SudokuSolver Forum

A forum for Sudoku enthusiasts to share puzzles, techniques and software
It is currently Thu Mar 28, 2024 5:14 pm

All times are UTC




Post new topic Reply to topic  [ 4 posts ] 
Author Message
 Post subject: Assassin 396
PostPosted: Mon Jun 01, 2020 5:57 am 
Offline
Grand Master
Grand Master

Joined: Wed Apr 16, 2008 1:16 am
Posts: 1040
Location: Sydney, Australia
Attachment:
a396.JPG
a396.JPG [ 91.99 KiB | Viewed 6674 times ]
Assassin 396

This is a really fun puzzle, then has a mighty sting in the tail. Perfect Assassin really....Come a long way since Ruud got this series started 2nd June, 2006. SudokuSolver gives it a 1.75, JSudoku uses 7 complex intersections (not default solvers)

triple click code:
3x3::k:2816:4097:4097:1794:3331:3331:5380:5380:5893:2816:6662:1794:1794:5380:5380:5380:5893:5893:6662:6662:6662:4103:1288:4361:4361:5898:5893:6662:6662:6667:4103:1288:7948:7948:5898:5893:3853:3853:6667:6667:6667:2574:7948:5898:5898:2837:2837:3853:0000:6667:2574:7948:7948:7948:3601:5394:5394:0000:6667:0000:0000:0000:3348:3601:3601:5394:0000:0000:3091:3091:3091:3348:2576:2576:5394:0000:0000:0000:4623:4623:4623:
Solution:
Code:
+-------+-------+-------+
| 3 9 7 | 4 8 5 | 1 2 6 |
| 8 5 2 | 1 6 3 | 9 4 7 |
| 1 6 4 | 7 2 9 | 8 3 5 |
+-------+-------+-------+
| 2 8 6 | 9 3 7 | 4 5 1 |
| 9 1 3 | 5 4 2 | 6 7 8 |
| 4 7 5 | 6 1 8 | 3 9 2 |
+-------+-------+-------+
| 5 3 1 | 8 7 4 | 2 6 9 |
| 7 2 8 | 3 9 6 | 5 1 4 |
| 6 4 9 | 2 5 1 | 7 8 3 |
+-------+-------+-------+


Top
 Profile  
Reply with quote  
 Post subject: Re: Assassin 396
PostPosted: Fri Jun 05, 2020 12:40 am 
Offline
Grand Master
Grand Master

Joined: Tue Jun 16, 2009 9:31 pm
Posts: 280
Location: California, out of London
Thanks Ed! No shortcuts on this one! :) A few easy placements to start followed by a tough middle and a tougher end!
I would expect to see some very different solving paths after the start.
Unusually for me I have a complete WT - which means there is more likely to be some errors/typos.
Some corrections & clarifications thanks to Ed and Andrew.
Assassin 396 WT:
1. 16(2)n1 = {79}
-> 13(2)n2 = {58}
-> 17(2)r3 = [98]
-> 16(2)c4 = [79]

2. 7(3)r1c4 = {124}
-> Whichever of (124) is in r2c3 goes in n2 in r3c5
-> r2c56 = {36}
-> Cells in 21(5)r1c7 in n3 are [{12}9] or [{14}7]

3. Remaining outie n36 -> r4c6 = 7
-> (79) both in n8 in r789c5

4. -> 26(6)r2c2 has no (79) -> must be {124568}
-> 3 in n1 in r1c1
-> 11(2)n1 = [38]
-> r4c2 = 8
Also (HS) -> r1c9 = 6

5. Remaining IOD n6 -> r3c8 = r4c9 + 2
Since r3c8 is max 5 -> r4c9 is max 3
-> Whatever goes in r4c9 must go in n3 in r1c78 (I.e., from (12))

6! Whichever of (124) is in r2c3 also goes in r4c1 and r3c5
-> r4c15 = +5(2)
It also goes in n3 in r1c78
-> r4c19 have the same values as in r1c78 i.e., either {12} or [41]
But since r4c15 = +5(2) -> r4c19 cannot also be +5(2)
-> r4c19 = {12}
-> r1c78 = {12} and r1c4 = 4 and r2c7 = 9

