The New Sudoku Players' Forum

by **999_Springs** » Tue Aug 27, 2013 9:45 pm

Pat wrote:
999_Springs wrote:

i'd love to try out the puzzles that have been posted,
but if you keep posting them at like 2 AM
then i'll never get a chance

i'm usually on Lunar Standard Time

just posted this one in the Patterns Game
— does this suit you?

.17...2..58.9.....6.9.....1.9.6....7.....76......3..5.2...6.3.......9.8...14....2
[ play ]

Code: Select all: . 1 7 . . . 2 . . 5 8 . 9 . . . . . 6 . 9 . . . . . 1 . 9 . 6 . . . . 7 . . . . . 7 6 . . . . . . 3 . . 5 . 2 . . . 6 . 3 . . . . . . . 9 . 8 . . . 1 4 . . . . 2

this is really hard

here's my solution path (spoiler) with a few thoughts and stuff

Hidden Text: Show

if anyone else wants to try doing this, go for it. if you could find a neater solution path too, that would be great. knowing this forum, that's not likely to happen

se rates it 8.6

by **Marty R.** » Wed Aug 28, 2013 1:12 am

For those of us who are not familiar with SE and its ratings, could you provide some perspective about the meaning of an 8.6?

by **champagne** » Wed Aug 28, 2013 5:39 am

Marty R. wrote:For those of us who are not familiar with SE and its ratings, could you provide some perspective about the meaning of an 8.6?

This means that in SE sequence the hardest step is something like this

Code: Select all: A B C |D E F |G H I 34 1 7 |358 45 6 |2 9 58 5 8 234 |9 127 123 |47 367 36 6 234 9 |23578 2457 23458 |58 347 1 ----------------------------------------------- 1348 9 2345 |6 12458 12458 |148 123 7 148 2345 2345 |1258 9 7 |6 123 348 1478 267 26 |128 3 1248 |9 5 48 ----------------------------------------------- 2 457 8 |157 6 15 |3 147 9 347 34567 3456 |12357 1257 9 |14 8 456 9 3567 1 |4 78 358 |57 67 2

[] 1r8c4 -> ~1r8c7 -> 4r8c7 -> ~4r2c7 -> 4r2c3 -> ~2r2c3 -> 2r3c2 -> ~2r3c6
[] 1r8c4 -> ~3r8c4 -> 3r9c6 -> ~3r3c6
[] 1r8c4 -> ~1r8c7 -> 1r7c8 -> ~4r7c8 -> 4r3c8 -> ~4r3c6
[] 1r8c4 -> ~1r7c6 -> 5r7c6 -> ~5r3c6
[] 1r8c4 -> ~1r8c7 -> 1r4c7 -> ~8r4c7 -> 8r3c7 -> ~8r3c6

<1r8c4>

This does not exclude a nicer solution using other rules.

by **999_Springs** » Wed Aug 28, 2013 4:25 pm

<6.5: se's solving methodology is highly flawed at this level
6.5-7.0: x-chains, x-cycles, y-cycles (not xy-chains. i think that's a bug though)
7.1-7.5: standard aic's (no grouping)
7.6-8.1: fish (expressed as nishio)
8.2-8.7: kraken row/column/block/cell (no grouping) (like what champagne is doing but backwards, so you start from a cell and work towards the elimination, but the logic is the same)
8.7-9.5: singles forcing nets
9.3-10.1: forcing nets with lc's and pairs and stuff
10.0-10.9: forcing nets with nested aic's
10.8+: forcing nets with nested... more stuff

from a manual solving perspective:

<6.5: easy
6.5-7.0: the easier end of stuff posted here
7.1-7.5: the harder end of stuff posted here, sometimes there's a one-step solution, sometimes not
7.6-8.1: you may wish to set aside a decent chunk of time to do this. look for various types of fish
8.2-8.7: not necessarily that much harder than above, except don't look for fish, and don't expect one-step solutions
8.7-9.5: expect a long hard chain-based slog. sometimes there's a quick solution if you're lucky, but usually not
9.3-10.1: excruciatingly painful, can still be done, but do you really want to
10.0+: may i suggest throwing toilet paper at a tree as a more productive use of your time

by **champagne** » Thu Aug 29, 2013 9:37 am

999_Springs wrote:from a manual solving perspective:
9.3-10.1: excruciatingly painful, can still be done, but do you really want to
10.0+: may i suggest throwing toilet paper at a tree as a more productive use of your time

If you limit yourself to chain nets, this is true.

