Bored89 (aka waiting for Assassin 89v2)

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
Nasenbaer
Master
Master
Posts: 167
Joined: Thu Jul 20, 2006 7:19 am
Location: Fellbach (Deutschland)

Bored89 (aka waiting for Assassin 89v2)

Post by Nasenbaer »

This weeks samurai is already done and no v2 for assassin 89 in sight. So I got bored and tried my hands on creating a nice killer. Not as easy as it looks. I got a lot of unsolvable puzzles, or they were too easy. Finally I managed to create two versions of a killer, Bored89-Easy and Bored89-Hard. They only differ in the center nonet.

Bored89-Easy (SudokuSolver score: 1,508)
Image
3x3:d:k:4352:4352:4352:3843:3843:3843:6406:6406:6406:4352:5642:5642:5642:3843:4622:4622:4622:6406:3858:3858:5642:3605:4374:4374:4622:3353:3353:3858:6428:3605:3605:4895:1312:3617:4130:3353:6428:6428:2342:1831:4895:1312:3617:4130:4130:3885:6428:2342:1831:4895:3890:3890:4130:4661:3885:3885:4408:3385:3385:3890:4924:4661:4661:5951:4408:4408:4408:4419:4924:4924:4924:4423:5951:5951:5951:4419:4419:4419:4423:4423:4423:
Bored89-Hard (SudokuSolver score: 1,976)
Image
3x3:d:k:4352:4352:4352:3843:3843:3843:6406:6406:6406:4352:5642:5642:5642:3843:4622:4622:4622:6406:3858:3858:5642:3605:4374:4374:4622:3353:3353:3858:6428:3605:3605:3871:3871:3617:4130:3353:6428:6428:2342:4135:3871:3871:3617:4130:4130:3885:6428:2342:4135:4135:3890:3890:4130:4661:3885:3885:4408:3385:3385:3890:4924:4661:4661:5951:4408:4408:4408:4419:4924:4924:4924:4423:5951:5951:5951:4419:4419:4419:4423:4423:4423:
No T&E is needed. I used one short chain for the hard version but I don't know if it is really necessary. Be sure to have a combo-table at hand! :wink:

Comments are appreciated. Have fun!
Nasenbaer
Afmob
Expert
Expert
Posts: 103
Joined: Sat Sep 22, 2007 5:36 pm
Location: MV, Germany

Post by Afmob »

What a tough Killer. There was only little progress you could make with no breakthrough moves at all but by picking off every candidate one by one :wink: I was able to solve it.

I think it would be best if we name A89 V2 (Easy) as A89 V1.5 since it fits with the difficulty though I haven't solved it but tried and it seems to be at least as tough as A89 V1.

We cannot expect that Ruud's makes all V2 for us so it's good that other users do this. But right now, I would love to have some easier assassins which aren't of rating 1.25-1.5 since they are the most fun to solve.

A89 V2 (Hard) Walkthrough:

1. N123
a) 17(2) = {89} locked for R3+N2
b) Outies N1 = 9(1+1): R4C1 <> 1,9
c) Outies N3 = 11(1+1) <> 1; R4C9 <> 2,3

2. N789+D\/
a) Outies R89 = 3(2) = {12} locked for R7
b) Outies N9 = 9(1+1) <> 9
c) R3C37+R5C5 <> 1,2 since they see all 1,2 of R7
d) 15(3) @ N7: R6C1 <> 9 because R7C12 >= 7
e) Outies N7 = 10(1+1): R8C4 <> 1

3. R123 + D\/
a) Outies R12 = 7(2) = {34} locked for R3
b) R5C5 <> 3,4 because it sees all 3,4 of R3
c) 15(4): R4C56+R5C6 <> 8,9 because R5C5 >= 5
d) 15(3): R4C1 <> 5 because R3C12 <> 3,4,8
e) Outies N1 = 9(1+1): R2C4 <> 4
f) 13(3) = {157/247/256} <> 8,9 (no {148} because R3C89 <> 4,8)
g) Outies N3 = 11(1+1): R2C6 <> 2,3
h) 22(4): R2C23 <> 1 because R2C4+R3C3 <= 11

4. N45
a) Innies N4 = 11(3) <> 9
b) Innies N5 = 14(2) = {59/68}
c) 15(3) @ N5: R6C7 <> 8,9 because R67C6 >= 8
d) Killer pair (56) of Innies N5 blocks {1356} of 15(4)
e) 15(4): R4C56+R5C6 <> 6 because R5C5 <> 2,3,4

5. N5689 + N7
a) Innies N5689 = 20(2+1): R8C4 <> 2,3 because R4C49 <= 16
b) Outies N7 = 10(1+1): R6C1 <> 7,8
c) 14(3): R4C3 <> 6 because R34C4 <> 3 and R4C4 <> 1,7
d) Innies N6 = 15(3): R6C9 <> 1 because R4C9+R6C7 <= 13
e) Outies N9 = 9(1+1): R8C6 <> 8

6. C789
a) Outies = 20(4): R8C6 <> 6 because (12) only possible there and R6C6 <> 3,4,7
b) Outies N9 = 9(1+1): R6C9 <> 3
c) Killer pair (12) locked in R7C7 + 17(4) for N9
d) Innies N6 = 15(3) = {168/267/348/357} since other combos blocked by Killer pairs (56,58) of 14(2)
e) 16(4) <> {1258/1456/2356} since they are blocked by Killer pairs (56,58) of 14(2)
f) 16(4) <> 5 because (137) is a Killer triple of Innies N6
g) Innies N6 = 15(3): R6C79 <> 4 because 3,8 only possible there
h) Innies N6 = 15(3): R6C7 <> 5 because R46C9 <> 3
i) Outies N9 = 9(1+1): R8C6 <> 5

7. N1
a) 17(4) <> 34{19/28} because R3C3 = (34)
b) 15(3) <> {348} because they are only possible @ R4C1
c) 15(3) <> 4 because {56}4 blocked by Killer pair (56) of 17(4)
d) Outies = 9(1+1): R2C4 <> 5

