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 Dec 06, 2010 3:53 pm

Mauricio wrote:
champagne wrote:2) try without the given in excess
. either you solve it now and find the given in the solution, the puzzle was not minimal
. or you fail and then you look for something else.
...

Counterexample to point 2
Code: Select all
.....1..2..3.4..5..6.7..8....1....3..4..9.1..7..1....6..5.7.....8.6...9.2....3..4
Minimal, not symmetric, remove 1@r6c4 (now multisolution, but symmetric), solve it (assuming uniqueness, wrongly), and you have r6c4=1 in the solution.

Now, if you remove one clue, giving it symmetry, solve it, and if the solution is not consistent with the original puzzle, then you now the clue you erased was not redundant; in other words, redundancy allows you to erase a clue, but IMO it is not easier to know if a clue is redundant than to solve the puzzle.


nice counter example, good to have the expert in symmetry.

is it a contradiction with point 2? I am not sure. and your reaction open some doors
Happily, in the wording, I did not stated that the puzzle was not minimal, so I have may be an escape lane.

Here it is a fact that the solution has a symmetry of given.
No easy way to establish it "forward". It could be that easy moves at the start lead to the symmetry, but in all these puzzles, the first move is very tough to establish.

Trying a kind of "backward" symmetry is still valid.

As you say, if the given in excess does not come, you know you could not erase it and you have to find something else..

If the erased given comes, as the puzzle is assumed to have a unique solution, it must be the solution, but it does not bring the probe that that digit was in excess.

Am I right??

champagne
champagne
2017 Supporter
 
Posts: 7453
Joined: 02 August 2007
Location: France Brittany

Re: The hardest sudokus (new thread)

