The New Sudoku Players' Forum

[ArkieTech]

Code: Select all

 *-----------*
 |...|..4|..3|
 |..1|.5.|...|
 |.3.|.62|.7.|
 |---+---+---|
 |.7.|..3|2..|
 |39.|.7.|.48|
 |..6|2..|.5.|
 |---+---+---|
 |.5.|48.|.6.|
 |...|.3.|7..|
 |6..|5..|...|
 *-----------*


Play/Print this puzzle online

[SpAce]

Sigh. Where's my spine? Simply too nice a puzzle to pass.

Code: Select all

.-----------------------.-------------------.-----------------------.
|   2589    268    589  |   7    d19    4   |  15689  1289    3     |
|   7       2468   1    |   3     5     89  |  4689   289     2469  |
|   4589    3      4589 | cd189   6     2   | d14589  7     e(1)49  |
:-----------------------+-------------------+-----------------------:
| e(1)458   7      458  |   168   14    3   |  2      19      169   |
|   3       9      2    |  c16    7     5   | d16     4       8     |
|  d148    d148    6    |   2     149  d189 |  3      5       7     |
:-----------------------+-------------------+-----------------------:
| a[29]-1   5      3    |   4     8     7   |  19     6     a[29]-1 |
|  b12489   1248  b489  |  b19    3     6   |  7      1289    5     |
|   6       18     7    |   5     2    c19  |  1489   3       149   |
'-----------------------'-------------------'-----------------------'

Almost Hidden Pair:

(29)r7c19 = (9,1)r8c134 - r35c4|r9c6 = r1c5&r5c7&r6c6 - r3c47|r6c12 = (1)r3c9&(1)r4c1 => -1 r7c19; stte

Btw, congrats to anyone who can spot the real short solution without computer help. I surely couldn't.

the short solution: Show

1:r35b9\r8c479 => -1 r8c4; stte

Looks like no human solver collected the prize for that.

Edit: added the short solution (found by Hodoku, not me).

[Cenoman]

Code: Select all

 +------------------------+--------------------+------------------------+
 |  2589    268    589    |  7    b19    4     |  15689  a1289   3      |
 |  7       2468   1      |  3     5     89    |  4689    289    2469   |
 |  4589    3      4589   | c189   6     2     |  14589   7      149    |
 +------------------------+--------------------+------------------------+
 | B1458    7      458    |  168   14    3     |  2      A19     169    |
 |  3       9      2      |  16    7     5     |  16      4      8      |
 | C148    C148    6      |  2     149  D189   |  3       5      7      |
 +------------------------+--------------------+------------------------+
 |  129     5      3      |  4     8     7     |  19      6      129    |
 |  12489   1248   489    |  9-1   3     6     |  7      z1289   5      |
 |  6       18     7      |  5     2    E19    |  1489    3      149    |
 +------------------------+--------------------+------------------------+

Single digit kraken (almost turbot-fish)
Kraken column (1)r148c8
(1)r1c8-r1c5=(1)r3c4
(1)r4c8-r4c1=r6c12-r6c6=(1)r9c6
(1)r8c8
=> -1 r8c4; ste

[TF(1)r8c8*=r1c8-r1c5=r3c4] = (1)r4c8-r4c1=r6c12-r6c6=(1)r9c6 => -1 r8c4; ste

[SteveG48]

Code: Select all

 *----------------------------------------------------------------------*
 | 2589   268    589    |  7     b19     4      | c15689 c1289   3      |
 | 7      2468   1      |  3      5      89     |  4689   289    2469   |
 | 4589   3      4589   |ah189    6      2      | c14589  7     d149    |
 *----------------------+-----------------------+-----------------------|
 | 1458   7      458    |ah168    14     3      |  2      19     169    |
 | 3      9      2      |ah16     7      5      |bg16     4      8      |
 | 148    148    6      |  2      149    189    |  3      5      7      |
 *----------------------+-----------------------+-----------------------|
 | 129    5      3      |  4      8      7      | f19     6     e129    |
 | 12489  1248   489    | g9-1    3      6      |  7     f1289   5      |
 | 6      18     7      |  5      2      19     | f1489   3     e149    |
 *----------------------------------------------------------------------*


1r345c4 = 1r1c5&r5c7 - b3p127 = b3p9 - b9p39 = b9p17|b9p5 - 1r5c6&r8c4 = 1r345c4 => -1 r8c4 ; stte

[eleven]

Code: Select all

 *----------------------------------------------------------------------*
 |  2589    268    589    |  7     19    4     |  5689-1  1289   3      |
 |  7       2468   1      |  3     5     89    |  4689    289    2469   |
 |  4589    3      4589   |  89-1  6     2     |  4589-1  7      149    |
 |------------------------+--------------------+------------------------|
 |  1458    7      458    |  68-1  14    3     |  2       19     169    |
 |  3       9      2      | a16    7     5     | b16      4      8      |
 |  148     148    6      |  2     149   189   |  3       5      7      |
 |------------------------+--------------------+------------------------|
 |  129     5      3      |  4     8     7     | c19      6      12-9   |
 |  12489   1248   489    | f19    3     6     |  7       28-19  5      |
 |  6       18     7      |  5     2    e19    | d489-1   3     d149    |
 *----------------------------------------------------------------------*