8. R789 + N6 !
a) Innies N6 = 15(3): R6C79 >= 8 because R4C9 <= 7
b) Outies R789 = 19(4): R6C16 <= 11 because R6C79 >= 8
c) ! Outies R789 = 19(4): R6C1 <> 5,6 because R6C16 would be {56} (step 8b) which would
force R6C79 = [17] -> no combo for Innies N6
d) Outies N7 = 10(1+1): R8C4 <> 4,5
e) Innies N8 = 15(3): R8C6 <> 7 because R7C6+R8C4 >= 9
f) Outies N9 = 9(1+1): R6C9 <> 2
g) 19(4): R8C78 <> 3,4 because 19(4) must have at least 2 of (56789) and they are only possible there
h) Innies N6 = 15(3) must have 1,2 xor 3 and it's only possible @ R6C7 -> R6C7 <> 6,7
i) 15(3) @ N8: R7C6 <> 3 because R6C67 <> 4,7
j) 17(4) @ N7: R8C23 <> 9 because R8C4 >= 6

9. C789
a) Outies = 20(4): R8C6 <> 4 because R267C6 would be >= 18
b) Outies N9 = 9(1+1): R6C9 <> 5
c) 16(4) <> 6 because either 3 in N6 is in 16(4) = 34{18/27} or in
Innies N6 = {357} -> 16(4) <> {1267}
e) 16(4) = 4{129/138/237} -> 4 locked for N6
f) 13(3) = 5{17/26} -> 5 locked between C9+N3 -> R12C9 <> 5
g) Outies N3 = 11(1+1): R2C6 <> 7

10. R789 + N6
a) Innies N8 = 15(3): R7C6 <> 9 because R8C4 >= 6
b) 15(3) @ N8: R7C6 <> 6 because R6C67 <> 4,7 and R6C67 = [81] clashes with Innies N6
c) Outies R789 = 19(4): R6C1 <> 2 because:
- <> 24{58/67} since 4 only possible there
- <> {2359} since (59) only possible @ R6C6
- <> {1279} since R6C79 = [17] clashes with Innies N6
- <> {2368} because R6C7 = 3 forces Innies N6 = {357} -> no 5,7 in Outies R789

d) Outies N9 = 9(1+1): R8C4 <> 8
e) 17(4) @ N7 <> 37{16/25} because (37) is a Killer pair of 19(4)
f) 17(4) @ N7: R8C23 <> 6,7 because R8C4 = (679)
g) 13(2) <> {58} since it's a Killer pair of 15(3) @ N7
h) 23(4) <> 58{19/37/46} since {58} is a Killer pair of 15(3) @ N7
i) 17(4) @ N7 <> {1358} because of Killer pair (58) of 15(3) @ N7
j) 23(4) <> {2489} since (24) is a Killer pair of 17(4) @ N7
k) 15(3) <> 7 because (57) is a Killer pair of 23(4)

11. R789 !
a) ! 17(4) @ N7 <> 3 because R8C4 = (679) and 17(4) = {1349} forces
19(4) = {1279} which clashes with R8C4 = 9
b) Killer pair (58) locked in 15(3) + 17(4) for N7
c) 23(4) = 79{16/34} -> 9 locked for N7
d) 2 locked in 17(4) @ N7 = 2{159/168/456} <> 7
e) Outies N7 = 10(1+1): R6C1 <> 3
f) 15(3) @ N7 must have 1 xor 4 and R6C1 = (14) -> R7C12 <> 4
g) Innies N8 <> 3 because R8C4 = (69)
h) 19(4) = {1279} -> {79} locked for R8+N9
i) Outies N7 = 10(1+1) = [46] -> R6C1 = 4, R8C4 = 6
j) 13(2) = {49} locked for R7+N8
k) 18(3) = 7{38/56} -> R6C9 = 7

12. R789
a) Hidden Single: R7C6 = 7 @ R7
b) Outies N9 = 9(1+1): R8C6 = 2
c) 15(3) @ N8 = 7{26/35}
d) R7C7 = 1, R7C3 = 2
e) Hidden Single: R3C6 = 9 @ C6, R3C5 = 8

13. N456
a) Innies N4 = 11(3) = [25/61]4
b) 9(2) = {18/36}
c) 14(3) = 5{18/27}
d) Innies N5 = 14(2) = [86] -> R4C4 = 8, R6C6 = 6
e) 15(3) @ R6C6 = {267} -> R6C7 = 2
f) Innies N6 = 15(3) = {267} -> R4C9 = 6
g) 13(3) = {256} -> {25} locked for R3+N3
h) 14(3) = {158} -> R3C4 = 1, R4C3 = 5
i) 15(3) = {267} -> R4C1 = 2; {67} locked for N1
j) 15(4) = {1347} -> R5C5 = 7; {134} locked for N5

14. N236
a) 14(2) = [95] -> R4C7 = 9, R5C7 = 5
b) 9 locked in 25(4) @ C9 = 79{18/36} -> 7,9 locked for N3
c) Innies N2 = 13(3) = {157} -> R2C4 = 7, R2C6 = 5
d) Hidden pair (26) in R12C5 for C5 locked for N2
e) Naked pair (34) locked in R1C46 for R1

15. C123
a) Hidden Single: R4C2 = 7 @ R4, R3C2 = 6, R3C1 = 7
b) 15(3) = 4{38/56}: R7C1 <> 5
c) Hidden Single: R1C1 = 5 @ C1, R9C3 = 7 @ C3
d) Hidden Single: R5C3 = 6 @ C3 -> R6C3 = 3
e) R3C3 = 4, R3C7 = 3

16. R123
a) 18(4) = {3456} -> 6 locked for R2+N3
b) R2C5 = 2, R1C5 = 6
c) Hidden Single: R1C2 = 2 @ R1
d) 22(4) = {3478} -> R2C2 = 3, R2C3 = 8

17. N7
a) 17(4) = {1268} -> R8C3 = 1, R8C2 = 8
b) R1C3 = 9, R1C9 = 1

18. Rest is singles without considering diagonals.

Rating: 1.75. I used some heavy combination analysis which could be considered contradiction chain-like though very short.
Last edited by Afmob on Sat Mar 22, 2008 8:32 am, edited 1 time in total.
sudokuEd
Grandmaster
Grandmaster
Posts: 257
Joined: Mon Jun 19, 2006 11:06 am
Location: Sydney Australia