Postby eleven » Wed Dec 08, 2010 10:23 am

Champagne,

i have made a new filter now. Can you find a hard one in this sample ?
Hidden Text: Show
.2.45....4....92....6.2...........383...9.5...7...3.1.5..9..3...8......6..1....7.
..3..6..9.5.1.....8...7.........436...4...9.....62...4.1.8.......6..2..37...6..5.
1...56.....7..9..6.6.7...4..4.6..3....5..7...9...1........7.82.......4...8.2....3
.2...6.894....9..3..97..5....4.7.9..8......2..3......4..1..76...4..........51....
...4...8...7..92......3...526...1.....19......7....1..5......4.71.8....3..6..29..
..34......5...9...6...2.1..2...6...1.4.5...6...6..3.....9...6.7......8.28...7..1.
.....6......18.2.66..32.....7.....5...9...4..8....2..33....1..8..5...9......4.37.
..3..67...5.1....66...2..4..8....9..3.18.......5.....8.....4.2..3.9....1....7.6..
1...5...9.56........92......1..6...7.....4.2....8..3....1.7...67......4..6...387.
.....678....1....6.6..7.4...7...1..4..5.2....9...3....3.....5...1...7..8..29...4.
...4.67....7....2.....7...5..6.4.5...9.6.1...8..3.....3..9...5..1......8..4.6.2..
..3..6.8....1..2.66...7...5..4......3....4.9..98.2.....3...8.6.8.....1.7...5.....
..3..6.8....1..2......7...4..9..8.6..3..4...1.7.2.....3....5..8..5...6..98.....5.
1...5...9......12...8...5...3.8.....5...6..9...4..7....7.6.4...6...2.9....53....6
1...5...9......12...8...5...3...7...8...6..9...48....2.7...4....8.3....66...2.9..
1..4.6....5.1.9.....6.2..1..8..4...5..9..2.4.7.....3..3.....8....42...6.........7
......7.94....9.3..8....4.5..18......6..2....7....53..5....7..3.7.6.......2.1..7.
..3.....94....92...9..7..1....56.9......1...8.6...7.5..1.7...9...2.....38.....4..
..3..6..9.5.1.....8...7.........43....4...92....62...4.1.8.......6..2..37...6..5.
...4...8...7..92......3...526...1.....19......7....1..5...9..4..9.8....3..6..29..
1...5...9..7...2...69..3.4....8..9.....91..7.9....5..83..5....1..2...6...4.......
.2.4....9..7...23.6.....5........69..9...4..2..5.9..7.3....8......2.1....1..4...8
1.......9..67...2..8...3.....76...4.36..1.....9..78.....5...2.....5..6.4.....7.5.
....56.8...71....6....27.....6..1..53......6..4.6..9....5.1...2.9....3..8......4.
...4.6..9..7.8..6......35..2...1....8......1...16.7.....2.6..9....9..3...9...5..4
1.....7.9.5.1......69....1.....4.8.7.....3..29..7...5.....2.4....4..8..36......7.
..3..6.8....1..2......7...4..9..8.6.....4.1...7.26......5.....383...5...96.....5.
1....67...571.......9....1..4....3.......89.29..7...6......24..5..6...9.....3...8
...4...8...7..92......3...52...7.1....19.....76...1...5......4...6..29...7.8....3
1.....7..4.7....36....7...42....8...5..9.......4.1.3....1.4.6...4...2.9.8..5.....
...4...8...7..92......3...52...7.1....19.....76...1...5...9..4...6..29.....8....3
....5.7..4...8...3.6.3...1.2....5.3..3.9..1......4.....9.7....1..2....7.8.....69.
...4...8...7..92......3...526...1.9...19.....7.....1..5...9..4...6..29.....8....3
...45....4....92....6.2......9.....83...9.5...75..3.1.5..9..3...8......6..1....7.
..34...8..5......6.....71..24.....9.3..9.......9.4.8....28...4.....6...7.....15.8
..34...8..5......6.....71..24.....9.3..9.......9.4.8....28...4.....64..7.....15..
..34......5...9...6...2...12...1...8.4.5..1....1..3.9...2.....77.....61.....7.8..
..34......5...92..6.9.7.....15....98.....1.2....5..1...9...58....4.6.......3...7.
..34......5...92..6...7.....15.....8.....1.259..5..1...9...58....4.6.......3...7.
..345.....5...92..6...7.....15.....8.....1.259.....1...9...58....4.6.......3...7.
.2.....8..5.1....6..9..71..2....3.......41.....47..3..6..9...2...2.1.9...8......5
.2..5..8......9..17.91.....2..5.8.6..6....8....8.6.4..3.............73...4..2..5.
.2.4....9..7.8..3.6...2.5....69.8....1..4...8.....1...3.....6...6...4..2..5....7.
.2.4...8.4.......68.9..7...2..9...7..7..1.6.......3..5......3....27...4..4...5..1
1....6..9.5.1...3...8...1....1...8...3.....5.9....4..2..2..7.9....24...77...6....
1....67.....18...6.6.73.....1...3..8...5.....5.4....2..3...1..7....9.4..9.2......
1...5...9.56........92......1..6...7.....462.6..8..3....1.7...67......4......38..
1...5...9.56........92......1..6...7.....4.2.6..8..3....1.7...67......4......387.
1...5...9......12...8...5....73....6.8...4..16...2.9...3...7...8...6..9...48.....
