .|. . 3|. 9 .
6 . .|. 9 .|8 . .
-----+-----+-----
. . .|3 . 5|. . .
. . 7|1 . .|. . .
. 3 .|. . 2|. 1 .
-----+-----+-----
. . 4|. . .|7 . .
. 9 .|5 . .|. 2 .
8 . .|. . .|. . 6
2 Jellies & 12 Swords, SE121 = 11.3


Greetz, Obi
User avatar
Obi-Wahn
 
Posts: 61
Joined: 05 January 2007
Location: Darmstadt, Germany

Postby tarek » Thu Oct 09, 2008 11:23 am

Thanks Obi-Wahn for your update ...

here are the puzzles from the patterns game (game 14 onwards) that had an ER >= 10.5 showing their ER/EP/ED

10.7/10.7/09.9 100002003020030040005600000003000002060010070800000900000004800070060050900500001 0040-079-Mauricio
10.7/10.7/09.5 100002030004100500060040001070000040005080006600000300020004900900500003006090070 0051-078-JPF
10.7/01.2/01.2 100200300000040050006003001700000800080060040009000006600300700050090000003008002 0020-205-JPF
10.6/10.6/10.6 100002003020030040005600000003000006060040070800000200000007900090020080600500001 0040-081-Mauricio
10.6/10.6/10.5 100200300000040050006007008300000700090060040002000009700900600050020000003008001 0020-192-JPF
10.6/10.6/10.1 100200300000030040005006007200000600080010090009000008800900200070060000004003005 0020-165-Mauricio
10.6/10.6/10.0 100002030004100500060070001020000080003090007500000300040007600600900003008060090 0051-060-JPF
10.6/10.6/09.8 100200300040050010006001007600800700080060090004002006400900100090020070005003009 0016-141-JPF
10.6/10.6/09.8 010000002304000000050060400000500001002070080000009600008004900000020070600300005 0037-129-Mauricio
10.6/10.6/07.8 100200300000040050004006002600000200070080090005000001200700400090030000008005009 0020-149-JPF
10.6/10.6/06.7 100200300000040050006007001600000100080050070009000008300400200060080000007003009 0020-137-m_b_metcalf
10.6/10.6/03.8 000100000002003004050060070800600500004001002060090040000900000007008003040070080 0022-088-JPF
10.6/10.6/02.6 001002003040000050600070100000803002009000800200706000007010008050000040800600900 0047-129-JPF
10.6/06.7/06.7 010000002304000000050060700000200001001050040000003600007008900000040050200900008 0037-096-gsf
10.6/06.7/03.4 001900000050007600200000030900300010000040000060008007080000004002100060000004500 0041-081-gsf
10.6/01.2/01.2 100002003040050010006700000003000004020080090900000100000004500050090020700800006 0040-039-JPF
10.6/01.2/01.2 001900000050007600200000030300200090000040000060008005080000004002600010000005700 0041-003-Pat
10.6/01.2/01.2 001002003040000050600070100000603008009000200800201000006030009050000040900700800 0047-057-JPF
10.5/10.5/10.5 100002003020040050006700000007000008080050060400000900000009400030080070200600001 0040-148-Mauricio
10.5/10.5/10.4 001002300040000090600700008005200007000000000300006200200003001090000040008500600 0025-089-gsf
10.5/10.5/10.3 100200300000040050006007008300000700070010040009000002200800600050020000008003009 0020-212-JPF
10.5/10.5/10.0 100002003020040050006700000005000008080090040400000700000005800090020060300600001 0040-072-Mauricio
10.5/10.5/09.5 100200300000040050006003007300000800010080090009000001400900200070030000005004006 0020-188-JPF
10.5/10.5/09.5 010000002304000000050060700000200001007080040000009500001004600000010080900500003 0037-107-gsf
10.5/10.5/09.5 001006003040000050200070100000703002003000600800609000008090007050000040100200800 0047-126-gsf
10.5/10.5/09.5 100002030002400500060070002070000050008060003900000800020009300600500004003040010 0051-068-JPF
10.5/10.5/09.4 100200300000010040005003001300000500060070080009000006200500900040060000008001007 0020-174-Mauricio
10.5/10.5/09.4 010000002304000000050060700000200008009030070000004500006005100000090030800100004 0037-062-Mauricio
10.5/10.5/09.4 100002003020040050006500000003000007080070020200000900000009700040080060500600001 0040-034-JPF
10.5/10.5/08.3 001002003040000050300060100000703008002000900800906000003070001050000040900200700 0047-088-Mauricio
10.5/10.4/06.7 100200300000040050005006007200000800030010020009000005700600400080030000004007009 0020-085-JPF
10.5/06.7/03.4 001900000050007600200000030900300010000040000060008005080000004002100060000004700 0041-080-gsf
10.5/01.2/01.2 010020300450000000006000007000300800200040050000007009300100040000050600007008000 0031-090-JPF
10.5/01.2/01.2 000001200030040050600700000100080700070409030009070004000004006050090040008200000 0034-106-JPF
10.5/01.2/01.2 001002003040000010500010600000703002003000500800506000005070009090000040700900800 0047-055-JPF