Post by sudokuEd »

Afmob wrote:we name A89 V2 (Easy) as A89 V1.5 since .. it seems to be at least as tough as A89 V1
Didn't seem quite that hard - though haven't checked my solution. I'll do a WT on the weekend for Bored89-Easy (SudokuSolver score: 1,508) and edit it in here. [edit: haven't had a chance to look again yet, will do a new post when finished]

BTW - the score is actually 1 point 508 (ie 3 decimal places): 1.508. I find the 3rd decimal place a bit distracting. Wish we could get scores accurate to even 1 decimal place! Every weeks killer seems to throw up a new challenge to SudokuSolver Scores. Richard's already done 3 lots of changes to SS this week to finally get A88 down to 1.37 - split out cage combining from Locked Candidates, added Killer triples and split out Simple Single House 45 Rule. As I keep saying, we'll give a proper ratings update when we have something we are happy with. Can't wait - it seems to keep me very quiet here!
Afmob wrote:But right now, I would love to have some easier assassins which aren't of rating 1.25-1.5 since they are the most fun to solve
That's good to know. I found an interesting version of an "old school" type V2 for A89, but couldn't toughen it up enough. Next time I won't ditch it quite so quick.

Thanks Nasenbaer for some fine Xs!

Cheers
Ed
Last edited by sudokuEd on Sun Feb 17, 2008 11:21 am, edited 1 time in total.
Afmob
Expert
Expert
Posts: 103
Joined: Sat Sep 22, 2007 5:36 pm
Location: MV, Germany

Post by Afmob »

By the way, the score of Nasenbaer for both puzzles is correct since the german 1,508 is the same as the english 1.508 so we use dot and comma the other way around.

Since A89 is of rating 1.25 (I rated it 1.5 but Andrew showed an easier way to solve it) I think that A89 V1.5 is as hard as A89 V1. But I can judge further on its rating when Ed's wt appears.
mhparker
Grandmaster
Grandmaster
Posts: 345
Joined: Sat Jan 20, 2007 10:47 pm
Location: Germany

Post by mhparker »

Hi guys,

How's everyone getting on with Bored 89 "Easy"? :roll:

I found time for a couple of sessions and have been chipping away at it. Despite the name, it's actually proving to be quite a tough puzzle (although compared to A90, it's all relative, I suppose...). So far, I agree with Afmob's description: lots of methodical work picking off individual candidates, but no apparent opportunity to by-pass this with any "blockbuster" type of move. I'm enjoying it though, although I still haven't managed to crack it yet.

Look forward to seeing how Ed did it. :wink:
Cheers,
Mike
sublue
Regular
Regular
Posts: 21
Joined: Fri Oct 12, 2007 3:35 am
Location: Champaign, Illinois, USA

Post by sublue »

I have a confession to make. Bored89 Easy was too hard. So I looked at Bored89 Hard. Couldn't do that one either. This is not good; I couldn't do 3 of the last 4 Assassins. It suddenly occurred to me that these 2 puzzles would have to have the same solution, since only N5 differed. So by comparing the two N5s I got a digit placed very easily. Suddenly I had a grip on the thing and was able to solve it (them) with only moderate annoyance. Thus, I don't have head north to Canada to meet Frank in a pub. I suspect my method is cheating, but a girl's gotta do what a girl's gotta do.
Susan
sudokuEd
Grandmaster
Grandmaster
Posts: 257
Joined: Mon Jun 19, 2006 11:06 am
Location: Sydney Australia

Post by sudokuEd »

mhparker wrote:Look forward to seeing how Ed did it. :wink:
Easy really. Just make the h15(3) in n8 a h17(3)!! Helps heaps. :wink:

I found a more orthodox way through, but again, haven't checked that solution. Got a real buzz doing an X killer again. Still my fav.

But the WT will have to wait until this weekend. Have an appalling work-rate at the moment - 1 WT every 8 weeks :wink: .

Cheers
Ed
Andrew
Grandmaster
Grandmaster
Posts: 300
Joined: Fri Aug 11, 2006 4:48 am
Location: Lethbridge, Alberta

Post by Andrew »

I've just finished Bored89-Easy. Nice puzzle! :D

It is easy but only after one finds the key move. It appears from the earlier messages that some people didn't find it. I must admit that I was also struggling until I spotted step 3a.

I wasn't sure how to rate this puzzle. Step 3a isn't a difficult move but it's hard to spot unless one is looking for that sort of move. Obviously I can't take into account the time "wasted" before that step so I'll rate Bored89-Easy as a 1.25.

Here is my walkthrough. Thanks Afmob for your comments which I've used for editing several steps. Fortunately my error in step 8 didn't affect other steps; that elimination is now made in step 26c.

This is a Killer-X. I've included eliminations along diagonals, for those not using software to do eliminations, because they are easy to overlook.

Prelims

a) R3C56 = {89}, locked for R3 and N2
b) R45C6= {14/23}
c) R45C7 = {59/68}
d) R56C3 = {18/27/36/45}, no 9
e) R56C4 = {16/25} (cannot be {34} which clashes with R45C6)
f) R7C45 = {49/58/67}, no 1,2,3
g) R456C5 = {289/379/469/478/568}, no 1

1. Killer pair 1,2 in R45C6 and R56C4, locked for N5

2. 45 rule on R12 2 outies R3C37 = 7 = {16/25/34}, no 7

3. 45 rule on R89 2 outies R7C37 = 3 = {12}, locked for R7
3a. CPE no 1,2 in R3C37, clean-up: no 5,6 in R3C37 (step 2)
3b. Naked pair {34} in R3C37, locked for R3
3c. CPE no 3,4 in R5C5

4. 15(3) cage at R3C1 = {159/168/258/267/357/456} (cannot be {249/348} because 3,4,8,9 only in R4C1
4a. 8,9 of {159/168} must be in R3C12 -> no 1 in R4C1
4b. 3,4,8,9 of {159/258/357/456} must be in R4C1 -> no 5 in R4C1

5. 13(3) cage at R3C8 = {157/247/256} (cannot be {139/148/238/346} because 3,4,8,9 only in R4C9), no 3,8,9