1...5..8...7.....6.....34..2..9.1........5.2...5.2.8...4...8..7..6...3..8...9..5.
1...5..8..5...9..6..9.2.5...7.3.......4..23.....54.....6......1..52..9..8......7.
1...5......7..9..6.6.7...4..4.6..3....5..7...9...1........7.82......84...8.2....3
1...5.7....6..9....8......42....3..1....2..37...5..2....46.....5...1.3...9...8..5
1...56.....6...1...8.2....4...3..84..3...49......2...3....4..7.5...7.....4.9....8
1...5...9......12...8...56.2...6..9..3.8.2..1..4..7....7...4...6...2.9.....3.....
1...5..8...7.....6.....34..2..9.1........5.2...5.2.8...4......7..6...3.88...9..5.
..345....4....9..2...1......1..7..9...539....9....42...7......4......6.86....8.2.
.2.4...8.....8...68....71.42..5...9..95.......4..3.........1..7..28...4.....6.3..
.2...6.8.4..18......972..........5...7.6...2.9.......35....7..4...2..3...1..6..7.
1...5...9..7...2...6.....4....83.......9...7.9....5..83..5....1..2..16...41.9....
.....67...5.1.......9.3...4...7..41......56....8.4...3..2...34.8...2....93......2
...45.7....7.8.........7.1.2......9.31......6..5..41...6.2....1..1..85..9......3.
.2...67.........32..9....6...854.....4...3.7.9...1.4....189....5.........9...4..7
1.......9..67...2..8...3..4..76...4..3..1....96...8.....5...2.7.......5....5..4.6
1...5...9.5.....3...9.......1..6...7.....4.2.6..8..3........87.7..3.2.4...1.7...3
.2..5....4....92...89......2..5...633...1...7..4..38...4....3.....6...15....7....
1...5...9.56........92......1..6...7.....4.2....8..3....1.7...67......4.9....387.
1..4..7....7..9..6.8.....5.....1....83..9......5..7.6.3.......5....4..92..2..56..
..3.5.7..4....9....8.2..1..2.....31...6.1...7......65..9.8...6......4.....5.3..7.
.2....7..4....9..3....2.4....1...6..8....3..4...67..1..8...5...5...3..4...98....5
.2...6.8...71.....6.9.3......1.9.......7..3...6...2.5......4..2..6....4..42...8.5
1...5...9......12...8...5..2...6..9..3.8.......4..7....7.6.4.1.6...2.9.....3....6
1...5...9......12...8...5..2...6..9..3.8.2.....4..7....7.6.4...6...2.9.....3....6
..3.5.7..4.7..9....8.7....1.7.8...1.8.5.4....9.........6.2..1.......3..8.......62
..3.5.7..45...9....8.2....52....4..3..5.3..7....8......9....61...6.1...7.......5.
..34...8..5......6.....71..24.....9.3..9.......9...8....28...4.....64..7.....15.8
.2....7.94.......6..93...1...5.3........98.5....5....1.7..4......18...9.6....3..2
1.......9..67...2..8...3.6...76...4.36...1....9...8...6....7.5....5..2....5.....4
..3.5...94....9..2...1......1..7..9...539....9....42...7......4......6.86....8.2.
....5.7..4...8...3.6.3...1.2....5....3.9..1......43....9.7....1..2....7.8.....69.
..34..7...5..8...6.....3....1...4...5.6.1....8..6....1...2..9...6..4...89....7.2.
....56.8...71....6....2.4...9..7....3..5..8....5..1..2.4.7.......6.1...59.....3..
....5....4....92...89....1.2..5.......4..29...3..6..7.3..7...2..4...83.....6....1
...4...8...7..92......3...526...1.....19.....7...2.1..5...9..4...6..29.....8....3
...4...8...7..92......3...526...1.....19.....7...2.1..5...9..4...6..29.....8....3
...4.6..9..7.8..6......35..26..1....8......1...1..7.....2.6..9....9..3...9...5..4
..3..6.8....1..2......7...4..9..8.6.3...4.1...7.2.......58......6.....5.93...58..
..3.5...94....92..8..2............7...1..5..2.6....8...7..12.4.....3.6....25....1
..34......5...9...6...2..1.2...6...7..4.......9...36.........788...7.1...7.5.8..2
.2...67..4.6.....37.8....6.2....84......9..1..8.5.......2..78.....1...9.....3...5
.2...6...4...8.2...89.....1...6....5..4.9.3.......1.7.3.......7.4..3.9....2..5.4.
.2.4....9..7.8.23.6...2.5........6...9...4..2..5....7.3..9........2.1....1..4...8
.2.......4....9..6..9.7.1....17..5..3....5.91.6.....2...8..79.....8....4.3..1....
.2.4...8...7..9.......3...52..6...4...6.....1.4...3.....28...7.7....21..8....59..
1......8..5......3..9..74....4.7.9..8......5..3..4...7..27..6......64.......92..1
1....67....7.8...6.6.3........9......8...7..19...3.5..5...4..2..1...3..7..2...4..
1....67...5..8..3...92...6....5...1..8..9....6....3..73....4.7...2...4........3.1
1....67...57.8..3........4.........3.1...56....6.4..2...1..79..5..9....2.9.8.....
1...5...9..7...2...6.....4....83.9.....9...7.9....5..83..5....1..2..16...4..9....
1...5...9.5.........92......1..6...7.....462.6..8..3....1.7...67......4..6...38..
eleven
 
