(Unofficial) Assassin 96 V1.5

Our weekly <a href="http://www.sudocue.net/weeklykiller.php">Killer Sudokus</a> should not be taken too lightly. Don't turn your back on them.
Post Reply
mhparker
Grandmaster
Grandmaster
Posts: 345
Joined: Sat Jan 20, 2007 10:47 pm
Location: Germany

(Unofficial) Assassin 96 V1.5

Post by mhparker »

Hi folks,

Building up on Jean-Christophe's UA96 (aka. "The puzzle formerly known as YAK 96"), here's a variation on the same theme that you might like to get your teeth stuck into, especially if you've got steel ones like this guy!


(Unofficial) Assassin 96 V1.5 (UA96 V1.5) (Est. rating: 1.5)

Image

3x3:d:k:4096:4096:2562:2562:1796:5125:5125:3079:3079:4096:5130:5130:2562:1796:5125:5391:5391:3079:2322:5130:5130:5653:11542:3607:5391:5391:3354:2322:2322:5653:5653:11542:3607:3607:3354:3354:2596:2596:11542:11542:11542:11542:11542:2603:2603:2861:2861:4911:4911:11542:3634:3634:4916:4916:2861:4919:4919:4911:11542:3634:5436:5436:4916:5183:4919:4919:3650:2627:3908:5436:5436:3655:5183:5183:3650:3650:2627:3908:3908:3655:3655:

Note: It's a Killer-X, where no repeated digits are allowed on either diagonal.

Enjoy your meal!

P.S. SudokuSolver v3.0 scores this beauty at 1.83, but that's probably making a bit too much of a meal of it! :)
Cheers,
Mike
Jean-Christophe
Addict
Addict
Posts: 92
Joined: Mon Apr 23, 2007 12:50 pm
Location: Belgium

Post by Jean-Christophe »

I don't have steel teeth Image, but enjoyed solving it. Thanks
Triple click to read the optimized walkthrough that I wrote:1. The easy part
a. Innies @ r1234 -> r34c5 = 16 = {79} (NP @ c5, 45/9)
b. Innies @ r6789 -> r67c5 = 4 = {13} (NP @ c5, 45/9)
c. Innies @ c5 -> r5c5 = 8
d. 10/2 @ r89c5 = {46} (NP @ n8, c5)
e. 7/2 @ r12c5 = {25} (NP @ n2)
f. Innies @ c1234 -> r5c34 = 10 = {46} (NP @ r5)
g. r5c67 = 7/2 = {25} (NP @ r5)

2. X-Wing on 9
a. 22/3 @ r34c4+r4c3 = {9..} and 16/2 @ r34c5 = X-Wing on 9
b. -> no 9 elsewhere in r34
c. 9 @ r3 locked @ r3c45 -> locked for n2

3. n123
a. 20/3 @ r1c67+r2c6 = {389|479|578} = {5|9..} -> r1c7 = {59}, r12c6 = {3478}
b. Innies @ n3 -> r1c7+r3c9 = 12 -> r3c9 = {37}
c. 10/3 @ r1c34+r2c4 = {1(27|36|45)} (no {89}); Since r12c4 <> {25} -> r1c3 <> {47}
d. 9/3 @ r34c1+r4c2 -> no {789}
e. Innies @ n1 -> r1c3+r3c1 = 9 = {36} | [54]
f. 10/3 @ r1c34+r2c4 = {1(36|45)} -> 1 locked @ r12c4 for n2, c4

4. r3
a. {125} @ r3 locked @ r3c2378
b. Outies @ r12 -> r3c2378 = 16 = {1258} (NQ @ r3)

5. r34
a. 22(3): R3C4 <> 6 because {679} blocked by R4C5 = (79) (Afmob step 2i)
b. r3c45 = NP {79} @ r3, n2
c. r3c9 = 3, r1c7 = 9, r12c6 = {38} (NP @ n2, c6)
d. Since r3c1 = {46}, 9/3 -> r4c12 = {2(1|3)} -> 2 locked for n4, r4
e. 13/3 @ r34c9+r4c8 -> r4c89 = {46} (NP @ n6, r4)
f. 22/3 @ r34c4+r4c3 = [958], , r34c5 = [79], r5c67 = [25]