A significant part of these puzzles can be easily solved using "exotic" patterns as SK loop, Exocets, big rank 0 logic. Most of them see the difficulty significantly reduced applying such rules.

In a puzzle where such patterns don't work, I would follow your advice.

by **Pat** » Thu Aug 29, 2013 10:33 am

999_Springs wrote:this is really hard

the 8.6 rating indicates a tough move

but then again, for me even 7.3 is too tough

.38...2..12.9.....4.9.....1.9.1....6.....28......5..7.2...8.7.......6.3...12....5

the real question:
how many tough moves does it take?

i tried to choose puzzles which require multiple tough moves

on my 7.3,
SuDoku Explainer says:
on my 8.6,
SuDoku Explainer says:

- . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - . - .

999_Springs wrote:6.5-7.0: x-chains, x-cycles, y-cycles

gsf's solver says the above puzzles
can be solved by X-cycles + Y-cycles

-q{NF}{B2B3}{H2T2H3T3H4T4}{W2W3W4}{XY}-G

7.3: Show

8.6: Show

by **999_Springs** » Thu Aug 29, 2013 4:53 pm

champagne wrote:
999_Springs wrote:from a manual solving perspective:
9.3-10.1: excruciatingly painful, can still be done, but do you really want to
10.0+: may i suggest throwing toilet paper at a tree as a more productive use of your time

If you limit yourself to chain nets, this is true.

A significant part of these puzzles can be easily solved using "exotic" patterns as SK loop, Exocets, big rank 0 logic. Most of them see the difficulty significantly reduced applying such rules.

In a puzzle where such patterns don't work, I would follow your advice.

ok this is news to me. i've been out of these forums for 3 years so i don't know much about these new methods or anything about how manual players can solve ultra hard puzzles using them. can you point me to some resources please? links that start from the beginning and have a strong emphasis on applying them to manual solving would be great

also did you do this? http://i.imgur.com/0faKbUv.jpg

pat: i have no idea what those weird symbols mean in your output but in sudoku explainer an x-cycle is a closed loop version of a forcing x-chain and a y-cycle is a closed loop version of an xy-chain, which are definitely confined to sub-7 territory. also what's up with the morse code in the middle of your post

by **champagne** » Thu Aug 29, 2013 5:36 pm

Hi 999_Springs,

a comprehensive answer would require more time, so, just a "flavour" of what can be done

false double exocet

this is a link within a thread dedicated to such powerful rules, but not in the list I mentioned.

Have a look to these puzzles rating at minimum SE diamond 10.3

Most of them should collapse after a short analysis of a visible potential complementary double AAHS

by **Pat** » Fri Aug 30, 2013 9:41 am

999_Springs wrote:in sudoku explainer, an x-cycle is a closed loop version of a forcing x-chain
and a y-cycle is a closed loop version of an xy-chain,
which are definitely confined to sub-7 territory

-q{NF}{B2B3}{H2T2H3T3H4T4}{W2W3W4}{XY}-G
in gsf's command-line [ manual ] —

level 1 = NF = "singles"
level 2 = B = b\l, l\b
level 3 = HT = subsets
level 4 = W = fish (not used)
level 5 = XY, quite possibly not matching the SuDoku Explainer definition

this is the 8.6 after "basic" moves [ play ] —

Code: Select all: +-------+-------+---------+ | . 1 7 | . . 6 | 2 9 58 | | 5 8 . | 9 . . | . . . | | 6 . 9 | . . . | 58 . 1 | +-------+-------+---------+ | . 9 . | 6 . . | . . 7 | | . . . | . 9 7 | 6 . . | | . . . | . 3 . | 9 5 . | +-------+-------+---------+ | 2 . 8 | . 6 . | 3 . 9 | | . . . | . . 9 | . 8 . | | 9 . 1 | 4 . . | . . 2 | +-------+-------+---------+

at this point, gsf's software offers 6 exclusions by Y-cycles —

Code: Select all: 5 5 Y6 [23][41][51][53]^3 [25][26]^4

exclude 3 at [23][41][51][53]
exclude 4 at [25][26]

does this make sense ?

by **Pat** » Mon Sep 02, 2013 11:09 am

i've now obtained the detailed report for those 6 exclusions —

Code: Select all: y-cycle 1 3 20a [23]-[11]>[43]- y-cycle 2 3 20a [41]-[11]>[43]- y-cycle 3 3 3a [51]-[29][59]- y-cycle 4 3 13a [53]-[59]>[83]- y-cycle 5 4 26a [25]-[15]>[83]-[72][78]- => [25]-[15]>[83]- y-cycle 6 4 26a [26]-[15]>[83]-[72][78]- => [26]-[15]>[83]-

does this make sense ?