6. 45 rule on N1 2 outies R2C4 + R4C1 = 9 = [18]/{27/36}/[54], no 9, no 4 in R2C4

7. 45 rule on N3 2 outies R2C6 + R4C9 = 11 = {47/56}, no 1,2,3

8. 45 rule on N7 2 outies R6C1 + R8C4 = 10 = {19/28/37/46}/[55]
[I first made a "doubles possible" error more than a year ago. Since then I've usually managed to avoid that error so I don't know why I made it this time. I think that every time I've noticed "doubles possible" for outies they have been eliminated before the final solution. Maybe sometime there will be a puzzle with this "trap" using the doubles as part of the final solution. :wink: ]


9. 45 rule on N9 2 outies R6C9 + R8C6 = 9 = {18/27/36/45}, no 9

10. 45 rule on N4 3 innies R4C13 + R6C1 = 11 = {128/137/146/236/245}, no 9, clean-up: no 1 in R8C4 (step 8)

11. 45 rule on C6789 3 innies R139C6 = 20 = {389/479/569/578}, no 1,2
11a. 3 of {389} must be in R1C6 -> no 3 in R9C6

12. 45 rule on N5 2 innies R4C4 + R6C6 = 14 = {59/68}

13. 7 in N5 locked in R456C5, locked for C5, clean-up: no 6 in R7C4
13a. R456C5 = {379/478}, no 5,6

14. Killer pair 8,9 in R3C5 and R456C5, locked for C5, clean-up: no 4,5 in R7C4

15. 45 rule on C1234 3 innies R179C4 = 16 = {169/178/259/268/349/358/367/457}
15a. 9 of {169/259/349} must be in R7C4 -> no 9 in R9C4

16. Hidden killer pair 1,2 in R45C6 and R8C6 for C6 -> R8C6 = {12}
16a. 45 rule on C789 2 outies R28C6 = 1 innie R6C7 + 5
16b. Max R28C6 = 9 -> max R6C7 = 4
16c. R28C6 = 6,7,8,9 = [51/52/61/62/71/72] (cannot be [42] which clash with R45C6), no 4, clean-up: R6C9 = {78} (step 9), no 7 in R4C9 (step 7)

17. 45 rule on N6 3 innies R4C9 + R6C79 = 15 = {168/267/348/357} (cannot be {456} because R6C9 only contains 7,8, cannot be {258} which clashes with R45C7)
17a. 3 of {348} must be in R6C7 -> no 4 in R6C7

18. R7C7 + R8C6 = {12} = 3 -> R8C78 = 16 = {79}, locked for R8 and N9, clean-up: no 1,3 in R6C1 (step 8)
18a. CPE no 1,2 in R8C9

19. Killer triple 7,8,9 in R4C4 + R6C6, R5C5 and R8C8, locked for D\

20. 15(3) cage at R6C6 = {168/249/258/267/348/357} (cannot be {159} which clashes with R139C6, cannot be {456} because R6C7 only contains 1,2,3)
20a. 9 of {249} must be in R6C6 -> no 9 in R7C6
20b. 3 of {348/357} must be in R6C7 -> no 3 in R7C6

21. 45 rule on N8 3 innies R7C6 + R8C46 = 15 = {168/258/267} (cannot be {348/357/456} because R8C6 only contains 1,2), no 3,4, clean-up: no 6,7 in R6C1 (step 8)
21a. 2 of {258/267} must be in R8C6 -> no 2 in R8C4, clean-up: no 8 in R6C1 (step 8)

22. Hidden killer pair 3,4 in R19C6 and R45C6 for C6 -> R19C6 must contain 3 or 4 -> R139C6 (step 11) = {389/479}, no 5,6, 9 locked for C6, clean-up: no 5 in R4C4 (step 12)

23. 15(3) cage at R6C6 (step 20) = {168/258/267/357}
23a. Hidden killer pair 5,6 in R2C6 and R67C6 for C6 -> R2C6 = {56}, clean-up: no 4 in R4C9 (step 7)

24. 13(3) cage at R3C8 (step 5) = {157/256}, CPE no 5 in R12C9

25. R4C9 + R6C79 (step 17) = {168/267/357}
25a. Killer pair 5,6 in R45C7 and R4C9 + R6C79, locked for N6

26. 45 rule on N2 3 innies R2C46 + R3C4 = 13 = {157/256}, no 3, 5 locked for N2, clean-up: no 6 in R4C1 (step 6)
26a. 6 in {256} must be in R2C6 (R23C4 cannot be {26} which clashes with R56C4) -> no 6 in R23C4, clean-up: no 3 in R4C1 (step 6)
26b. Killer pair 1,2 in R23C4 and R56C4, locked for C4
26c. 5 in C5 locked in R789C5, locked for N8

27. R4C13 + R6C1 (step 10) = {128/245} (cannot be {137} because R6C1 only contains 2,4, cannot be {146/236} because 1,3,6 only in R4C3), no 3,6,7, 2 locked for N4, clean-up: no 2 in R2C4 (step 6), no 7 in R56C3
27a. 1,5 must be in R4C3 -> R4C3 = {15}
27b. 2 locked in R46C1, locked for C1

28. 14(3) cage at R3C4 = {158/167} (cannot be {257} because 2,7 only in R3C4), no 2,9, clean-up: no 5 in R6C6 (step 12)

29. Naked pair {68} in R4C4 + R6C6, locked for N5 and D\, clean-up: no 4 in R46C5 (step 13a), no 1 in R56C4

30. Naked pair {25} in R56C4, locked for C4 and N5, clean-up: no 4 in R4C1 (step 6), no 3 in R45C6

31. Naked pair {14} in R45C6, locked for C6 -> R8C6 = 2, R7C7 = 1, locked for D\, R7C3 = 2, locked for D/ -> R56C4 = [25], 5 locked for D/, R6C9 = 7 (step 9), clean-up: no 4 in R5C3, no 7 in R19C6 (step 11) -> R1C6 = 3