7. 7 in n6 only in r5c89
-> 7 in n4 in r6c123
-> 7 in 15(3)n4 or 11(2)n4

Innies n14 = r245c3 = +11(3) - > No 9
-> 9 also in 15(3)n4 or 11(2)n4

-> One of:
(A) 11(2)n4 = {29} -> 15(3)n4 = [537] -> r4c15 = [14] -> r45c3 = [64]
(B) 11(2)n7 = {47} -> 15{3}n4 = {159} -> r4c15 = [23] -> r45c3 = [63]

I.e., r4c3 = 6 and r4c5 = r5c3 (either 3 or 4)

8! IOD r123456 -> r7c5 = r6c4 + 1
-> 26(6) must contain a 1 which can only go in n5

6 already in 26(6)-> r7c5 cannot be 6 -> r6c4 cannot be 5
-> 5 in n5 in 26(6)

Since r4c5 from (34) -> [r6c4,r7c5] cannot be [34]
-> 3 must be in 26(6) either in r5c3 or in n5

-> 26(6) = {1356(29|47)} (No 8)

-> Either 10(2)n5 = {28} and r6c4 = 6 (and r7c5 = 7)
or 10(2)n5 = {46} and r6c4 = 8 (and r7c5 = 9)

9! r1c78 = {12} prevents 31(6)r4c6 = [7{124}{89}] (Simpler - r4c9 from (12) does the same thing)
-> In n6 one of (89) in r6c89 and the other in r5c89
Since one of (89) in r6c89 and one of (82) in r6c46 -> 11(2)n4 cannot be {29}!
-> 11(2)n4 = {47}
-> 15(3)n4 = {159}
-> r4c1 = 2 -> r2c34 = [21] -> 5(2)r3c5 = [23] and r5c3 = 3
Also -> r4c9 = 1 -> r3c8 = 3
Also r2c56 = [63]

10. 23(5)n3 = [6{457}1] with 7 in r2c89
-> 7 in n9 in c7r789
-> 13(2)n9 from {49} or {58}
-> r2c8 from (45)
-> r2c9 = 7
-> r5c8 = 7
Also -> r6c8 from (89)

11. Either r6c4 = 6 or 10(2)n5 = [46]
-> HS 6 in n6 -> r5c7 = 6
r4c78 = {45}
-> 23(4) = [3578] or [3479]
-> (HP) r6c79 = {23}
-> 8 in n5 either in r6c4 or r6c6
-> r6c8 = 9
-> 23(4) = [3578]
-> r4c7 = 4

12. Also -> 13(2)n9 = {49}
-> r2c8 = 4 and r3c9 = 5
-> r2c2 = 5 and r3c123 = {146}

13. Only possible solution for 18(3)n9 = [783]
-> r6c79 = [32]

14. 8 in n7 in r78c3
HP r78c2 = {23}
-> 2 in r9 in n8 in r9c46
Either r9c4 = 2 -> 10(2)n5 = [28] -> 13(2)n2 = [85]
Or r9c6 = 2 -> r5c4 = 2 -> 5 in r56c4 -> 13(2)n2 = [85]
Either way -> 13(2)n2 = [85]
(Simpler is: 8 in c5 in r1c5 or r8c5.
But the latter puts r7c5 = 7 puts r6c4 = 6 and 10(2)n5 = [28] which contradicts 13(2)n2 = {58}
-> 8 in c5 only in r1c5)


15. r8c2 is either 2 in r8c2 -> r78c1 = {57}
or r8c2 is 3 -> r78c1 = {56}
Either way 5 in n7 in r78c1
-> 5 in n8/r9 in r9c45

16! Trying 2 in r9c6 puts 10(2)n5 = [46] puts r56c4 = [28] puts r789c5 = [974] puts r9c4 = 5
This leaves no place for 6 in r9
-> HS r9c4 = 2

17. -> 10(2)n5 = [28]
-> r6c4 = 6 and r7c5 = 7
Also r9c5 = 5
-> r8c5 = 9
-> 13(2)n9 = [94]
Also r5c4 = 5
-> r56c5 = [41]
Also -> 15(3)n4 = [{19}5]

18. Also r78c4 = {38} and r789c6 = {146} with r8c6 from (16)
-> 12(3)r8 only from [156] or [651]
-> 2 in n9 in r7c78
-> r78c2 = [32]
-> r78c4 = [83] and r8c3 = 8
Also r78c1 = [57]
-> 11(2)n4 = [47]
-> 16(2)n1 = [97]
-> (HS 6 in n7) 10(2)n7 = [64]
-> r79c3 = [19]
-> r3c123 = [164]
-> r5c12 = [91]
Also -> r789c6 = [461]
-> 12(3)r8 = [651]
-> r1c78 = [12]
-> r7c78 = [26]


Last edited by wellbeback on Sun Jun 14, 2020 7:02 pm, edited 1 time in total.