by **999_Springs** » Wed Sep 11, 2013 9:09 pm

unfortunately i don't get it... sudoku explainer lists all 6 of those eliminations as dynamic fc's, the shortest one being 3r1 --> 3r5c1 off:

Hidden Text: Show

Chain 1: If R1C1 contains the value 3, then R5C1 cannot contain the value 3 (View 1):
(1) If R1C1 contains the value 3, then R5C1 cannot contain the value 3 (the value can occur only once in the column)

Chain 2: If R1C4 contains the value 3, then R5C1 cannot contain the value 3 (View 2):
(1) If R1C4 contains the value 3, then R3C6 cannot contain the value 3 (the value can occur only once in the block)
(2) If R1C4 contains the value 3 (initial assumption), then R2C6 cannot contain the value 3 (the value can occur only once in the block)
(3) If R2C6 does not contain the value 3 and R3C6 does not contain the value 3 (1), then R9C6 must contain the value 3 (only remaining possible position in the column)
(4) If R9C6 contains the value 3, then R9C6 cannot contain the value 8 (the cell can contain only one value)
(5) If R9C6 does not contain the value 8, then R9C5 must contain the value 8 (only remaining possible position in the block)
(6) If R9C5 contains the value 8, then R1C5 cannot contain the value 8 (the value can occur only once in the column)
(7) If R1C4 contains the value 3 (initial assumption), then R1C4 cannot contain the value 8 (the cell can contain only one value)
(8) If R1C4 does not contain the value 8 and R1C5 does not contain the value 8 (6), then R1C9 must contain the value 8 (only remaining possible position in the row)
(9) If R1C9 contains the value 8, then R6C9 cannot contain the value 8 (the value can occur only once in the column)
(10) If R6C9 does not contain the value 8, then R6C9 must contain the value 4 (only remaining possible value in the cell)
(11) If R6C9 contains the value 4, then R5C9 cannot contain the value 4 (the value can occur only once in the block)
(12) If R1C9 contains the value 8 (8), then R5C9 cannot contain the value 8 (the value can occur only once in the column)
(13) If R5C9 does not contain the value 8 and R5C9 does not contain the value 4 (11), then R5C9 must contain the value 3 (only remaining possible value in the cell)
(14) If R5C9 contains the value 3, then R5C1 cannot contain the value 3 (the value can occur only once in the row)

and i can't construct that at all out of gsf's output, and it also doesn't even involve r2c9 anywhere.

certainly none of these eliminations can be done with an xy-chain or loop. in particular, the xy-chain in my solution isn't there until you kill 8r1c5 - which gsf does not do

meh i'll just ask gsf

by **Pat** » Thu Sep 12, 2013 3:10 pm

well gsf may be busy

so i changed the reporting to "-v3" (to get greater detail)

-v3: Show

then changed the reporting to "-v4" (to get even greater detail)

-v4: Show

Code: Select all: x-hinge 1 6 [41]-[78] x-hinge 1 6 [45]-[78] x-hinge 1 6 [46]-[78] x-hinge 3 7 [32]-[84] x-hinge 3 7 [52]-[84] x-hinge 4 3 [23]-[78] x-hinge 4 3 [25]-[78] x-hinge 4 3 [26]-[78] x-hinge 5 7 [52]-[89] x-hinge 5 9 [82]-[37] x-hinge 5 9 [83]-[37] x-hinge 5 9 [84]-[37] x-hinge 5 9 [85]-[37] x-hinge 6 7 [89]-[62] x-hinge 7 7 [84]-[62] x-hinge 7 7 [85]-[62] x-hinge 7 7 [87]-[62] x-hinge 8 6 [41]-[19] x-hinge 8 6 [45]-[19] x-hinge 8 6 [46]-[19] y-edge 3 22 [11]>[28] y-edge 3 18 [11]>[43] y-edge 3 8 [11]=[59] y-edge 4 9 [15]>[27] y-edge 4 19 [15]>[52] y-edge 4 7 [15]>[69] y-edge 4 11 [15]>[78] y-edge 4 21 [15]>[83] y-edge 3 11 [59]>[83] y-edge 3 6 [59]=[96] y-cycle 1 3 20a [23]-[11]>[43]- y-cycle 2 3 20a [41]-[11]>[43]- y-cycle 3 3 3a [51]-[29][59]- y-cycle 4 3 13a [53]-[59]>[83]- y-cycle 5 4 26a [25]-[15]>[83]-[72][78]- => [25]-[15]>[83]- y-cycle 6 4 26a [26]-[15]>[83]-[72][78]- => [26]-[15]>[83]- 5 5 Y6 [23][41][51][53]^3 [25][26]^4

