HATMAN wrote:
Interesting box interactions
This is still a
HUS puzzle for me. Here are examples of the interesting interactions I found of which there are other eliminations but nothing to crack the puzzle. Hope someone else can see something. If we can't crack it, perhaps
HATMAN can give us a hint after the weekend.
HS8 Boxesprelims
i. 14(4) at r1c3, r1c6, r6c8 & r8c2: no 9
ii. 26(4) at r2c8, r3c1, r5c5, r6c6, r7c1 & r8c6: no 1
1. "45" on n5: 25(4)+26(4): 1 overlap cell r5c5 - 6 = 2 innies r4c6 + r6c4
1a. -> r5c5 = 9
1b. r4c6+r6c4 = {12}: both locked for n5
1c. 22(4)r5c3 must have 1/2 for r6c4 -> no 1,2 elsewhere in 22(4)
1d.
deleted wrong step: thanks Andrew2. "45" on n78: 26(4)+16(4) -> 10 innies = 48
2a. "45" on c34: 14(4)+18(4)+22(4) -> 6 innies = 36
2b. step 2 - step 2a -> 4 outies r9c1256 = 12 = {1236/1245}
2c. 1 and 2 both locked for r9
3. "45" on r56: 6 innies r56c789 = 27
3a. -> "45" on n6: 3 remaining innies r4c789 = 18 (no eliminations yet)
3b. "45" on n3: 5 innies = 31
3c. adding steps 3a + 3b = 49
3d. then subtracting the 21(4)r3c7 -> 4 remaining outies r1234c9 = 28
3e. h28(4) = {4789/5689}(no 1,2,3): 8 and 9 locked for c9
4. From step 3e, 5 remaining innies c9 r56789c9 = 17
4a. "45" on n9: 5 innies = 27
4b. -> from step 4 and 4a, r9c78 - 10 = r56c9
4c. -> min. r9c78 = 13 (no 3)
4d. and max. r56c9 = 7 (no 7)
5. deleted: thanks to Andrew for picking up a flaw and showing how to rework the next couple.6. 9 in c6 is in
r123c6 or 16(4)n8
6a. however, can't be in 16(4): like this.
6b. 26(4)
at r5c5 must have 9 = 9{368/458/467} = at least one of 4/6 in c56
6c. the only combo in 25(4)n2 without 9 is {4678} = both 4&6
6d. when 16(4)n8 has 9 it must be {1249} = one of 4/6
6e. ie, all 4 & 6 taken for c56
6f. but this forces 16(4)n8 to clash with h12(4)r9 since r9c56 can only be {35}
6g. -> no 9 in r78c6
7. 9 in c6 only in
r123c6: locked for n2
8. "45" on n7: 5 innies = 19
8a. 3 of those cells r8c3 + r9c23 overlap the 14(4)n7 -> r7c3 + r9c1 - 5 = r8c2
8b. -> no 5 in r7c3 nor r9c1 (IOU)
[Andrew pointed out a much simpler way to do the previous step;
“45” on n7 26(4)+14(4) 1 overlap cell r8c2 + 5 = 2 innies r7c3 + r9c1 -> no 5 in ... (IOU).
Nice!]Perhaps the same sort of IOU step is available when 3 cages overlap. Just can't see it!
Cheers
Ed