Top
 Profile  
Reply with quote  
 Post subject: Re: Assassin 396
PostPosted: Wed Jun 10, 2020 8:06 am 
Offline
Grand Master
Grand Master

Joined: Wed Apr 16, 2008 1:16 am
Posts: 1040
Location: Sydney, Australia
Really enjoyed wellbeback's WT. So many lovely clone sequences. Love 6. Really like the simplicity of 8 as well. Think I'd call that a type of "Locking-out cages". In all, I counted (a record?) 5 !s in wellbeback's WT! (note: two ! in step 9, which deserves them)

I only managed to find one clone which was very important for my solution (step 13) [edit: check out Andrew's step 3 for a really neat way to avoid this clone]. So, as predicted, quite a different WT but as often happens, we worked in the same areas. My 10, 13, 14 & 28 are the main differences. [Edit: ahhh, so many typos! Thanks Andrew and wellbeback for checking my WT]

Assassin 396 walk-through.
32 steps:
Prelims courtesy of SudokuSolver
Cage 16(2) n1 - cells ={79}
Cage 16(2) n25 - cells ={79}
Cage 17(2) n23 - cells ={89}
Cage 5(2) n25 - cells only uses 1234
Cage 13(2) n2 - cells do not use 123
Cage 13(2) n9 - cells do not use 123
Cage 10(2) n7 - cells do not use 5
Cage 10(2) n5 - cells do not use 5
Cage 11(2) n4 - cells do not use 1
Cage 11(2) n1 - cells do not use 1
Cage 7(3) n12 - cells ={124}

note: this WT is optimised so not all clean-up done, only what is shown. Also, naked singles may be available but not placed because they are not essential at that point.
1. 17(2)r3c6 = {89}: both locked for r3
1a. r34c4 = [79]

2. 16(2)n1 = {79}: both locked for r1 and n1

3. 13(2)n2 = {58}: both locked for r1 and n2
3a. r3c67 = [98]

4. 7(3)r1c4 = {124}, r2c56 sees all those -> no 1,2,4 (Common Peer Elimination, CPE)

5. r2c56 = {36}: both locked for r2, 21(5), 3 for n2
5a. r2c56 = 9 -> r1c78 + r2c7 = 12 = {129/147}
5b. r2c7 = (79)
5c. 1 locked for n3, r1
5d. naked triple 1,2,4 in r1c478, 2 & 4 locked for r1

6. r2c56 = 9, r3c6 = 9
6a. -> from "45" on n36: 1 remaining outie r4c6 = 7
6b. 7 in n6 only in r5: locked for r5

7. 26(6)r2c2 = {124568}(no 3)
7a. r1c1 = 3 (hsingle n1), r2c1 = 8
7b. r1c9 = 6
7c. 26(6) must have 8 -> r4c2 = 8
7d. 5 & 6 in n1 only in 26(6) -> no 5,6 in r4c1

8. "45" on c123: 3 innies r245c3 = 11 (no 9)

9. hidden killer pair 7,9 in n4
9a. -> 11(2)n4 must have one of 7,9 = {29/47}(no 3,5,6)
9b. -> 15(3)n6 must have one of 7,9: but {249/267} blocked by 11(2)
9c. = {159/357}(no 2,4,6)
9d. must have 5: locked for n4

10. "45" on n12346: 1 remaining outie r4c5 + 6 = 2 innies r45c3
10a. since r4c35 see each other, the 3rd cell cannot be the Innie/outie Difference of 6 (IOU)
10b. no 6 in r5c3
10c. r4c3 = 6 (hsingle n4)

11. 6 in n5 only in 10(2) = {46} or in r6c4
11a. -> no 4 in r6c4 (Locking out cages)