i'm still in the dark

"-v5" pours out tons of extra stuff ("y-trip")
which may explain what's really happening
but not something a human can follow

yes we'll have to wait for gsf

Website · by **denis_berthier** » Thu Sep 12, 2013 3:15 pm

Full resolution path using Subsets and g-whips. Longest chain = 6

Hidden Text: Show

Code: Select all: (solve ".17...2..58.9.....6.9.....1.9.6....7.....76......3..5.2...6.3.......9.8...14....2") ***** SudoRules 16.2 based on CSP-Rules 1.2, config: gW-S ***** 24 givens, 221 candidates, 1451 csp-links and 1451 links. Initial density = 1.49 singles ==> r9c1 = 9, r6c7 = 9, r5c5 = 9, r7c3 = 8 hidden-pairs-in-a-block: b3{r1c9 r3c7}{n5 n8} ==> r3c7 <> 7, r3c7 <> 4, r1c9 <> 9, r1c9 <> 6, r1c9 <> 4, r1c9 <> 3 singles ==> r1c8 = 9, r7c9 = 9, r1c6 = 6 biv-chain[4]: b8n8{r9c5 r9c6} - b8n3{r9c6 r8c4} - r1n3{c4 c1} - r1n4{c1 c5} ==> r1c5 <> 8 biv-chain[4]: r1c5{n4 n5} - c9n5{r1 r8} - r9c7{n5 n7} - r2c7{n7 n4} ==> r2c6 <> 4, r2c5 <> 4 whip[4]: r1n3{c1 c4} - r1n8{c4 c9} - r5c9{n8 n4} - r6c9{n4 .} ==> r5c1 <> 3 whip[5]: r6c9{n4 n8} - r1c9{n8 n5} - r1c5{n5 n4} - r3n4{c5 c8} - r7n4{c8 .} ==> r6c2 <> 4 whip[5]: r6c9{n4 n8} - r1c9{n8 n5} - r1c5{n5 n4} - r3n4{c5 c2} - r7n4{c2 .} ==> r5c8 <> 4 whip[4]: c3n6{r6 r8} - c9n6{r8 r2} - c9n3{r2 r5} - r5n4{c9 .} ==> r6c3 <> 4 whip[5]: r6c9{n4 n8} - r1c9{n8 n5} - r1c5{n5 n4} - r3n4{c5 c2} - r7n4{c2 .} ==> r4c8 <> 4 whip[3]: r2c7{n7 n4} - c8n4{r2 r7} - b9n1{r7c8 .} ==> r8c7 <> 7 whip[4]: c7n1{r8 r4} - c8n1{r4 r7} - b9n4{r7c8 r8c9} - b6n4{r6c9 .} ==> r8c7 <> 5 whip[5]: r5n4{c1 c9} - r6c9{n4 n8} - r4c7{n8 n1} - r8c7{n1 n4} - r2n4{c7 .} ==> r4c3 <> 4 whip[6]: b9n7{r9c8 r7c8} - c8n4{r7 r3} - r2n4{c7 c3} - r1c1{n4 n3} - r8c1{n3 n4} - b9n4{r8c9 .} ==> r9c2 <> 7 whip[6]: r5n4{c1 c9} - r6c9{n4 n8} - r4c7{n8 n1} - b4n1{r4c1 r5c1} - r5n8{c1 c4} - r1n8{c4 .} ==> r6c1 <> 4 g-whip[6]: r6c9{n4 n8} - r1n8{c9 c4} - r1n3{c4 c1} - c1n4{r1 r456} - r5n4{c2 c1} - r5n8{c1 .} ==> r8c9 <> 4 hidden-pairs-in-a-block: b9{r7c8 r8c7}{n1 n4} ==> r7c8 <> 7 whip[1]: b9n7{r9c8 .} ==> r9c5 <> 7 biv-chain[3]: r9c5{n8 n5} - c7n5{r9 r3} - c7n8{r3 r4} ==> r4c5 <> 8 whip[5]: c7n1{r4 r8} - b8n1{r8c5 r7c4} - r6n1{c4 c1} - c1n7{r6 r8} - b8n7{r8c4 .} ==> r4c6 <> 1 whip[5]: c7n1{r4 r8} - c4n1{r8 r7} - r6n1{c4 c1} - c1n7{r6 r8} - b8n7{r8c4 .