tarek
User avatar
tarek
 
Posts: 2622
Joined: 05 January 2006

from game 41

Postby Pat » Fri Oct 10, 2008 3:04 am

game 41 also had these puzzles posted by Mauricio ( he did not provide ratings ) --
    10.7-10.7-10.4 ..12......3...45..4......6.6..1...7.....8.....9...3..8.5......4..27...1......89..

    10.6-10.6-9.9 ..12......3...45..6......7.2..7...6.....8.....9...3..4.8......9..51...2......73..
User avatar
Pat
 
Posts: 3438
Joined: 18 July 2005

Postby Obi-Wahn » Fri Oct 10, 2008 11:29 am

There are no new puzzles requiring a Jelly check in these lists.

I planned to publish a list of all Jelly+ puzzles known to me so far (162). And I tried to build in a new feature into my solver. So far I counted the most complicated checks for the complete solution of a puzzle. Now, I thought I just count the Jelly and Sword checks till the next candidate elimination. This way I can solve a puzzle step by step and get the count for the most complicated candidate elimination, just like Sudoku Explainer. I hoped that this rating (most complicated step instead of the sum for the complete solution) would be better correlated to the SE rating.

When I applied this method to the Golden Nugget something weird happened. For the complete solution my solver only needs 1 Jelly and 20 Sword checks.
Code: Select all
.---------------------.---------------------.---------------------.
| 25678  14568  124567| 268    2467   4678  | 1247   3      9     |
| 26789  4689   2467  | 23689  23467  1     | 247    2467   5     |
| 2679   1469   3     | 269    5      4679  | 8      12467  1247  |
:---------------------+---------------------+---------------------:
| 235    345    8     | 135    9      357   | 123457 1247   6     |
| 3569   7      456   | 13568  136    2     | 13459  1489   1348  |
| 1      3569   256   | 4      367    35678 | 23579  2789   2378  |
:---------------------+---------------------+---------------------:
| 367    136    9     | 1236   8      346   | 12347  5      12347 |
| 3578   2      157   | 1359   134    3459  | 6      14789  13478 |
| 4      13568  156   | 7      1236   3569  | 1239   1289   1238  |
'---------------------'---------------------'---------------------'

For the first steps I got the following counts:
1) R7C4<>3, 2 Swords
2) R4C4<>3, 2 Swords
3) R2C1<>7, 1 Jelly, 4 Swords
4) R1C1<>8, 1 Jelly, 4 Swords
5) R7C9<>3, 1 Jelly, 5 Swords
6) R1C3<>6, 1 Jelly, 5 Swords
7) R1C3<>5, 1 Jelly, 6 Swords
8) R2C1<>2, 1 Jelly, 6 Swords
9) R3C2<>9, 1 Jelly, 6 Swords
10) R2C2<>6, 1 Jelly, 7 Swords
Code: Select all
.---------------------.---------------------.---------------------.
| 2567   14568  1247  | 268    2467   4678  | 1247   3      9     |
| 689    489    2467  | 23689  23467  1     | 247    2467   5     |
| 2679   146    3     | 269    5      4679  | 8      12467  1247  |
:---------------------+---------------------+---------------------:
| 235    345    8     | 15     9      357   | 123457 1247   6     |
| 3569   7      456   | 13568  136    2     | 13459  1489   1348  |
| 1      3569   256   | 4      367    35678 | 23579  2789   2378  |
:---------------------+---------------------+---------------------:
| 367    136    9     | 126    8      346   | 12347  5      1247  |
| 3578   2      157   | 1359   134    3459  | 6      14789  13478 |
| 4      13568  156   | 7      1236   3569  | 1239   1289   1238  |
'---------------------'---------------------'---------------------'

So far, so good. And now step 11:
R2C4<>9 and R3C1<>9, 3 Jellies, 12 Swords!

That doesn't make sense. How can a single step be more complicated than the complete solution? Okay, so I thought my step solving function was flawed. But when I started my original solver on the reduced candidate grid, it gave: 4 Jellies & 14 Swords for the complete solution.

So the question is now: how can a puzzle get more complicated for my solver by correctly removing candidates?
I'm quite puzzled ...

Greetz, Obi
User avatar
Obi-Wahn
 
Posts: 61
Joined: 05 January 2007
Location: Darmstadt, Germany

Postby coloin » Sat Oct 11, 2008 12:16 pm

Interesting that you have highlighted those new puzzles - which surprisingly arnt in the taxonomy list.