12. 10(2)n5 = {28/46}(no 1,3)
12a. 5(2)r3c5 = {14}/[23]

13. "45" on n12346: 1 remaining outie r4c5 = 1 innie r5c3 = (134)
13a. -> from cage sum 26(6)r4c3, r4567c5+r5c4 = 20
13b. -> but {12458} blocked since must have 4 in r7c5 to escape clash with 10(2)n5, and also means [41] in r34c5: but two 4's in c5
13c. = {12359/13457}(no 8)
13d. must have 7 or 9 with are only in r7c5 -> r7c5 = (79)
13e. and 26(6)r4c3 = [6]{12359/13457}: note, has both 3 & 5

14. 15(3)n4 = {159}/{357}
14a. but {357} as [537] only, leaves no place for both 3 & 5 in 26(6)r4c3
14b. = {159} only: 1 & 9 locked for n4

15. 11(2)n4 = {47}: 4 locked for r6 & n4
15a. r4c1 = 2, r5c3 = 3 -> r4c5 = 3 (step 13), r3c5 = 2, r2c56 = [63], r12c4 = [41], r2c3 = 2
15b. no 6 in r5c6

16. r1c78 = {12} = 3 -> r2c7 = 9 (cage sum)

17. 23(5)r1c9 = {14567} only (no 3)
17a. -> r4c9 = 1
17b. r3c8 = 3 (hsingle n3)
17c. -> 23(4)r3c8, which must have 7 for r5 = [3]{479/578}(no 2,6), r5c89 = {78/79}
17d. r5c7 = 6 (hsingle r5)

18. 13(2)n9 = {49/58}(no 7)

19. killer pair 4,5 in r3c9 with 13(2)n9: both locked for c9
19a. r2c9 = 7
19b. r5c8 = 7 (hsingle n6)

20. r24c8 = {45}: both locked for c8

21. killer pair 8,9 in r5c9 with 13(2)n9: both locked for c9

22. 10(2)n7 = {19/46}/[73] (no 2, 7 in r9c2)
22a. 18(3)n9: must have 2 or 3 for r9c9
22b. but {369} blocked by 10(2)n7
22c. = {279/378}(no 1,4,5,6)
22d. -> r9c7 = 7, r9c89 = 11 = [92/83]
22e. no 3 in 10(2)n7

23. killer pair 8,9 in r9c8 and 13(2)n9: both locked for n9

24. hidden pair 2,3 in n7 -> r78c2 = {23}

25. 14(3)n7 must have one of 2 or 3 for r8c2
25a. but {47}[3] blocked by r6c1 = (47)
25b. = {57}[2]/{56}[3](no 1,4,9)
25c. 5 locked for c1 and n7

26. "45" on r1..6: 1 innie r6c4 + 1 = 1 outie r7c5
26a. = [67/89]

27. 26(6)r4c3 = [63]([2]{15}[9]/[5417])
27a. note: has 5 in r5c4 or 9 in r7c5
27b. 1 locked for c5

Took many permutations of this step to get it to the simplest. Hope it's correct.
28. "45" on r9: 4 innies r9c3456 = 17 and must have 5 for r9
28a. = {1259/1358/2456}
28b. 4 in {2456} must be in r9c3 -> no 4 in r9c56
28c. 5 only in n8, locked for n8
28d. the only permutation with 5 in r9c4 = [1592] but this is blocked by 26(6)r4c3 (step 27a)
28e. -> no 5 in r9c4
28f. -> r5c4 = 5 (hsingle c4)
28g. -> r567c5 = [417] (step 27)

29. 10(2)n5 = {28}: both locked for c6, 8 for n5

30. r5c12 = {19}: 9 locked for r5 and n4
30a. r5c9 = 8
30b. -> r9c8 = 8 (hsingle c8)
30c. r9c9 = 3 (cage sum)

31. r1c56 = [85], r8c5 = 9
31a. r78c9 = [94]
31b. r89c6 = {16}: both locked for n8

32. 12(3)r8c6 = {156} only
32a. -> r8c7 = 5, 1 & 6 both locked for r8

Singles now.
Cheers
Ed


Top
 Profile  
Reply with quote  
 Post subject: Re: Assassin 396
PostPosted: Wed Jun 10, 2020 11:04 pm 
Offline
Grand Master
Grand Master

Joined: Wed Apr 23, 2008 6:04 pm
Posts: 1893
Location: Lethbridge, Alberta, Canada
Thanks Ed for this latest Assassin. A hard one right to the end, definitely a sting in the tail; it took me a while to find an acceptable way to do that bit.

