## The Sudoku Gremlin is Back!

Post the puzzle or solving technique that's causing you trouble and someone will help

### The Sudoku Gremlin is Back!

[ Edit by Pappocom: In case someone should think that this Topic is a complaint about a bug of some kind in the Sudoku program by Pappocom, let me assure you it is not! ]

Last year a <sudoku gremlin> was being a minor nuisance by erasing all the single digits in my candidate grids ,
but now it's worse -- The gremlin is erasing all the single digits and also all the naked pairs (if there are any)!
Here's the latest vandalised grid:
Code: Select all
`        347           | 3456   46            |        3567   37         3478   468    | 1346          1346   |        3678   378 268    38     268    | 3569          3569   | 568               ---------------------+----------------------+------------------ 1678                 | 14678  468           |                   678                  | 3678          367    |               38         18     68     | 13689  689    1369   |                   ---------------------+----------------------+------------------               248    | 45789         4579   | 58     5789   278 128    189           | 5689          569    | 158    589        18     1489          |        489    479    |        789       `

Can anyone recover the whole grid and solve the original puzzle [without guessing where any missing candidates are located, and without using brute- force T&E]?

[Edited to make the puzzle more interesting, and to make the intended humor a bit clearer.]
Last edited by r.e.s. on Thu May 25, 2006 9:45 pm, edited 4 times in total.
r.e.s.

Posts: 337
Joined: 31 August 2005

I will post my solution in tiny text, not to spoil this nice challenge for others.

With the assumption of 180 degrees rotational symmetry, this would be your original puzzle:

901008200500020900000070041053000009029050410400000720360010000007030004005200306

After placing a few singles, the gremlins attacked. You were here:

901008200500020900000070041053002009029050410400000725360010000007030004005200306

The puzzle then solves with singles upto here. I advise you to use Gremlin Repel (extra strong) in this stage. Do NOT add water!

901068207570020908002070041053042009029050413400000725360010002207030104005200306

The next solving step requires a long hard look:

Unrecognizable rubbish >>>> XY-wing in R2C3. <<<< and more rubbish

It's all singles from here. So this is the solution:

000000000000000000000000000000000000000000000000000000000000000000000000000000000

.....

Uh-oh...

Ruud.
Ruud

Posts: 664
Joined: 28 October 2005

### Re: The Sudoku Gremlin is Back!

not done by hand
plugged 123456789 in each pilfered cell and let the solver at it
not sure if this would work for all pilfered pm grids

9 347 1 | 3456 46 8 | 2 357 37
5 3478 468 | 1346 2 1346 | 9 378 378
268 38 268 | 3569 7 3569 | 56 4 1
------------------+------------------+------------------
17 5 3 | 147 4 2 | 8 6 9
678 2 9 | 3678 5 367 | 4 1 3
4 18 68 |13689 689 1369 | 7 2 5
------------------+------------------+------------------
3 6 248 |45789 1 4579 | 5 5789 278
128 189 7 | 5689 3 569 | 15 589 4
18 1489 5 | 2 489 479 | 3 789 6
gsf
2014 Supporter

Posts: 7306
Joined: 21 September 2005
Location: NJ USA

Ok, guys .... I obviously should have chosen an asymmetrical puzzle to thwart that kind of guessing, and disallowed that kind of computering! (There's a more interesting way to solve it without guessing where any candidates are located, and without brute-force T&E.)
r.e.s.

Posts: 337
Joined: 31 August 2005

By hand using logic: When the gremlin hit ...
9.1..82..5...2.9......7..41.53..2..9.29.5.41.4.....72536..1......7.3...4..52..3.6

The solution ...
941368257576421938832975641753142869629857413418693725364719582287536194195284376
ronk
2012 Supporter

Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

r.e.s. wrote:Ok, guys .... I obviously should have chosen an asymmetrical puzzle to thwart that kind of guessing, and disallowed that kind of computering! (There's a more interesting way to solve it without guessing where any candidates are located, and without brute-force T&E.)

solving from the pm grid with 123456789 entries required no guessing
my computer will abstain from future ones
gsf
2014 Supporter

Posts: 7306
Joined: 21 September 2005
Location: NJ USA

Solving the grid is not a problem (as gsf mentioned, plugging the holes with 123456789 should do the trick without guessing), I would be interested to see how the original puzzle was recovered ??

tarek

tarek

Posts: 2650
Joined: 05 January 2006

tarek wrote:I would be interested to see how the original puzzle was recovered ?

I did this manually.
For every vandalized cell, check the 20 peers for each of the 9 digits. When 8 of 9 are candidates in any peer, the 9th digit can be plugged into the cell. A good place to start is R9C4, followed by other cells in houses with few holes.

The last 2 holes must be a naked pair, but you already have the original puzzle then. It is not difficult to find the naked pair using the original clues.

I finally removed digits until the symmetry was restored (only 2 digits needed to be removed.

Ruud.
Ruud

Posts: 664
Joined: 28 October 2005

tarek wrote:Solving the grid is not a problem (as gsf mentioned, plugging the holes with 123456789 should do the trick without guessing), I would be interested to see how the original puzzle was recovered ??

solve the plugged pm grid
if there is only one solution then fill in each pilfered cell with the corresponding solution value
gsf
2014 Supporter

Posts: 7306
Joined: 21 September 2005
Location: NJ USA

gsf wrote:solve the plugged pm grid
if there is only one solution then fill in each pilfered cell with the corresponding solution value

I did (I have the candidate grid as it should be), I didn't know which combinations of clues at hand to use in the reproduction of the original puzzle.... furthermore some plugins would reduce to polyvalued cells....were these clues or were they blanks originally ???

tarek

tarek

Posts: 2650
Joined: 05 January 2006

gsf wrote:solving from the pm grid with 123456789 entries required no guessing
my computer will abstain from future ones

Sorry gsf! I misunderstoood what you meant by "plugged 123456789 in each pilfered cell" -- I thought you were just being lazy and let your machine do T&E on the possibilities.
r.e.s.

Posts: 337
Joined: 31 August 2005

r.e.s. wrote:Sorry gsf! I misunderstoood what you meant by "plugged 123456789 in each pilfered cell" -- I thought you were just being lazy and let your machine do T&E on the possibilities.

and I skimmed over the part where naked pairs could be pilfered
I think there was one set pilfered so my solution has at least 2 extra clues
gsf
2014 Supporter

Posts: 7306
Joined: 21 September 2005
Location: NJ USA