6. c6, n6, n9
a. r36c6 = HP {46} @ c6
b. r5c4&r6c6 = NP {46} @ n5
c. Two 14/3 @ c67 -> r46c7 = {137}
d. r6c89 = HP {28}; 19/3 -> r7c9 = 9
e. Innies @ n9 -> r9c7 = 1, 15/3 -> r89c6 = {59}
f. 10/2 @ r5c89 = [91] (HS @ n6)
g. 11/3 @ r6c12+r7c1 -> no 9; r6c3 = 9 (HS @ r6)
...
So, it does not require any Killer subset, but a X-Wing and a contradiction will unlock it.
Triple click to read the original harder walkthrough that I wrote: Steps 1 to 4 as above

5. c6
a. 8 @ n2 locked @ r12c6 -> locked for c6, 20/3 = {8(39|57)}
b. {46} @ c6 locked @ r346c6
c. 14/3 @ r34c6+r4c7 <> {46}+4 -> Killer HP {46} @ c6
d. -> r6c6 = {46}, 14/3 = {4|6..} within r34c6
e. r5c4&r6c6 = NP {46} @ n5
f. r36c6 = HP {46} @ c6

6. n9, c6
a. Innies @ n9 -> r7c9+r9c7 = 10 -> no 5
b. 19/3 @ r6c89+r7c9 -> r7c9 <> 1 -> r9c7 <> 9
c. 15/3 @ r89c6+r9c7 = {159|249|258|267|357} = {2|5..}
d. Since r89c6 <> {468} -> r9c7 <> 2
e. -> r5c6 & r89c6 = Killer NP {25} @ c6

7. c6, n6
a. Two 14/3 @ c67 -> no 8 because <> {158|248}
b. Two 14/3 @ c67 -> r46c7 <> 6 because r36c6 = {46}
c. {68} @ n6 locked @ r46c89
d. 13/3 @ r34c9+r4c8 <> {68..}
e. Since r7c9 <> 5 (6a); 19/3 -> r6c89 <> {68}
f. -> Killer HP {68} @ n6, each of 13/3 & 19/3 must include one of {68} within r46c89
g. -> 13/3 @ r34c9+r4c8 = {238|346} (no 7)
h. r3c9 = 3, r1c7 = 9, r12c6 = {38} (NP @ n2, c6)

8. Almost done
a. r3c45 = HP {79} @ n2
b. Since r3c1 = {46}, 9/3 -> r4c12 = {2(1|3)} -> 2 locked for n4, r4
c. 13/3 @ r34c9+r4c8 -> r4c89 = {46} (NP @ n6, r4)
d. r4c3 = 8 (HS @ r4), 22/3 -> r34c4 = [95], r34c5 = [79], r5c67 = [25]
e. r6c89 = HP {28}; 19/3 -> r7c9 = 9
f. Innies @ n9 -> r9c7 = 1, 15/3 -> r89c6 = {59}
g. 10/2 @ r5c89 = [91] (HS @ n6)
h. 3 @ n5 locked @ r6c45 -> locked for r6
i. -> r46c7 = [37]
...
Edited to fix typos and optimize WT
Last edited by Jean-Christophe on Fri Apr 04, 2008 6:45 pm, edited 5 times in total.
Afmob
Expert
Expert
Posts: 103
Joined: Sat Sep 22, 2007 5:36 pm
Location: MV, Germany

Post by Afmob »

It seems Mike found another Killer where the SS rating is quite off. Though this one is in a different league than UA 96 V2 they nearly got the same rating :shock:, in fact this one is "0.04 tougher".

UA 96 V1.5 Walkthrough

1. R6789
a) Innies = 4(2) = {13} locked for C5 + 45(9)
b) 10(2) <> 7,9
c) Innies N7 = 6(2) = {15/24}
d) Innies N9 = 10(2) <> 5; R9C7 <> 9

2. R1234
a) Innies = 16(2) = {79} locked for 45(9)
b) 7(2) = {25} locked for C5+N2
c) 9 locked Innies + 22(3) = 9{58/67} for R34
d) Innies N1 = 9(2) = {27/36/45}; R1C3 <> 2
e) 10(3) = 1{36/45} -> 1 locked for C4+N2
f) Innies N1 = 9(2) <> 2
g) 10(3): R1C3 <> 4 because R12C4 <> 5
h) Innies N1 = 9(2): R3C1 <> 5
i) 22(3): R3C4 <> 6 because {679} blocked by R4C5 = (79)
j) 9 locked in R3C45 for N2