32. Naked pair {89} in R39C6, locked for C6 -> R67C6 = [67], R6C7 = 2 (step 23), R2C6 = 5, R4C9 = 6 (step 7), R4C4 = 8, R3C4 = 1 (step 28), R4C3 = 5, R2C4 = 7, R7C45 = [94], R9C6 = 8, R3C56 = [89], R89C4 = [63], R1C4 = 4, R1C1 = 5, locked for D\, R46C1 = [24], R45C7 = [95], R8C78 = [79], 9 locked for D\ -> R5C5 = 7, 7 locked for D/, R46C5 = [39], clean-up: no 3 in R5C3
32a. R7C12 = 11 = {38}/[65], no 6 in R7C2

33. Naked pair {67} in R3C12, locked for R3 and N1
33a. Naked pair {25} in R3C89, locked for N3

34. R7C3 + R8C4 = 8 -> R8C23 = 9 = {18}, locked for R8 and N7

and the rest is naked singles, remembering eliminations along the diagonals

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

I'll have a try at the "Hard" version once I've caught up with a couple of other walkthroughs. A fair number of the steps used for "Easy" can still be used because of the way the puzzle wraps itself around N5.
Last edited by Andrew on Mon Mar 24, 2008 5:21 am, edited 1 time in total.
mhparker
Grandmaster
Grandmaster
Posts: 345
Joined: Sat Jan 20, 2007 10:47 pm
Location: Germany

Bored 89 Piece of Cake

Post by mhparker »

Hi folks,
I wrote:So far, I agree with Afmob's description: lots of methodical work picking off individual candidates, but no apparent opportunity to by-pass this with any "blockbuster" type of move.
Well, finally finished it. :D

I was actually very close to the breakthrough (step 14) when I stopped after the last session. I now retract my comment above regarding picking off of individual candidates.

All in all, this was a nice puzzle, which looked impossible at the beginning, due to the huge number of candidates still left after the (relatively few) preliminaries. There were also some good X-moves in there. The difficulty level was just right for me. Many thanks to Nasenbaer for a great little puzzle! :D

As for the rating, I'll go along with Andrew's 1.25.

Edit: Thanks to Andrew for minor addition to step 18b!

Bored 89 Easy Walkthrough

Prelims:

a) 17(2) at R3C5 = {89}, locked for R3 and N2
b) 19(3) at R4C5 = {289/379/469/478/568} (no 1) = {(8/9)..}
c) 5(2) at R4C6 = {14/23} = {(3/4)..}
d) 14(2) at R4C7 = {59/68}
e) 9(2) at R5C3 = {18/27/36/45} (no 9)
f) 7(2) at R5C4 = {16/25} = {(2/6)..}
(Note: {34} blocked by 5(2) at R4C6 (prelim c))
g) 13(2) at R7C4 = {49/58/67} (no 1..3)

1. 19(3) at R4C5 (prelim. b) and R3C5 form killer pair on {89} within C5
1a. -> no 8,9 elsewhere in C5
1b. cleanup: no 4,5 in R7C4

2. Innies N5: R4C4+R6C6 = 14(2) = {59/68} = {(8/9)..}

3. 7 in N5 locked in 19(3) at R4C5 (prelim b) = {379/478} (no 2,5,6) = {(3/4)..}
3a. 7 locked in R456C5 for C5
3b. cleanup: no 6 in R7C4

4. Outies R89: R7C37 = 3(2) = {12}, locked for R7
4a. no 1,2 in R3C37 (CPE)

5. Outies R12: R3C37 = 7(2) = {34} (last combo), locked for R3
5a. no 3,4 in R5C5 (CPE)

6. 13(3) at R3C8 = {157/247/256} (no 3,8,9)
(Note: {139/148/238/346} unplaceable)

7. Outies N3: R2C6+R4C9 = 11(1+1) = {47/56} (no 1..3)

8. Innies C6789: R139C6 = 20(3) = {389/479/569/578} (no 1,2)
8a. 3 of {389} must go in R1C6
8b. -> no 3 in R9C6

9. Hidden killer pair on {12} in C6 at R45C6 and R8C6
9a. -> R8C6 = {12}

10. 19(4) at R7C7 = {1279} (last combo)
10a. -> R8C78 = {79}, locked for R8 and N9
10b. no 1,2 in R8C9 (CPE)

11. R8C8, R5C5 and R4C4+R6C6 (step 2) form killer triple on {789} in D\
11a. -> no 7..9 elsewhere in D\
11b. 8 locked in R4C4+R5C5+R6C6 for N5

12. Outies N9: R6C9+R8C6 = 9(2)
12a. -> R6C9 = {78}

13. Innies N6: R4C9+R6C79 = 15(3) = {168/258/267/348/357}
(Note: {159/249/456} unplaceable)
13a. must have 1 of {123}, only available in R6C7
13b. -> R6C7 = {123}
13c. can only have 1 of {78}, which must go in R6C9
13d. -> no 7 in R4C9
13e. cleanup: no 4 in R2C6 (step 7)

14. R2C3 and 17(4) at R1C1 form hidden killer pair on {89} within N1
14a. -> R2C3 = {89} (no 1..7)
14b. 17(4) at R1C1 = {(8/9)..} = {1259/1268/1358} (no 4,7)
(Note: {1349/2348} both blocked by R3C3)

15. 7 in N1 locked in R3C12 for R3 and 15(3)
15a. 15(3) at R3C1 = {7..} = {267/357} (no 1,4,8,9)
15b. 3 only available in R4C1
15c. -> no 5 in R4C1

16. Outies N1: R2C4+R4C1 = 9(1+1)
16a. -> R2C4 = {367} (no 1,2,4,5)

17. 13(3) at R3C8 (step 6) = {256} (no 1,4) (last combo)
17a. 2 locked in R3C89 for R3 and N3
17b. cleanup: no 7 in R2C6

18. Innies N2: R2C46+R3C4 = 13(3) = [751] (last permutation)
18a. -> R4C19 = [26] (steps 7,16)
18b. cleanup: no 3 in R5C6; no 5 in R3C12 (step 15a); no 6 in R56C4; no 6 in R7C5; no 8 in R45C7; no 7 in R56C3