Ed mentioned clones in his and wellbeback's walkthoughs; I didn't knowingly use any but perhaps in at least one place I used an implied one.

Thanks Ed for pointing out the simpler way for step 11a; I also then spotted a simpler way for it.

Here is my walkthrough for Assassin 396:
Prelims

a) R12C1 = {29/38/47/56}, no 1
b) R1C23 = {79}
c) R1C56 = {49/58/67}, no 1,2,3
d) R34C4 = {79}
e) R34C5 = {14/23}
f) R3C67 = {89}
g) R56C6 = {19/28/37/46}, no 5
h) R6C12 = {29/38/47/56}, no 1
i) R78C9 = {49/58/67}, no 1,2,3
j) R9C12 = {19/28/37/46}, no 5
k) 7(3) cage at R1C4 = {124}

1a. Naked pair {79} in R1C23, locked for R1 and N1, clean-up: no 2,4 in R12C1, no 4,6 in R1C56
1b. Naked pair {58} in R1C56, locked for R1 and N2, clean-up: no 3,6 in R2C1
1c. R3C67 = [98] -> R34C3 = [79], clean-up: no 1 in R56C6
1d. 7(3) cage at R1C4 = {124}, CPE no 1,2,4 in R2C56
1e. Naked pair {36} in R2C56, locked for R2 and 21(5) cage at R1C7, 3 locked for N2, clean-up: no 2 in R4C5
1f. R2C56 = 9 -> R1C78 + R2C7 = 12 = {129/147} -> R2C7 = {79}, R1C78 = {12/14}, 1 locked for R1 and N3
1g. Naked triple {124} in R1C478, 2,4 locked for R1
1h. 7(3) cage at R1C4 = {124}, 1 locked for R2

2a. 45 rule on N36 using R1C78 + R2C7 = 12, 1 remaining outie R4C6 = 7, clean-up: no 3 in R56C6
2b. 7 in N6 only R5C89, 7 locked for R5
2c. 1 in C6 only in R789C6, locked for N8
2d. 45 rule on N14 3 innies R245C3 = 11 = {128/146/236/245}, no 9

3a. 45 rule on R123456 1 outie R7C5 = 1 innie R6C4 + 1, no 8 in R7C5
3b. 26(6) cage at R4C3 = {123569/123578/124568/134567} (cannot be {123479} because 7,9 only in R7C5
3c. 7,9 of {123569/123578/134567} must be in R7C5 -> no 3 in R7C5, clean-up: no 2 in R6C4
3d. R6C4 + R7C5 = [12/45/56/67/89] (cannot be [34] which clashes with R34C5 = [41]), no 3 in R6C4, no 4 in R7C5
3e. R6C4 + R7C5 = [45/56/67/89] (cannot be [12] which clashes with R34C5 = [23] because 3 in N5 must be in R4C5 when 26(6) cage = {124568}), no 1 in R6C4, no 2 in R7C5
3f. R6C4 + R7C5 = [56/67/89] (cannot be [45] because R56C6 = {28} and 26(6) cage = {124568} would require 2,4,8 in R45C3), no 4 in R6C4, no 5 in R7C5
3g. R6C4 + R7C5 = [67/89] (cannot be [56] because R56C6 = {46}, only other place for 6 in N5 and R7C5 clash with R2C56), no 5 in R6C4, no 6 in R7C5
[Some of those can be considered to be equivalent to forcing chains.]
3h. 26(6) cage at R4C3 = {123569/123578/134567}
3i. 5 in N5 only in R5C45 + R6C5, locked for 26(6) cage, no 5 in R45C3
3j. Killer pair 6,8 in R56C6 and R6C4, locked for N5

4a. 45 rule on N3 using R1C78 + R2C7 = 12, 1 innie R3C8 = 1 outie R4C9 + 2 -> R3C8 = {3456}, R4C9 = {1234}
4b. 23(5) cage at R1C9 = {12569/13469/14567/23459/23567} (cannot be {12479} because R1C9 must contain one of 3,6)
4c. 23(5) cage = {12569/14567/23459/23567} (cannot be {13469} which clashes with R3C8 + R4C9 = [31]), 5 locked for N3, clean-up: no 3 in R4C9