3. R456+N8
a) Hidden Killer pair (13) in both 10(2) for R5 -> Both 10(2) = {19/37}
b) Innies C6789 = 7(2) = {25} locked for R5
c) Innies C1234 = 10(2) = {46} locked for R5
d) R5C5 = 8
e) 10(2) @ N8 = {46} locked for N8

4. N123 !
a) 20(3) <> 6 because 5,9 only possible @ R1C7; 20(3) = {389/479/578}
b) ! Killer triple (789) locked in R3C45+20(3) for N2
-> 20(3) must have exactly one of (789) @ N2
c) 20(3) must have 2 of (789) -> R1C7 <> 3,4,5
d) 20(3) = 9{38/47} -> R1C7 = 9
e) Innies N3 = 12(3) = {39} -> R3C9 = 3
f) 13(3) = 3{28/46}
g) 9(3) = 2{16/34} because R3C1 = (46) -> 2 locked R4+N4; R4C12 <> 4,6
h) 13(3) = {346} -> {46} locked for R4+N6
i) 22(3) = {589} -> 5 locked for R4
j) 14(3) = 7{16/34} -> 7 locked for R4

5. N12
a) Innies N1 = 9(2): R1C3 <> 6
b) 10(3) must have 3 xor 5 and R1C3 = (35) -> R12C4 <> 3
c) 3 locked in 20(3) @ N2 = {389} -> {38} locked for N2+C6
d) 22(3) = {589} -> R3C4 = 9, R4C4 = 5, R4C3 = 8

6. N478
a) Killer triple (123) locked in R4C12 + 10(2) for N4
b) Hidden Killer pair (12) in 11(3) @ R7C1 -> R7C1 = (12)
c) Innies N7 = 6(2) = [15/24]
d) R5C6 = 2, R5C7 = 5
e) 15(3) = 5{19/37} because R89C6 = (1579) -> 5 locked for N8; R89C6 <> 1
f) 1 locked in R7C56 for R7
g) Innies N7 = 6(2) = [24] -> R7C1 = 2, R9C3 = 4
h) 11(3) = {245} -> {45} locked for R6
i) 19(3) @ N8 = {379} -> R6C3 = 9, {37} locked for C4

7. N369
a) Innies N9 = 10(2) = [73/91]
b) 19(3) = 9{28/37} -> R7C9 = 9; R6C8 <> 7
c) 14(3) @ R6C6 = {167} -> R6C6 = 6
d) 14(3) @ R3C6 = {347} -> R3C6 = 4, R4C6 = 7, R4C7 = 3
e) R9C7 = 1
f) 21(4) @ N3 = 28{47/56} because R3C7 = (26) -> 2 locked for N3

8. Rest is singles.

Rating: (Easy) 1.25. I used a Killer triple to crack it.
Andrew
Grandmaster
Grandmaster
Posts: 300
Joined: Fri Aug 11, 2006 4:48 am
Location: Lethbridge, Alberta

Post by Andrew »

Thanks Mike. A fun puzzle. Unfortunately I missed one key move (J-C's step 2a, Afmob's step 2c) which I ought to have spotted. When I went through the other walkthroughs I realised that this is a design feature of this puzzle. Sorry Mike for missing it. :oops:

I'm pleased to see that J-C kept his original walkthrough as well as posting his improved version. There are some interesting steps in the original which weren't required in the improved version.

I rate the way I solved this puzzle as Hard 1.25. However if the key move is used, then it becomes Easy 1.25 as rated by Afmob.

Here is my walkthrough. It often happens that if one misses one important move, then other interesting moves are brought into play.

Prelims

a) R12C5 = {16/25/34}, no 7,8,9
b) R5C12 = {19/28/37/46}, no 5
c) R5C89 = {19/28/37/46}, no 5
d) R89C5 = {19/28/37/46}, no 5
e) 10(3) cage at R1C3 = {127/136/145/235}, no 8,9
f) 20(3) cage at R1C6 = {389/479/569/578}, no 1,2
g) 9(3) cage at R3C1 = {126/135/234}, no 7,8,9
h) 22(3) cage at R3C4 = {589/679}
i) 11(3) cage at R6C1 = {128/137/146/236/245}, no 9
j) 19(3) cage at R6C3 = {289/379/469/478/568}, no 1
k) 19(3) cage at R6C8 = {289/379/469/478/568}, no 1
l) 20(3) cage in N7 = {389/479/569578}, no 1,2

1. 45 rule on R1234 2 innies R34C5 = 16 = {79}, locked for C5 and 45(9) cage, clean-up: no 1,3 in R89C5
[This would give a UR no 6 in R4C3 because it would be impossible to determine the placement of 7,9 in R34C45. Of course I don’t use such moves.]