} ==> r4c5 <> 1 biv-chain[3]: c5n1{r2 r8} - r8c7{n1 n4} - r2c7{n4 n7} ==> r2c5 <> 7 whip[1]: b2n7{r3c5 .} ==> r3c8 <> 7 hidden-pairs-in-a-column: c8{n6 n7}{r2 r9} ==> r2c8 <> 3 whip[3]: r3c8{n3 n4} - r2n4{c7 c3} - r1c1{n4 .} ==> r3c2 <> 3 whip[3]: b1n3{r1c1 r2c3} - b4n3{r5c3 r5c2} - c9n3{r5 .} ==> r8c1 <> 3 biv-chain[3]: r1n4{c5 c1} - r8c1{n4 n7} - c5n7{r8 r3} ==> r3c5 <> 4 biv-chain[4]: b8n3{r9c6 r8c4} - r8n2{c4 c5} - c5n7{r8 r3} - c5n8{r3 r9} ==> r9c6 <> 8 hidden-single-in-a-block ==> r9c5 = 8 biv-chain[4]: r9c6{n3 n5} - b9n5{r9c7 r8c9} - c9n6{r8 r2} - b3n3{r2c9 r3c8} ==> r3c6 <> 3 whip[4]: b8n2{r8c5 r8c4} - b8n3{r8c4 r9c6} - r2c6{n3 n1} - r2c5{n1 .} ==> r3c5 <> 2 whip[5]: c7n5{r3 r9} - c9n5{r8 r1} - r1n8{c9 c4} - c4n3{r1 r8} - r9c6{n3 .} ==> r3c4 <> 5 whip[5]: b8n2{r8c5 r8c4} - c4n7{r8 r3} - c4n3{r3 r1} - r1c1{n3 n4} - r8c1{n4 .} ==> r8c5 <> 7 hidden-single-in-a-column ==> r3c5 = 7 biv-chain[4]: b9n1{r8c7 r7c8} - r7c6{n1 n5} - r3n5{c6 c7} - c7n8{r3 r4} ==> r4c7 <> 1 singles ==> r8c7 = 1, r7c8 = 4, r3c8 = 3, r5c9 = 3, r2c5 = 1 whip[1]: c2n3{r9 .} ==> r8c3 <> 3 whip[1]: r5n4{c1 .} ==> r4c1 <> 4 whip[1]: b3n4{r2c7 .} ==> r2c3 <> 4 biv-chain[3]: b1n4{r3c2 r1c1} - r1n3{c1 c4} - r8n3{c4 c2} ==> r8c2 <> 4 biv-chain[3]: b7n3{r8c2 r9c2} - r9n6{c2 c8} - r8c9{n6 n5} ==> r8c2 <> 5 biv-chain[3]: b6n8{r4c7 r6c9} - r1n8{c9 c4} - r5n8{c4 c1} ==> r4c1 <> 8 biv-chain[3]: c4n3{r8 r1} - r1c1{n3 n4} - r8c1{n4 n7} ==> r8c4 <> 7 singles ==> r7c4 = 7, r7c2 = 5, r7c6 = 1 naked-pairs-in-a-column: c2{r3 r5}{n2 n4} ==> r6c2 <> 2 x-wing-in-rows: n5{r3 r9}{c6 c7} ==> r4c6 <> 5 whip[2]: r2n2{c6 c3} - r6n2{c3 .} ==> r4c6 <> 2 naked-pairs-in-a-row: r4{c6 c7}{n4 n8} ==> r4c5 <> 4 singles to the end 317546298 582913476 649278531 193654827 425897613 876132954 258761349 764329185 931485762

Considering only Subsets and whips, the max. length is 7.
Allowing braids and/or g-braids doesn't change the resolution paths.

The New Sudoku Players' Forum

pat's 8.6 in the patterns game

pat's 8.6 in the patterns game

Re: pat's 8.6 in the patterns game

Re: pat's 8.6 in the patterns game

Re: pat's 8.6 in the patterns game

Re: pat's 8.6 in the patterns game

re: Pat # 212 # 56

Re: pat's 8.6 in the patterns game

Re: pat's 8.6 in the patterns game

re: XY

Re: pat's 8.6 in the patterns game

re: XY in gsf's software

Re: pat's 8.6 in the patterns game - gwhip solution