Posts: 3142
Joined: 10 February 2008

Re: The hardest sudokus (new thread)

Postby champagne » Wed Dec 08, 2010 12:13 pm

This is a very homogeneous lot. Very close as well of the primary selection I am doing.

I just erased a small number of puzzles I would not have taken.

Most of these puzzles have an EXOCET or a "Quasi EXOCET" pattern.
All are solved using the 2 cells AC2, not more

Sometimes the EXOCET pattern works well, sometimes not, so about 20 of them are still very hard, even using the EXOCET property.

Difficult to predict the SE rating, but from my experience in the pattern game, you have the best chance to find high ratings in the top of the below list.

I would bet for the best ones in the range 10.5 11.5 and not many below 10.0 , but this is in the field I explored. You came with surprising results.

champagne

I think your hardest is in that short list

Hidden Text: Show
100050009000000120008000500200060090030802000004007000070604000600020900000300006
100050009050000000009200000010060007000004620600800300001070006700000040060003800
000400080007009200000030005260001000001900000070000100500090040090800003006002900
100000009006700020080003000007600040360010000090078000005000200000500604000007050
020400009007000230600000500000000690090004002005090070300008000000201000010040008
020006080007100000609030000001090000000700300060002050000004002006000040042000805
100050009000000120008000500030007000800060090004800002070004000080300006600020900
100050009007000200069003040000800900000910070900005008300500001002000600040000000
000406700007000020000070005006040500090601000800300000300900050010000008004060200
020000709400000006009300010005030000000098050000500001070040000001800090600003002


and here is a near complete list sorted to have in the top the highest SE ratings