5a. 45 rule on N124 using R2C56 = 9, 2 innies R45C3 = 1 outie R4C5 + 6, IOU no 6 in R5C3
5b. R245C3 (step 2d) = {128/146/236}
5c. 6 of {146/236} must be in R4C3 -> no 3,4 in R4C3

[I ought to have spotted this obvious step earlier but hadn’t checked and found that 26(6) cage without 7 or 9 has one combination; I ought to have realised that since it’s missing 7 and 9 it must also be missing 3 to make a total of 19 missing. I’d also used combinations in step 3 for the other 26(6) cage.]
6a. 26(6) cage at R2C2 = {124568}, no 3
6b. R1C1 = 3 (hidden single in N1) -> R2C1 = 8, R1C9 = 6, clean-up: no 3,8 in R6C2, no 7 in R78C9, no 2,7 in R9C2, no 4 in R4C9 (step 4a)
6c. 26(6) cage = {124568} -> R4C2 = 8, clean-up: no 2 in R9C1
6d. 5,6 in N1 only in R2C2 + R3C123, locked for 26(6) cage, no 5,6 in R4C1
6e. 26(6) cage at R4C3 (step 3h) = {123569/134567} -> R4C3 = 6, clean-up: no 5 in R6C12
6f. 5 in N4 only in 15(3) cage at R5C1 = {159/357}, no 2,4
6g. 7 of {357} must be in R6C3 -> no 3 in R6C3
6h. 45 rule on N4 2 remaining innies R4C1 + R5C3 = 5 = [14/23/41], no 2 in R5C3
6i. 18(3) cage at R9C7 = {279/369/378/567} (cannot be {189/459/468} which clash with R78C9), no 1,4
6j. R9C12 = {19/46} (cannot be [73] which clashes with 18(3) cage), no 3,7
6k. 18(3) cage = {279/378/567} (cannot be {369} which clash with R9C12), 7 locked for R9 and N9
6l. 14(3) cage at R7C1 = {239/257/347/356} (cannot be {149/167} which clash with R9C12), no 1
6m. 3 of {239/347/356} must be in R8C2 -> no 4,6,9 in R8C2
6n. 45 rule on N9 4 innies R78C78 = 14 = {1238/1256/1346} (cannot be {2345} which clashes with R78C9), no 9

7a. 5 in R4 only in R4C78, locked for N6
7b. 7 in R5 only in 23(4) cage at R3C8 = [35]{78}/{34}{79}, 3 locked for C8
7c. 45 rule on R1234 2 outies R5C89 = 1 remaining innie R4C7 + 11
7d. R5C89 = {78/79} = 15/16 -> R4C7 = {45}

[Something else I ought to have spotted earlier.]
8a. 3 in N4 only in R5C23, locked for R5
8b. 3 in N5 only in R46C5, locked for C5 -> R2C56 = [63]

9a. 15(3) cage at R5C1 (step 6f) = {159/357}, 14(3) cage at R7C1 (step 6l) = {239/257/347/356}
9b. Consider combinations for R4C1 + R5C3 (step 6h) = [14/23/41]
R4C1 + R5C3 = {14} => 15(3) cage = [537] => 14(3) cage = {27}5
or R4C1 + R5C3 = [23] => 15(3) cage = {159} => R6C12 = {47}=> 14(3) cage = {257/356} (cannot be {29}3 which clashes with R4C1, cannot be {47}3 which clashes with R6C1)
-> 14(3) cage = {257/356}, no 4,9, 5 locked for N7