2. 45 rule on R6789 2 innies R67C3 = 4 = {13}, locked for C5 and 45(9) cage, clean-up: no 4,6 in R12C5

3. Naked pair {25} in R12C5, locked for C5 and N2, clean-up: no 8 in R89C5

4. Naked pair {46} in R89C5, locked for C5 and N8

5. R5C5 = 8 (naked single), locked for both diagonals, clean-up: no 2 in R5C12, no 2 in R5C89

6. 45 rule on C1234 2 innies R5C34 = 10 = {46} (only remaining combination), locked for R5
[At this stage I realised that there’s potential for more URs. There are in fact already one candidate elimination and one other combination elimination in a different cage that could be made using UR.]

7. 45 rule on N1 2 innies R1C3 + R3C1 = 9 = {36/45}/[72], no 1, no 2 in R1C3

8. 45 rule on N3 2 innies R1C7 + R3C9 = 12 = {39/48/57}, no 1,2,6

9. 45 rule on N7 2 innies R7C1 + R9C3 = 6 = {15/24}

10. 45 rule on N9 2 innies R7C9 + R9C7 = 10 = {28/37/46}/[91], no 5, no 9 in R9C7

11. 10(3) cage at R1C3 = {136/145}, no 7, 1 locked in R12C4, locked for C4 and N2, clean-up: no 2 in R3C1 (step 7)
11a. 5 of {145} must be in R1C3 -> no 4 in R1C3, clean-up: no 5 in R3C1 (step 7)

12. 9(3) cage at R3C1 = {126/135/234}
12a. 6 of {126} must be in R3C1 -> no 6 in R4C12

13. 20(3) cage at R1C6 = {389/479/569/578}
13a. 4 of {479} must be in R12C4 (R12C4 cannot be {79} which clashes with R3C5), no 4 in R1C7, clean-up: no 8 in R3C9 (step 8)

14. 22(3) cage at R3C4 = {589/679}
14a. 6 of {679} must be in R4C34 (R4C34 cannot be {79} which clashes with R4C5), no 6 in R3C4

15. 13(3) cage at R3C9 = {139/148/157/238/247/256/346}
15a. 3 of {139} must be in R3C9 (cannot be 9{13} which clashes with R5C89), no 9 in R3C9, clean-up: no 3 in R1C7 (step 8)

16. 20(3) cage at R1C6 = {389/479/569} (cannot be {578} which clashes with R3C45, 2-cell or ALS cage block)
16a. Killer triple 7,8,9 in R12C6 and R3C45, locked for N2

17. 14(3) cage at R3C6 = {149/167/239/248/347/356} (cannot be {158/257} because R3C6 only contains 3,4,6)
17a. 6 of {167/356} must be in R3C6 (R4C67 cannot be {56} which clashes with R3C34), no 6 in R4C67

18. 1,2 in R3 locked in R3C2378
18a. 45 rule on R12 4 outies R3C2378 = 16 = {1249/1258/1267}, no 3
18b. Killer triple 7,8,9 in R3C2378 and R3C45, locked for R3, clean-up: no 5 in R1C7 (step 8)

19. 20(3) cage at R1C6 (step 18) = {389/479}, no 6

20. 13(3) cage at R3C9 = {139/148/238/247/256/346} (cannot be {157} which clashes with R5C89)
20a. 5 of {256} must be in R3C9 -> no 5 in R4C89

21. 45 rule on N4 3 innies R456C3 = 2 outies R37C1 + 15
21a. Min R37C1 = 4 -> min R456C3 = 19, max R45C3 = 15 -> min R6C3 = 4
21b. R456C3 can only total 19,20,21,23 (cannot be 22 because {679} clashes with R5C12) -> R37C1 can only total 4,5,6,8 = [31/32/35/41/42/62], no 4 in R7C1, clean-up: no 2 in R9C3 (step 9)

22. 14(3) cage at R8C4 = {158/248/257/347} (cannot be {149} because 1,4 only in R9C3, cannot be {239} because R9C3 only contains 1,4,5), no 9