Hidden Text: Show
100050009000000120008000500200060090030802000004007000070604000600020900000300006
020006080007100000609030000001090000000700300060002050000004002006000040042000805
020400080400000006809007000200900070070010600000003005000000300002700040040005001
100056000007009006060700040040600300005007000900010000000070820000000400080200003
100000009006700020080003000007600040360010000090078000005000200000500604000007050
100050009000000120008000500030007000800060090004800002070004000080300006600020900
000450700007080000000007010200000090310000006005004100060200001001008500900000030
020050000400009200089000000200500063300010007004003800040000300000600015000070000
100000700407000036000070004200008000500900000004010300001040600040002090800500000
003050009400009002000100000010070090005390000900004200070000004000000608600008020
000400080007009200000030005260001000001900000070000100500090040090800003006002900
020400009007000230600000500000000690090004002005090070300008000000201000010040008
100050009050000030009000000010060007000004020600800300000000870700302040001070003
100050080007000006000003400200901000000005020005020800040000007006000308800090050
100050009000000120008000500030800000500060090004007000070604000600020900005300006
100050080007000006000003400200901000000005020005020800040008007006000300800090050
020450000400009200006020000000000038300090500070003010500900300080000006001000070
003006700050100006600020040080000900301800000005000008000004020030900001000070600
003006009050100000800070000000004360004000900000620004010800000006002003700060050
100050009050000000009200000010060007000004620600800300001070006700000040060003800
020000709400000006009300010005030000000098050000500001070040000001800090600003002
003006080000100206600070005004000000300004090098020000030008060800000107000500000
003400000050009200600070000015000008000001025900500100090005800004060000000300070
003450000050009200600070000015000008000001025900000100090005800004060000000300070
003400000050009200609070000015000098000001020000500100090005800004060000000300070
000006700050100000009030004000700410000005600008040003002000340800020000930000002
100050009056000000009200000010060007000004020600800300001070006700000040000003870
100050009056000000009200000010060007000004620600800300001070006700000040000003800
100000080050000003009007400004070900800000050030040007002700600000064000000092001
100050009056000000009200000010060007000004020000800300001070006700000040060003870
000400080007009200000030005260001090001900000700000100500090040006002900000800003
003006009050100000800070000000004300004000920000620004010800000006002003700060050
020400080007009000000030005200600040006000001040003000002800070700002100800005900
020000700400009003000020400001000600800003004000670010080005000500030040009800005
100050080050009006009020500070300000004002300000540000060000001005200900800000070
100050009007000200060000040000830000000900070900005008300500001002001600041090000
000400080007009200000030005260001000001900000700020100500090040006002900000800003
000400080007009200000030005260001000001900000700020100500090040006002900000800003
003400000050009000600020001200010008040500100001003090002000007700000610000070800
020006089400009003009700500004070900800000020030000004001007600040000000000510000
100000009006700020080003060007600040360001000090008000600007050000500200005000004
003450000400009002000100000010070090005390000900004200070000004000000608600008020
100006700000180006060730000010003008000500000504000020030001007000090400902000000
000400080007009200000030005260001000001900000070000100500000040710800003006002900
020006700000000032009000060008540000040003070900010400001890000500000000090004007
100006700057100000009000010040000300000008902900700060000002400500600090000030008
100050009000000120008000500007300006080004001600020900030007000800060090004800000
100050009056000000009200000010060007000004020000800300001070006700000040900003870
003050700400009000080200100200000310006010007000000650090800060000004000005030070
100050000007009006060700040040600300005007000900010000000070820000008400080200003
100006009050100030008000100001000800030000050900004002002007090000240007700060000
000000709400009030080000405001800000060020000700005300500007003070600000002010070
100050009007000200060000040000830900000900070900005008300500001002001600040090000
100006700057080030000000040000000003010005600006040020001007900500900002090800000
100050009007000200069003040000800900000910070900005008300500001002000600040000000
000406700007000020000070005006040500090601000800300000300900050010000008004060200
003006080000100200000070004009008060030040001070200000300005008005000600980000050
003400000050009000600020010200060007004000000090003600000000078800070100070508002
020000000400009006009070100001700500300005091060000020008007900000800004030010000
003050700407009000080700001070800010805040000900000000060200100000003008000000062
003000009400009200090070010000560900000010008060007050010700090002000003800000400
003400080050000006000007100240000090300900000009040800002800040000064007000001500
003400080050000006000007100240000090300900000009040800002800040000060007000001508
003050009400009200800200000000000070001005002060000800070012040000030600002500001
100050700006009000080000004200003001000020037000500200004600000500010300090008005
003006080000100200000070004009008060300040100070200000005800000060000050930005800
003050700450009000080200005200004003005030070000800000090000610006010007000000050
000006780000100006060070400070001004005020000900030000300000500010007008002900040
020006080400180000009720000000000500070600020900000003500007004000200300010060070
000406009007080060000003500260010000800000010001007000002060090000900300090005004
100006700007080006060300000000900000080007001900030500500040020010003007002000400
000056080007100006000020400090070000300500800005001002040700000006010005900000300
100406000050109000006020010080040005009002040700000300300000800004200060000000007
003006080000100200000070004009008060000040100070260000005000003830005000960000050
003400700050080006000003000010004000506010000800600001000200900060040008900007020
100000709050100000069000010000040807000003002900700050000020400004008003600000070
000056080007100006000027000006001005300000060040600900005010002090000300800000040
000406009007080060000003500200010000800000010001607000002060090000900300090005004
100006700050080030009200060000500010080090000600003007300004070002000400000000301
020400080000080006800007104200500090095000000040030000000001007002800040000060300
003400080050000006000007100240000090300900000009000800002800040000064007000001508
000450000400009200006020000009000008300090500075003010500900300080000006001000070
000050700400080003060300010200005000030900100000043000090700001002000070800000690
000006000000180206600320000070000050009000400800002003300001008005000900000040370
020000080050100006009007100200003000000041000004700300600900020002010900080000005
100050009000000120008000500200060090030800000004007000070604010600020900000300006
100050009000000120008000560200060090030802001004007000070004000600020900000300000
020050080000009001709100000200508060060000800008060400300000000000007300040020050
000400080007009200000030005200070100001900000760001000500090040006002900000800003
100000009006700020080003004007600040030010000960008000005000207000000050000500406
003400000050009000600020100200060001040500060006003000009000607000000802800070010
100056000006000100080200004000300840030004900000020003000040070500070000040900008
100400700007009006080000050000010000830090000005007060300000005000040092002005600
020400009007080230600020500000000600090004002005000070300900000000201000010040008
020400009007080030600020500006908000010040008000001000300000600060004002005000070
020006700406000003708000060200008400000090010080500000002007800000100090000030005
020006000400080200089000001000600005004090300000001070300000007040030900002005040
champagne
2017 Supporter
 
Posts: 7453
Joined: 02 August 2007
Location: France Brittany

Re: The hardest sudokus (new thread)

Postby eleven » Wed Dec 08, 2010 1:32 pm

Thanks,

the puzzles all have ER >= 11.2 and q2 > 98000, i applied the new filter additionally.
But i know better now, how to select puzzles for you.
eleven
 
Posts: 3142
Joined: 10 February 2008

Re: The hardest sudokus (new thread)

Postby tarek » Wed Dec 08, 2010 1:42 pm

eleven wrote:The puzzles all have ER >= 11.2 and q2 > 98000, i applied the new filter additionally.
But i know better now, how to select puzzles for you.