1r5c4 = r5c7 - (1=9)r7c7 - r9c79 = r9c6 - (9=1)r8c4, loop => -1r34c4, -1r139c7, -19r8c8, -9r7c9; stte
(either 1 is in 1r5c4 & 1r7c7 or in 1r5c7 & 1r8c4, either 9 in r9c79 or in r7c7)

[SpAce]

1r5c4 = r5c7 - (1=9)r7c7 - r9c79 = r9c6 - (9=1)r8c4, loop => -1r34c4, -1r139c7, -19r8c8, -9r7c9; stte
(either 1 is in 1r5c4 & 1r7c7 or in 1r5c7 & 1r8c4, either 9 in r9c79 or in r7c7)

Very clever! I think I saw that loop but dismissed it as not having stte-potential. I didn't see the crucial -1r8c8, as it's not a normal loop elimination with adjacent end-points but a result of a subchain (with end-points 1r7c7 and 1r8c4). It's perfectly acceptable, especially with your explanation, but it'd be clearer if it were marked on the chain somehow. Personally I use '@' to mark embedded end-points (left or right side of the node, depending on which way the subchain goes). For example, here it would be on the left side because the subchain goes right:

1r5c4 = r5c7 - @(1=9)r7c7 - r9c79 = r9c6 - (9=1)r8c4, loop

For better or worse (probably worse), I used those a lot here (steps 5,6,9) to avoid several chains with mostly the same structure. It got a bit out of hand, I admit, but it's always fun to get several seemingly unrelated eliminations with a single chain.

[Leren]

Code: Select all

*---------------------------------------------------*
| 2589  268  589  | 7   19  4   | 15689  1289  3    |
| 7     2468 1    | 3   5   89  | 4689   289   2469 |
| 4589  3    4589 |*189 6   2   |*14589  7    *149  |
|-----------------+-------------+-------------------|
| 1458  7    458  | 168 14  3   | 2      19    169  |
| 3     9    2    |*16  7   5   |*16     4     8    |
| 148   148  6    | 2   149 189 | 3      5     7    |
|-----------------+-------------+-------------------|
| 129   5    3    | 4   8   7   |*19     6    *129  |
| 12489 1248 489  | 9-1 3   6   | 7     f1289  5    |
| 6     18   7    | 5   2   19  |*1489   3    *149  |
*---------------------------------------------------*

Finned Franken Swordfish in 1's r35b9 c479 with a fin r8c8 => - 1 r8c4; stte

Leren

[eleven]

I didn't see the crucial -1r8c8, as it's not a normal loop elimination with adjacent end-points but a result of a subchain (with end-points 1r7c7 and 1r8c4). It's perfectly acceptable, especially with your explanation, but it'd be clearer if it were marked on the chain somehow. Personally I use '@' to mark embedded end-points (left or right side of the node, depending on which way the subchain goes).

I see it, as i wrote and explained it. I don't think, that extra symbols would make it clearer, as long as there is no official definition for them.
You are invited to start a sticky thread for that, but it seems, that you have other plans.

[SpAce]

I see it, as i wrote and explained it.

I understand, and it's of course a valid way to see and explain it. Instead of a subchain it can be seen as a parity thing, but even then I think it would benefit from a corresponding notation:

1'r5c4 = 1"r5c7 - (1'=9")r7c7 - 9'r9c79 = 9"r9c6 - (9'=1")r8c4, loop

...or more likely (and clearly) just the relevant bit:

1r5c4 = r5c7 - (1'=9)r7c7 - r9c79 = r9c6 - (9=1")r8c4, loop

Hmm. That might actually be a better way to mark it?

Anyway, your explanation was good, but it wasn't immediately clear to me why one was needed, as at first it seemed like a normal loop with normal loop eliminations -- as it was except for the one. Either mentioning it explicitly, or marking the relevant node(s) or the special elimination would have drawn attention to why it needed an explanation. I probably would have missed the significance if I hadn't considered that loop myself and wondered how you got it to work as stte. (Then again, I'm one who actually likes these kinds of side-puzzles, so I'm not really even complaining for myself!)

I don't think, that extra symbols would make it clearer, as long as there is no official definition for them. You are invited to start a sticky thread for that, but it seems, that you have other plans.

Well, to me it's just important that there's some clue (and you did have that). At this point it's probably hard to push any "official" standards for anything, as people already use various markings for the same things and probably like their own best. This particular situation is pretty rare anyway, so I think ad hoc notations/explanations are to be expected and accepted. The worst option is to have nothing to hint that some eliminations need a subchain. Without any hint I'd only look at the chain end-points or adjacent nodes in a loop, and would reject any rogue eliminations. (And yes, I'm still planning to take a break.)