19. 1 in N5 locked in 5(2) at R4C6 = {14}, locked for C6 and N5

--- NOTE: cleanups omitted from here on for simplicity ---

20. Naked single at R8C6 = 2
20a. -> R7C7 = 1; R7C3 = 2; R6C9 = 7 (outie N9)
20b. -> R6C7 = 2 (innie N6); R56C4 = [25]

21. Naked pair at R45C7 = {59}, locked for C7 and N6
21a. -> R8C78 = [79]
21b. -> Split 13(2) at R4C34 = [58]
21c. -> R5C5 = 7; R67C6 = [67]; R45C7 = [95]; R7C45 = [94]
21d. -> split 12(2) at R46C5 = [39]; R19C6 = [38]
21e. -> R3C56 = [89]; R1C1 = 5

22. Innie N8: R8C4 = 6

Rest is now just singles and simple cage sums.
Cheers,
Mike
sudokuEd
Grandmaster
Grandmaster
Posts: 257
Joined: Mon Jun 19, 2006 11:06 am
Location: Sydney Australia

Post by sudokuEd »

About Bored89-Easy Andrew wrote:It is easy but only after one finds the key move
Don't agree with 'easy' even with your step 3a. I managed to see that one early but still found it a really good tussle.

Working backwards through my solution I felt that removing 4 from r2c6 was the key move. I did it through innies n2, Andrew through a very nice use of i/o on c6 (which I missed) (his hidden step 16a) and Mike through h(cage) placement from innies of n6 taking out 7 from r4c9 (which I missed) (his step 13c,d).

Sudoku Solver ignores these ways completely and instead finds a really useful IOU in i/o of c1234 -> no 3 in r1c4. This soon leads to a hidden single in r9c4.

Fortunately, all five of us agree on the rating of 1.25 (SSscore of 1.30 with current unpublished version of SudokuSolver. Only 2 decimal places :D ). So Afmob was correct
Afmob wrote:Since A89 is of rating 1.25.. I think that A89 V1.5 (ie Bored89-Easy) is as hard as A89 V1
Nice prediction Afmob! Very enjoyable puzzle Nasenbaer - many ways to skin the 1,25 bear :wink: !

Now can start work on updating the sticky since A89 is finally put to rest.

Cheers
Nasenbaer
Master
Master
Posts: 167
Joined: Thu Jul 20, 2006 7:19 am
Location: Fellbach (Deutschland)

Post by Nasenbaer »

It seems that all of you had some fun solving these puzzles. :) And some great work with these walkthroughs! They were a lot better than what I had. Congratulations! =D>

With the new version of SudokuSolver (v2.2.3) the rating dropped a little: 1.347 for the easy version, 1.729 for the hard one.

BTW, I didn't name it Assassin 89v2 because
a) it wasn't based on the original Assassin 89,
b) I didn't think it had the quality to be named Assassin, and
c) I think Assassin should be reserved for Ruud's killers. But that's just my opinion.

Cheers,
Nasenbaer
Andrew
Grandmaster
Grandmaster
Posts: 300
Joined: Fri Aug 11, 2006 4:48 am
Location: Lethbridge, Alberta

Post by Andrew »

Good to know that you enjoyed the posted walkthroughs.

I'm hoping to have a go at Hard once I've caught up on a couple of things in my "backlog".
Nasenbaer wrote:BTW, I didn't name it Assassin 89v2 because
a) it wasn't based on the original Assassin 89,
b) I didn't think it had the quality to be named Assassin, and
c) I think Assassin should be reserved for Ruud's killers. But that's just my opinion.
Interesting reasons. I agree completely with a), variants using the Assassin name followed by a variant number should be either the same cage pattern with different totals and solution or slightly modified cage pattern which normally takes away something easy at the beginning.

You will be pleased to know that I disagree with b). Easy was definitely of Assassin quality and I'm sure that Hard will be even more so.

If a puzzle meets my comments about a), then in my opinion it's valid to call it an Assassin variant. Of course Ruud is the only one who should post original Assassins and long may he continue to do so.
mhparker
Grandmaster
Grandmaster
Posts: 345
Joined: Sat Jan 20, 2007 10:47 pm
Location: Germany

Post by mhparker »

Nasenbaer wrote:BTW, I didn't name it Assassin 89v2 because
...
c) I think Assassin should be reserved for Ruud's killers...
...or (as pointed out by Andrew) for variants closely related to the corresponding original Assassin. But I think this is what you mean anyway.

An alternative would have been to call it a Maverick (e.g., in this case, it would have been the M3). The idea is that anyone can add to this "series", not just me (which is not a bad idea, considering that I've only come up with 2 of them so far! :)).

BTW, congratulations to Andrew on reaching Grandmaster status. :salut:
Cheers,
Mike
Andrew
Grandmaster
Grandmaster
Posts: 300
Joined: Fri Aug 11, 2006 4:48 am
Location: Lethbridge, Alberta

Post by Andrew »

I solved Bored89-Hard last week but only went through Afmob's walkthrough yesterday.

A really challenging puzzle. Thanks Nasenbaer! Changing the cages in N5 took away the useful 5(2) cage in C6 making it a lot harder than the Easy version.

Afmob used some heavy combination analysis. I must admit I couldn't follow the details of some of the sub-steps since I work with an Excel worksheet and Ruud's combination calculator. They are probably easier to see and understand by those using software solvers such as SumoCue and SS.

Some of the key solving involves 4 outies, particularly in C4 and then R6. After that, as Ed pointed out in his discussion of the Easy version, the elimination of 4 from R2C6 is a key move. In my solution for Hard, this came from an easy 45 test, step 32, after the harder work had been done.

My methodical combination and permutation analysis in steps 21, 25 and 29 may at first sight look complicated but it only involved simple clashes. I've given summaries after those three steps for anyone who may want to skip the details.

Since my combination analysis was only moderately heavy, I'll rate this puzzle as Hard 1.5


Here is my walkthrough

This is a Killer-X. I've included eliminations along diagonals, for those not using software to do eliminations, because they are easy to overlook.