10a. 1 in C9 only in R46C9, locked for N6
10b. 31(6) cage at R4C6 = {145678/234679/235678} (cannot be {124789} which clashes with R5C89, cannot be {135679} which clashes with 23(4) cage at R3C8, step 7b)
10c. 31(6) cage = {234679/235678} (cannot be {145678} = [75{46}81] which clashes with R6C12 + R6C46, killer ALS block), no 1, 3 locked for R6 and N6
10d. Naked pair {45} in R4C78, 4 locked for R4 and N6
10e. R4C9 = 1 (hidden single in N6) -> R3C8 = 3 (step 4a), R4C1 = 2, R4C5 = 3 -> R3C5 = 2, R12C4 = [41], R2C3 = 2, R5C3 = 3 (step 6h), clean-up: no 9 in R6C12
10f. Naked pair {47} in R6C12, 4 locked for R6, 7 locked for N4, clean-up: no 6 in R5C6
10g. R1C78 = {12}, R2C56 = [63] -> R2C7 = 9 (cage sum)
10h. R9C7 = 7 (hidden single in C7) -> R9C89 = 11 = {29}/[65/83], no 5 in R9C8, no 8 in R9C9
10i. 14(3) cage at R7C1 (step 9b) = {257/356} -> R8C2 = {23}, R78C1 = {56/57}, 5 locked for C1
10j. R78C2 = {23} (hidden pair in C2)

11a. Consider combinations for 21(4) cage at R7C2 = 2{478}/3{189}
21(4) cage = 2{478}, 7 locked for C3 => R1C3 = 9 => R6C3 = {15}
or 21(4) cage = 3{189}, 1,9 locked for C3 => R6C3 = 5
-> R6C3 = {15}
[Ed pointed out 21(4) cage at R7C2 = 2{478}/3{189} -> killer pair 7,9 in R1C3 and R789C3, locked for C3.
I also realised that R6C35 = {15} (hidden pair for R6) was simpler.
So I’d seen this step as a forcing chain when both killer pair and hidden pair were simpler.]

11b. 9 in N4 only in R5C12, locked for R5
11c. R5C89 = {78}, 8 locked for R5 and N6, R3C8 = 3 -> R4C8 = 5 (cage total), clean-up: no 2 in R6C6
11d. R4C7 = 4, R5C7 = 6 (hidden single in R5)
11e. 5 in N3 only in R23C9, locked for C9, clean-up: no 8 in R78C9, no 6 in R9C8 (step 10h)
11f. Naked pair {49} in R78C9, locked for C9 and N9 -> R23C9 = [75], R2C28 = [54], clean-up: no 2 in R9C89 (step 10h)
11g. R6C3 = 5 (hidden single in N4) -> R6C5 = 1
11h. R9C89 = [83]
11i. 2,5 in R9 only in R9C456, locked for N8
11j. 12(3) cage at R8C6 = {156/246}, no 8, 6 locked for R8
11k. 12(3) cage = {156/246} = [156/651/426], no 1 in R8C7 = {25}, no 2 in R8C8
11l. R59C4 = {25} (hidden pair in C4)

[I’ve now reached Ed’s “sting in the tail”, how to eliminate 12(3) cage at R8C6 = [156]. I refuse to use Unique Rectangle, R8C6 cannot be 1 => R1C78 and R7C78 both {12} whose order cannot be resolved, as that relies on the solution being unique but walkthroughs are meant to reach a unique solution.]

[Or maybe that’s not initially the right approach.]
12a. R56C6 = [28/46], R7C5 = R6C4 + 1 (step 3a), 14(3) cage at R7C1 (step 9b) = {257/356}
12b. Consider permutations for 12(3) cage at R8C6 (step 11k) = [156/651/426]
12(3) cage = [156] => 14(3) cage = [572], 6 in N7 only in R9C12 = {46}, R9C35 = [19] (hidden pair in R9) => R7C5 = 7, R6C4 = 6 => R56C6 = [28]
or 12(3) cage = [426/651] => R8C6 = {46} => R56C6 = [28] (cannot be [46] which clashes with R8C6)
-> R56C6 = [28], R5C45 = [54], R6C4 = 6, R7C5 = 7, R1C56 = [85], R89C5 = [95], R78C9 = [94]
12c. 12(3) cage = [156/651] -> R8C7 = 5, 1 locked for R8
12d. R8C1 = 7 -> R7C1 + R8C2 = 7 = [52], R6C1 = 4
12e. 6 in N7 only in R9C12 = [64]

and the rest is naked singles.

Rating Comment:
I'll rate my walkthrough for A396 at 1.5. I used IOD analysis and later a few forcing chains, the last of which feels like a full 1.5.


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 4 posts ] 

All times are UTC


Who is online

Users browsing this forum: No registered users and 10 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group