23. 15(3) cage at R8C6 = {159/168/249/258/267/348/357} (cannot be {456} because 4,6 only in R9C7)
23a. 2,4,6 of {168/258/348} must be in R9C7 (R89C6 cannot be {25} which clashes with R5C6), no 8 in R9C7, clean-up: no 2 in R7C9 (step 10)

24. 2 in C3 locked in R2378C3
24a. 45 rule on C12 4 outies R2378C3 = 15 = {1239/1248/1257} (cannot be {2346} which clashes with R5C3), no 6, 1 locked for C3, clean-up: no 5 in R7C1 (step 9)

25. Hidden killer triple 7,8,9 in R2378C3 and R46C3 for C3 -> R46C3 = {789}
25a. Killer triple 7,8,9 in R46C3 and R5C12, locked for N4
25b. 8 in N4 locked in R46C3, locked for C3, clean-up: no 4 in R2378C3 (step 24a)

26. 22(3) cage at R3C4 = {589/679}
26a. 5,6 only in R4C4 -> R4C4 = {56}

27. 11(3) cage at R6C1 = {236/245} (cannot be {146} which clashes with R5C3) -> R7C1 = 2, R6C12 = {36/45}
27a. R9C3 = 4 (step 9), R5C34 = [64], R89C5 = [46], clean-up: no 3 in R3C1 (step 7), no 3 in R6C12 (step 27)
27b. Naked pair {45} in R6C12, locked for R6 and N4

28. R4C2 = 2 (hidden single in N4)

29. 10(3) cage at R1C3 (step 13) = {136} (only remaining combination) -> R1C3 = 3, R3C1 = 6 (step 7), R4C1 = 1 (step 12), R12C4 = {16}, locked for C4 -> R4C4 = 5, locked for D\, R5C67 = [25], clean-up: no 7 in R3C2378 (step 18a), no 9 in R2378C3 (step 24a), no 9 in R5C12

30. Naked pair {37} in R5C12, locked for R5 and N4
30a. Naked pair {19} in R5C89, locked for N6

31. 22(3) cage at R3C4 = {589} (only remaining combination), no 7

32. R3C5 = 7 (hidden single in R3), R4C5 = 9, R46C3 = [89], R3C4 = 9, clean-up: no 4 in R3C2378 (step 18a)

33. 20(3) cage at R1C6 (step 19) = {389} (cannot be {479} because 7,9 only in R1C7) -> R1C67 = [89], R2C6 = 3, R34C6 = [47], 7 locked for D/, R4C7 = 3, R3C9 = 3 (step 8)

34. R6C3 = 9 -> R67C4 = 10 = [37], 3 locked for D/, R6C56 = [16], 6 locked for D\, R7C5 = 3

35. R6C6 = 6 -> R6C7 + R7C6 = 8 -> R6C7 = 7, R7C6 = 1, R7C3 = 5, locked for D/, R7C7 = 4, locked for D\, R1C1 = 7, locked for D\, R5C12 = [37], R89C1 = [89], R789C2 = [613], 1 locked for D/, R8C3 = 7, R3C7 = 2, R23C3 = [21], 1 locked for D\, R2C2 = 9, locked for D\

and the rest is naked singles

7 5 3 6 2 8 9 1 4
4 9 2 1 5 3 8 6 7
6 8 1 9 7 4 2 5 3
1 2 8 5 9 7 3 4 6
3 7 6 4 8 2 5 9 1
5 4 9 3 1 6 7 2 8
2 6 5 7 3 1 4 8 9
8 1 7 2 4 9 6 3 5
9 3 4 8 6 5 1 7 2

It's interesting that J-C and Afmob didn't use URs in their solutions. Maybe they aren't very helpful.

Now to have another go at J-C's V2 to see if I can make further progress.
Last edited by Andrew on Mon Apr 28, 2008 10:42 pm, edited 1 time in total.
Afmob
Expert
Expert
Posts: 103
Joined: Sat Sep 22, 2007 5:36 pm
Location: MV, Germany

Post by Afmob »

In the other UA 96 thread (where I didn't post because of possible future wts) I suggested that we might want to discuss who is going to post the next UA on friday.

So for this friday I want to provide the next UA (97). If nobody is against this or really wants to post his creation then I'll post my assassin between thursday night/friday morning (GMT +01:00 :!:). But I would only post V1, others can feel free, if requested, to post other version(s) of my killer.
Post Reply