Prelims

a) R3C56 = {89}, locked for R3 and N2
b) R45C7 = {59/68}
c) R56C3 = {18/27/36/45}, no 9
d) R7C45 = {49/58/67}, no 1,2,3

1. 45 rule on R12 2 outies R3C37 = 7 = {16/25/34}, no 7

2. 45 rule on R89 2 outies R7C37 = 3 = {12}, locked for R7
2a. CPE no 1,2 in R3C37 and R5C5, clean-up: no 5,6 in R3C37 (step 1)
2b. Naked pair {34} in R3C37, locked for R3
2c. CPE no 3,4 in R5C5

3. 15(4) cage at R4C5 = {1239/1248/1257/1347/1356/2346}
3a. 8,9 of {1239/1248} must be in R5C5 -> no 8,9 in R4C56 + R5C6

4. 45 rule on N1 2 outies R2C4 + R4C1 = 9 = [18]/{27/36/45}, no 1,9 in R4C1

5. 45 rule on N3 2 outies R2C6 + R4C9 = 11 = [29/38]/{47/56}, no 1, no 2,3 in R4C9

6. 45 rule on N7 2 outies R6C1 + R8C4 = 10 = {19/28/37/46}/[55]

7. 45 rule on N9 2 outies R6C9 + R8C6 = 9 = {18/27/36/45}, no 9

8. 45 rule on N4 3 innies R4C13 + R6C1 = 11 = {128/137/146/236/245}, no 9, clean-up: no 1 in R8C4 (step 6)

9. 45 rule on N5 2 innies R4C4 + R6C6 = 14 = {59/68}

10. 13(3) cage at R3C8 = {157/247/256} (cannot be {148} because 4,8 only in R4C9), no 8,9, clean-up: no 2,3 in R2C6 (step 5)

11. 15(3) cage at R3C1 = {168/258/267/357/456} (cannot be {348} because 3,4,8 only in R4C1)
11a. 3,4,8 of {258/357/456} must be in R4C1 -> no 5 in R4C1, clean-up: no 4 in R2C4 (step 4)

12. Hidden killer pair 8,9 in N1, 17(4) cage cannot contain both of 8,9 -> R2C23 must contain at least one of 8,9
12a. 22(4) cage at R2C2 = {1489/2389/2479/3469/3478/3568} (cannot be {1579/1678/2569/2578} because R3C3 only contains 3,4, cannot be {4567} which doesn’t contain 8,9)
12b. 1 of {1489} must be in R2C4 -> no 1 in R2C23

13. 14(3) cage at R3C4 = {149/158/167/239/248/257/356} (cannot be {347} because no 3,4,7 in R4C4)
13a. 6 of {167} must be in R4C4, 3 of {356} must be in R4C3 -> no 6 in R4C3

14. 15(3) cage at R6C6 = {159/168/249/258/267/348/357/456}
14a. 1,2 of {159/249} must be in R6C7 -> no 9 in R6C7
14b. 2 of {258} must be in R6C7, 8 of {348} must be in R6C6 -> no 8 in R6C7

15. 45 rule on N689 2 innies R4C9 + R8C4 = 1 outie R6C6 + 6, min R6C6 = 5 -> min R4C9 + R8C4 = 11, no 2,3 in R8C4, clean-up: no 7,8 in R6C1 (step 6)

16. 45 rule on N69 2 innies R4C9 + R6C7 = 1 outie R8C6 + 6, max R4C9 + R6C7 = 13 -> max R8C6 = 7, clean-up: no 1 in R6C9 (step 7)

17. 45 rule on N6 3 innies R4C9 + R6C79 = 15 = {168/267/348/357} (cannot be {258/456} which clash with R45C7)
17a. 4 of {348} must be in R4C9 -> no 4 in R6C79, clean-up: no 5 in R8C6 (step 7)
[At this stage I noticed that R4C9 + R6C79 and 15(3) cage at R6C6 cannot both be {348} because of clash between R6C9 and R6C6 but either one can be {348} if the other one is {357}.]

18. 16(4) cage in N6 = {1249/1267/1348/2347} (cannot be {1258/1456/2356} which clash with R45C7, cannot be {1357} which clashes with R4C9 + R6C79), no 5

19. 45 rule on C789 4 outies R2678C6 = 20 = {1469/1478/1568/2459/2468/2567/3458/3467} (cannot be {1289} which clashes with R3C6, cannot be {1379/2369} because 15(3) cage at R6C6 cannot be = [933], cannot be {2378} because no 4 in R6C7 to make combination with R67C6 = [83])
19a. 1,2 of {1469/1568/2468/2567} must be in R8C6, 6 of {3467} must be in R6C6 -> no 6 in R8C6, clean-up: no 3 in R6C9 (step 7)

20. R4C9 + R6C79 (step 17) = {168/267/348/357}
20a. 3 of {357} must be in R6C7 -> no 5 in R6C7

Consider the combinations of step 19 in more detail. Most of the following could have been done in step 19 but one depends on step 20 so it’s all together here.

21. 45 rule on N78 2 innies R78C6 = 1 outie R6C1 + 5, min R78C6 = 6
21a. R2678C6 (step 19) = {1469/1478/1568/2459/2468/2567/3458/3467}
21b. {1469} cannot be 4{69}1 because R67C6 must total less than 15, cannot be [6941] because min R78C6 = 6 -> cannot be {1469}
21c. {1478} cannot be 4{78}1 because R67C6 must total less than 15, cannot be [7841] because min R78C6 = 6 -> cannot be {1478}
21d. {1568} = 5{68}1 / 6{58}1
21e. {2459} = 4{59}2 / [5942]
21f. {2468} = 4{68}2 / [6842]
21g. {2567} = [5672/6572] (cannot be 7{56}2 because no 4 in R6C7)
21h. {3458} = 4{58}3 / [5843] (cannot be [5834] because no 4 in R6C7)
21i. 6 of {3467} must be in R6C6 = [4673] (cannot be [7643] because no 5 in R6C7, cannot be [7634] because 15(3) cage at R6C6 cannot be [663])

Summary of step 21, R2C678C6 = {1568/2459/2468/2567/3458/3467}, R2C6 = {456}, R67C6 = {58/59/68}/[57/67/84/94], no 3, R8C6 = {123}, clean-up: no 4 in R4C9 (step 5), no 2,5 in R6C9 (step 7)