I wish you luck in clarifying your recent solver problems . Maybe this is of help.

Thanks for your analysis of the batch of puzzles.
Code: Select all
Cloudy Bay      # No Jelly, no Sword
Bronze Medalian # 5 Swords
Silver Plate    # 1 Jelly, 10 Swords
Platinum Blonde # 1 Sword
Tungston Rod    # No Jelly, no Sword
Golden Nugget   # 1 Jelly, 20 Swords
DukDiamond1     # 1 Sword
DukDiamond2     # 1 Sword
WeeKender2      # 1 Sword
WeeKender1      # 1 Sword
Fata Morgana    # 1 Sword
Trompe-l'oeil   # 2 Swords 

The posting of the other puzzle has highlighted an interesting observation however.

The making of these puzzles was derived by using a 16 clue template
- obtaining all the possible ways to complete the puzzle by filling in the central box
- then remove clues from the central box.

16 clue templates commonly had ~ 500,000 grid completions, where a random 16 clue might have 4 million or more.

Now particular 16 clue template that I have found is this one - a slight modification to colx006
Code: Select all
1.......5.6.2...8...3...7.........4...........8.......7.....3...4...6.2...5.....1

The middle clues [even numbers] in B1379 occur as a single clue in B2468
Code: Select all
+---+---+---+
|1..|...|..5|
|.6.|2..|.8.|
|..3|...|7..|
+---+---+---+
|...|.3.|.4.|
|...|.71|...|
|.8.|4..|...|
+---+---+---+
|7..|...|3..|
|.4.|..6|.2.|
|..5|...|..1|
+---+---+---+    Its easy to give an example of 20 clue SE 9.0/8.6 /1.2   
many of the puzzles with this template have a difficult first elimination
Here is the initial pms for this complete puzzle -not many bivalue positions or cells it would appear............
Code: Select all
+----------------------+----------------------+----------------------+
| 1      279    24789  | 36789  4689   34789  | 2469   369    5      |
| 459    6      479    | 2      1459   34579  | 149    8      349    |
| 24589  259    3      | 15689  145689 4589   | 7      169    2469   |
+----------------------+----------------------+----------------------+
| 2569   12579  12679  | 5689   3      2589   | 125689 4      26789  |
| 234569 2359   2469   | 5689   7      1      | 25689  3569   23689  |
| 23569  8      12679  | 4      2569   259    | 12569  135679 23679  |
+----------------------+----------------------+----------------------+
| 7      129    12689  | 1589   124589 24589  | 3      569    4689   |
| 389    4      189    | 135789 1589   6      | 589    2      789    |
| 23689  239    5      | 3789   2489   234789 | 4689   679    1      |
+----------------------+----------------------+----------------------+
 

These are the "initial" pms for the 16 clue template [no clues in central box]
Code: Select all
+---+---+---+
|1..|...|..5|
|.6.|2..|.8.|
|..3|...|7..|
+---+---+---+
|...|...|.4.|
|...|...|...|
|.8.|...|...|
+---+---+---+
|7..|...|3..|
|.4.|..6|.2.|
|..5|...|..1|
+---+---+---+

Code: Select all
+-------------------------------+-------------------------------+-------------------------------+
| 1         279       24789     | 346789    346789    34789     | 2469      369       5         |
| 459       6         479       | 2         134579    134579    | 149       8         349       |
| 24589     259       3         | 145689    145689    14589     | 7         169       2469      |
+-------------------------------+-------------------------------+-------------------------------+
| 23569     123579    12679     | 1356789   12356789  1235789   | 125689    4         236789    |
| 234569    123579    124679    | 13456789  123456789 12345789  | 125689    135679    236789    |
| 234569    8         124679    | 1345679   12345679  1234579   | 12569     135679    23679     |
+-------------------------------+-------------------------------+-------------------------------+
| 7         129       12689     | 14589     124589    124589    | 3         569       4689      |
| 389       4         189       | 135789    135789    6         | 589       2         789       |
| 23689     239       5         | 34789     234789    234789    | 4689      679       1         |
+-------------------------------+-------------------------------+-------------------------------+

Hovever this template has a very lowly 132274 grid solutions

This means there will be many more valid puzzles with a complete central box.