I have always been testing minimal puzzles for difficulty ... Your search for a "Hardest" non minimal puzzle would be exploring uncharted waters.

tarek
User avatar
tarek
 
Posts: 3762
Joined: 05 January 2006

Re: The hardest sudokus (new thread)

Postby eleven » Wed Dec 08, 2010 2:39 pm

I dont search for non minimal hardest puzzles, i just dont test for minimality in the moment :)
This can be done later easily, it will not change the basic ER rating in almost all cases, but lower the other two.
eleven
 
Posts: 3142
Joined: 10 February 2008

Re: The hardest sudokus (new thread)

Postby champagne » Wed Dec 08, 2010 3:02 pm

hi eleven,

From my experience, the right cut off to catch all "hardest puzzles" should be lower than 11.2 my last "hardest" had only a rating 10.8.
10.6 would be I think a good cut off.

champagne
champagne
2017 Supporter
 
Posts: 7453
Joined: 02 August 2007
Location: France Brittany

Re: The hardest sudokus (new thread)

Postby tarek » Wed Dec 08, 2010 3:22 pm

champagne wrote:From my experience, the right cut off to catch all "hardest puzzles" should be lower than 11.2 my last "hardest" had only a rating 10.8.
10.6 would be I think a good cut off.

When we were attempting the generation of "Hardest puzzles" 3+ years ago. We used to hit a brick wall with SE at the 10.6 mark. That is why JPF at the time collated the posted puzzles with a 10.6+ rating.

This reminded me of that time in the ancient past when Ocean hit that 10.0 mark for the 1st time 8-) . How far back did the goal posts move :!:

tarek
User avatar
tarek
 
Posts: 3762
Joined: 05 January 2006

Re: The hardest sudokus (new thread)

Postby champagne » Wed Dec 08, 2010 3:58 pm

tarek wrote:
champagne wrote:From my experience, the right cut off to catch all "hardest puzzles" should be lower than 11.2 my last "hardest" had only a rating 10.8.
10.6 would be I think a good cut off.

When we were attempting the generation of "Hardest puzzles" 3+ years ago. We used to hit a brick wall with SE at the 10.6 mark. That is why JPF at the time collated the posted puzzles with a 10.6+ rating.

This reminded me of that time in the ancient past when Ocean hit that 10.0 mark for the 1st time 8-) . How far back did the goal posts move :!:

tarek



We are speaking here of a cut off to analyze puzzles, not for a final publication. But may be we should have that discussion in another thread.

Reversely, frankly I don't believe at all we can rely on SE rating alone to extract hardest puzzles.

I did not publish many puzzles up to now and I have in cache some puzzles that would be eligible with your 11.2 filter, but they have nothing attractive on my view.

(and I don't see a rush to enter new puzzles) :D :D

Anyway, I don't want to be a source of trouble and if you think it's better, we open a separate thread

champagne
champagne
2017 Supporter
 
Posts: 7453
Joined: 02 August 2007
Location: France Brittany

Re: The hardest sudokus (new thread)

Postby tarek » Wed Dec 08, 2010 4:52 pm

champagne wrote:We are speaking here of a cut off to analyze puzzles, not for a final publication. But may be we should have that discussion in another thread.
I was only confirming your suspicions through our attempts back in 2006/2007

champagne wrote:Reversely, frankly I don't believe at all we can rely on SE rating alone to extract hardest puzzles.
I think it was said before that SE ratings >=9.5 are difficult to judge because all available solvers will go to T&E at this stage. q1, q2, Sx9 & Sxt all have flaws ... There is a chance that we are excluding puzzles with SE>=9.5 from the hardest list.

champagne wrote:I did not publish many puzzles up to now and I have in cache some puzzles that would be eligible with your 11.2 filter, but they have nothing attractive on my view..
As far as I'm concerned, this thread is to discuss the "Hardest puzzles" ... Discussions, Posted puzzles & ratings should NOT be limited by the filters I'm employing.