22. 15(3) cage at R6C6 (step 14) = {159/168/249/258/267/348/357} (cannot be {456} because R67C6 cannot be {45} from combinations in step 21)
22a. 1,2,3 only in R6C7 -> R6C7 = {123}

23. R4C9 + R6C79 (step 17) = {168/267/357}
23a. {168} must be [618] -> R67C6 (summary of step 21) = {58/59}/[57/67/68/84/94] (cannot be [86] which clashes with R6C9), no 6 in R7C6

24. Max R7C7 + R8C6 = 5 -> min R8C78 = 14, no 1,2,3,4
24a. 19(4) cage at R7C7 = {1279/1369/1378/2359/2368}

25. R4C9 + R6C79 (step 23) = {168/267/357} -> R6C79 = [18/26/27/37]
25a. 45 rule on R789 4 outies R6C1679 = 19 = {1279/1378/1468/2368/2467/3457} (cannot be {1459/2359} because R6C9 only contains 6,7,8, cannot be {1369/1567/2458} which don’t match with R6C79)
25b. {1279} = [1927]
25c. {1378} = [1837]
25d. {1468} = [4618]
25e. {2368} = [3826]
25f. {2467} = [4627]
25g. {3457} = [4537]

Summary of step 25, R6C1 = {134}, R6C67 = [53/61/62/82/83/92], clean-up: no 4,5,8 in R8C4 (step 6)

26. R6C67 (step 25) = [53/61/62/82/83/92] -> R7C6 = {4578}, no 9

27. R4C13 + R6C1 (step 8) = {128/137/146/236/245}
27a. 6 of {146} must be in R4C1, 4 of {245} must be in R6C1 -> no 4 in R4C1, clean-up: no 5 in R2C4 (step 4)

28. 13(3) cage at R3C8 (step 10) = {157/256}, CPE no 5 in R12C9

29. 45 rule on C123 4 outies R2348C4 = 22 = {1579/1678/2389/2569/2578/3568}
29a. {1579} = [1759/7159]
29b. {1678} = [6187/7186]
29c. {2389} = [3289]
29d. {2569} = [2569/2659]
29e. {2578} = [2587]
29f. {3568} = [3586]

Summary of step 29 R34C4 = [15/18/28/56/58/65/75], no 9, clean-up: no 5 in R6C6 (step 9)

30. R34C4 (step 29) = [15/18/28/56/58/65/75] -> R4C3 = {123458}, no 7

31. 17(4) cage at R7C3 = {1259/1268/1349/1367/1457/2357/2456} (cannot be {1358/2348} because R8C4 only contains 6,7,9)
31a. 9 of {1259/1349) must be in R8C4 -> no 9 in R8C23

32. 45 rule on N2 3 innies R2C46 + R3C4 = 13 = {157/256} (cannot be {247} because R23C4 cannot be {27}, cannot be {346} because R23C4 cannot be [36], both from step 29), no 3,4, 5 locked for N2, clean-up: no 2 in R3C4 (step 29c), no 6 in R4C1 (step 4), no 7 in R4C9 (step 5)
32a. 2 of {256} must be in R2C4 -> no 6 in R2C4, clean-up: no 3 in R4C1 (step 4)

33. Killer pair 5,6 in R45C7 and R4C9, locked for N6, clean-up: no 3 in R6C1 (step 25e), no 7 in R8C4 (step 6), no 3 in R8C6 (step 7)

34. 19(4) cage at R7C7 (step 24a) = {1279} (only remaining combination)
34a. Naked pair {79} in R8C78, locked for R8 and N9 -> R8C4 = 6, R6C1 = 4 (step 6), clean-up: no 5 in R56C3, no 8 in R6C6 (step 9), no 7 in R7C45

35. R6C1 = 4 -> R7C12 = 11 = {38/56}, no 7,9
35a. R7C45 = {49} (cannot be {58} which clashes with R7C12), locked for R7 and N8

36. R7C6 = 7 (hidden single in R7), R6C67 = 8 = [62], R4C4 = 8 (step 9), 6,8 locked for D\, R2C6 = 5, R4C9 = 6 (step 5), R7C7 = 1, locked for D\, R8C6 = 2, R7C3 = 2, locked for D/, R6C9 = 7 (step 7), clean-up: no 1 in R2C4 (step 4), no 3,7 in R5C3, no 8 in R5C7
[While checking my walkthrough I noticed that the clean-up makes R2C4 + R4C1 into naked pair {27} -> CPE no 2,7 in R2C1. These are both eliminated in the next step.]

37. R2C6 = 5 -> R23C4 = [71] (step 32), R4C1 = 2 (step 4), R4C3 = 5 (step 27), R45C7 = [95], R8C78 = [79], 9 locked for D\, R5C5 = 7, locked for D\ and D/

38. R3C6 = 9 (hidden single in C6), R3C5 = 8, R9C6 = 8 (hidden single in C6)
38a. 1 in C6 locked in R45C6, locked for N5

39. R5C4 = 2 (hidden single in R5), R6C45 = 14 = {59}, locked for R6

40. R4C9 = 6 -> R3C89 = {25} (step 28), locked for R3 and N3

41. R4C2 = 7 (hidden single in R4) -> R3C12 = [76]
41a. R9C3 = 7 (hidden single in C3)

42. Naked pair {34} in R1C46, locked for R1 and N2 -> R1C1 = 5, locked for D\

43. R7C3 = 2, R8C4 = 6 -> 17(4) cage at R7C3 (step 31) = {1268/2456}, no 3
43a. 4 of {2456} must be in R8C3 -> no 4 in R8C2

44. R5C3 = 6 (hidden single in C3), R6C3 = 3, R3C3 = 4, locked for D\, R3C7 = 3, locked for D/

45. R2C6 = 5, R3C7 = 3 -> R2C78 = 10 = {46}, locked for R2 and N3 -> R12C5 = [62], R1C7 = 8, R2C2 = 3, locked for D\, R9C9 = 2, R2C3 = 8 (step 12a), R8C3 = 1, R1C3 = 9, R12C9 = [19], 1 locked for D/, R1C2 = 2, R1C8 = 7, R2C1 = 1, R4C56 = [34], 4 locked for D/

and the rest is naked singles

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