The reason for this is :

the "real" pms are this [* represents a "real" elimination]
Code: Select all
+-------------------------------+-------------------------------+-------------------------------+
| 1         279       *4*8*     | 346789    346789    34789     | 2*6*      3*9       5         |
| *59       6         *79       | 2         13*57*    13*57*    | 149       8         349       |
| *4*8*     259       3         | 145689    145689    14589     | 7         1*9       2*6*      |
+-------------------------------+-------------------------------+-------------------------------+
| 23569     1*357*    12679     | 1356789   12356789  1235789   | 125689    4         236789    |
| 234569    1*357*    124679    | 13456789  123456789 12345789  | 125689    135*7*    236789    |
| 234569    8         124679    | 1345679   12345679  1234579   | 12569     135*7*    23679     |
+-------------------------------+-------------------------------+-------------------------------+
| 7         1*9       *26**     | 14589     124589    124589    | 3         569       4*8*      |
| 389       4         189       | 1357**    1357**    6         | 5*9       2         7*9       |
| 2*6**     *39       5         | 34789     234789    234789    | 4*8*      679       1         |
+-------------------------------+-------------------------------+-------------------------------+   

This means that this template has got these reductions - and that is without any given clues in the central box!

Achieving any of these reductions even when solving a complete puzzle envolves a a hard SE >9.3 if attempted before simpler methods.

This will be a form of "SK loop" in the puzzles with this template.

Easter Monster has a template which has eliminations - without the central box.
Code: Select all
+-------------------------------+-------------------------------+-------------------------------+
| 1         478       34578     | 356789    356789    35678     | 3489      369       2         |
| 238       9         378       | 4         123678    123678    | 138       5         368       |
| 23458     248       6         | 123589    123589    12358     | 7         139       3489      |
+-------------------------------+-------------------------------+-------------------------------+
| 234689    5         134789    | 1236789   12346789  1234678   | 12389     123679    36789     |
| 234689    124678    134789    | 12356789  123456789 12345678  | 123589    123679    356789    |
| 23689     12678     13789     | 12356789  12356789  1235678   | 123589    4         356789    |
+-------------------------------+-------------------------------+-------------------------------+
| 7         148       14589     | 12358     123458    123458    | 6         239       3459      |
| 456       3         145       | 12567     124567    9         | 245       8         457       |
| 45689     468       2         | 35678     345678    345678    | 3459      379       1         |
+-------------------------------+-------------------------------+-------------------------------+


+-------------------------------+-------------------------------+-------------------------------+
| 1         478       345*8     | 356789    356789    35678     | 3489      369       2         |
| 238       9         378       | 4         12*67*    12*67*    | 138       5         368       |
| *3458     248       6         | 123589    123589    12358     | 7         139       3489      |
+-------------------------------+-------------------------------+-------------------------------+
| 234689    5         134789    | 1236789   12346789  1234678   | 12389     12*67*    36789     |
| 234689    12*67*    134789    | 12356789  123456789 12345678  | 123589    12*67*    356789    |
| 23689     1267*     13789     | 12356789  12356789  1235678   | 123589    4         356789    |
+-------------------------------+-------------------------------+-------------------------------+
| 7         148       *4589     | 12358     123458    123458    | 6         239       3459      |
| 456       3         145       | 12*67     12**67    9         | 245       8         457       |
| 45*89     468       2         | 35678     345678    345678    | 3459      379       1         |
+-------------------------------+-------------------------------+-------------------------------+


These are many of the initial elimination clues envolved in the EM SK loop which were commented on by Allan Barker..... here and here in 3D

Golden Nuggett however wouldn't /doesnt have these "real" eliminations revealed if/when the most populated box is cleared.

champagnes solver perhaps finds similar eliminations, and consequently is able to classify these SK loop puzzles [including 04/13-1600 and 04/13-1414] as slightly easier !

C
coloin
 
Posts: 1637
Joined: 05 May 2005

ER/EP/ED

Postby AllanTrojan » Sat Oct 11, 2008 7:45 pm

I just joined this forum today though I have been reading it for a little while.

Tonight I completed writing a Sudoku program in visual basic
and it seems to have no difficulty in solving the 'hardest' problems
in a few seconds.

I would be grateful if someone could explain to me
what the notations ER/EP/ED mean
and how they are calculated.

Presumably they are explained in this forum in some earlier postings.
A link to these would be appreciated.
AllanTrojan
 
Posts: 3
Joined: 11 October 2008

Postby StrmCkr » Sat Oct 11, 2008 7:57 pm

Code: Select all
Tonight I completed writing a Sudoku program in visual basic
and it seems to have no difficulty in solving the 'hardest' problems
in a few seconds.


and what is your solver based on?

they are hardest only becasue logic based or cover set elliminations based solvers find steps that are beyond most human solvers to allocate and complete. requiring emence amount of compution to show and prove any give ellimination befor progreess is made.

a brute force solver ie: "dancing links" etc will crack these in fractions of secods.

our ratings are only from solvers that dont rely on trial and error till completion.
Some do, some teach, the rest look it up.
User avatar
StrmCkr
 
Posts: 647
Joined: 05 September 2006

PreviousNext

Return to General