If you suspect that you have the hardest puzzle in a group of puzzles then feel free to post them.

tarek
User avatar
tarek
 
Posts: 3762
Joined: 05 January 2006

Re: The hardest sudokus (new thread)

Postby eleven » Thu Dec 09, 2010 8:21 pm

Here is one for the q2 list :geek:
Code: Select all
99450 .2.4.....4....92...98.....4...6....75...1.....4...38..3......6....7....1..4..85.. Green Glasses


This grid only has 2 strong links (but a low 10.x ER)
Code: Select all
..3....8.4.........7..3...4...6..8...6.5.2.9...7.1...6.8.9..2..7...6...8.1.....5.
+-------------------------+-------------------------+-------------------------+
| 12569   259     3       | 1247    24579   145679  | 15679   8       12579   |
| 4       259     125689  | 1278    25789   156789  | 135679  12367   123579  |
| 125689  7       125689  | 128     3       15689   | 1569    126     4       |
+-------------------------+-------------------------+-------------------------+
| 12359   23459   12459   | 6       479     3479    | 8       12347   12357   |
| 138     6       148     | 5       478     2       | 1347    9       137     |
| 23589   23459   7       | 348     1       3489    | 345     234     6       |
+-------------------------+-------------------------+-------------------------+
| 356     8       456     | 9       457     13457   | 2       13467   137     |
| 7       23459   2459    | 1234    6       1345    | 1349    134     8       |
| 2369    1       2469    | 23478   2478    3478    | 34679   5       379     |
+-------------------------+-------------------------+-------------------------+


champagne wrote:All are solved using the 2 cells AC2, not more

Still i dont know, what AC2 really is and i did not find a definition in this forum. Can you point me to one please and demonstrate an AC2 move in this grid (#9 in my list), so that we can see, how complex it is ?

Code: Select all
 *----------------------------------------------------------------------*
 | 1       23478  23478  | 3467   5      678     | 2467   368    9      |
 | 2348    5      6      | 13479  13489  1789    | 1247   138    12348  |
 | 348     3478   9      | 2      1348   1678    | 14567  13568  13458  |
 |-----------------------+-----------------------+----------------------|
 | 234589  1      23458  | 359    6      259     | 459    589    7      |
 | 35689   3789   3578   | 13579  139    4       | 1569   2      158    |
 | 24569   2479   2457   | 8      129    12579   | 3      1569   145    |
 |-----------------------+-----------------------+----------------------|
 | 234589  23489  1      | 459    7      2589    | 259    359    6      |
 | 7       2389   2358   | 1569   1289   125689  | 1259   4      1235   |
 | 2459    6      245    | 1459   1249   3       | 8      7      125    |
 *----------------------------------------------------------------------*


I have uploaded 5000 puzzles to zippy share f3sudoku.zip. They all passed 3 filters, which throw out 90 % of the known hardest including swampy oil and Mauricio's puzzle, but probably dont have a very high ER. Maybe you can find something in this set.
eleven
 
Posts: 3142
Joined: 10 February 2008

Re: The hardest sudokus (new thread)

Postby tarek » Thu Dec 09, 2010 9:45 pm

eleven wrote:Here is one for the q2 list :geek:
Code: Select all
99450 .2.4.....4....92...98.....4...6....75...1.....4...38..3......6....7....1..4..85.. Green Glasses
Hi eleven,
Good catch ... Unfortunately I think it is a non minimal isomorph of Fata Morgana .... This will happen often when your search widens.

My tedious task will also check if the puzzles posted here have been posted before ... Your non minimal puzzles have to be minimized then checked :x
I have already a database with top lists of the past & submissions from coloin to check against but it will require updating with new puzzles submitted here.

tarek
User avatar
tarek
 
Posts: 3762
Joined: 05 January 2006

Re: The hardest sudokus (new thread)

Postby eleven » Thu Dec 09, 2010 11:34 pm

Wow, i really found an old one from random puzzles ? What a surpise !
eleven
 
Posts: 3142
Joined: 10 February 2008

Re: The hardest sudokus (new thread)

Postby champagne » Fri Dec 10, 2010 11:38 am

eleven wrote:
champagne wrote:All are solved using the 2 cells AC2, not more

Still i dont know, what AC2 really is and i did not find a definition in this forum. Can you point me to one please and demonstrate an AC2 move in this grid (#9 in my list), so that we can see, how complex it is ?


You should find a detailed answer either in that thread


full-tagging-t5624.html

or on my website, but let me answer briefly.

My main process is a search of AIC's and AIC's nets to eliminate candidates.

Code: Select all
level 1: stong links and weak links generated from rows/columns and boxes including groups.

level 2: AHS are added with new strong links and new weak links.

level 3: are added rowx/ columns or boxes patterns in which 2 digits are "unknown" to solve it
         each possible solution (a couple of digits in 2 groups) is a super-candidate.
         These super-candidates are eliminated in the same way as candidates.

         2 kinds of new patterns are considered
           2 cells 3 digits giving 3 super candidates 
           AC2 defined as {n cells n-2 known digits 4 free digits} and 6 "super candidates";
              the smallest AC2 is made of 2 cells 4 digits and no known digit (group)


In the last version of my solver (still bugged at the last level but I had no time to fix the bug), I defined a level 3.1 using only the smallest AC2.
. I thought and this is verified that most puzzles would be solved just using these ones
. They are much easier to see for a player.
. Last but no least, EXOCET patterns have such an AC2 as base, and the SK loop is solved just using them.

You can find examples of solutions using AC2 on my website;

champagne
champagne
2017 Supporter
 
Posts: 7453
Joined: 02 August 2007
Location: France Brittany

Re: The hardest sudokus (new thread)

Postby coloin » Fri Dec 10, 2010 9:32 pm

eleven wrote:Wow, i really found an old one from random puzzles ? What a surprise !


well a while back i was generating 21-clue puzzles and passing them on to champagne - and it was pleasing to see his solver and subsequent rater program advancing.

it seemed I was generating puzzles which just happened to give high ratings with -q2 and sxt/sx9
and these were the best that i got some time ago, just making the crieria.
Code: Select all
3149 ,    1115 , .......2........49.2.9..7.......8...5...1.6...3.4....7..1..7...87..5.3...4.3.....# 99166 FNBP C21.m/M3.625.850
3564 ,    2012 , ..3.......5.4...8.1.......7.9..8........94.6.5.62.............3.6.9...4...7.2.1..# 97003 FNBP C21.m/M2.5.22304
5008 ,    1212 , ...2....871......9..6.9.5....8.6...5.....23...4....6....9.3..5.4....1........7...# 96665 FNBP C21.m/M2.10.4592


It was also unclear which puzzles to pick to continue to advance with a {-2+2} process - pick the hard ones or perhaps the not so hard ones ???? It is impossible to know how extensive or complete the trawl was !

The 22-clue search gave many more puzzles per {-2+2} and was more protracted. I think i gave up.

Anyhow

just noticed this thread here

it might be a more direct way of making hard puzzles.......

i made this non-minimal 27 clue puzzle - by hand
probably fairly easy to generate them
the diagonal pattern and the clues in each box are not fixed
Code: Select all
+---+---+---+
|3..|..9|.4.|
|.2.|.7.|6..|
|..1|8..|..5|
+---+---+---+
|..6|3..|..7|
|.5.|.2.|.8.|
|4..|..1|9..|
+---+---+---+
|7..|..6|..1|
|..8|.5.|3..|
|.9.|4..|.2.|
+---+---+---+

instead of easy it was a bit hard SE 6.8/6.8/6.8
removing clues - probably going to be harder still - and there will be many puzzles .
Code: Select all
3....9.4..2..7.6....1.....5..63......5..2....4....19..7....6..1..8.5.3.....4...2.# SE=9.2/9.1/9.0

If we can generate a non-minimal 27 clue with SE 9.0 - there will be many harder puzzles generated im sure.

C
coloin
 
Posts: 2487
Joined: 05 May 2005
Location: Devon

PreviousNext

Return to General