## February 5, 2019

Post puzzles for others to solve here.

### Re: February 5, 2019

rjamil wrote:
SpAce wrote:
XY-Wing Transport: 348 @ r4c9 r9c49 Strong Link ERI 3 @ b4r4c3 => -3 @ r9c3; stte

Btw, what do you mean by the 'Strong Link' here? What is strongly linked to what?

It is based on Strong Link point 5.

So it means the internal strong link within the ER? That's redundant information because there's always a strong link within an ER between its row and column candidates. (Btw, it's true whether the ERI point is occupied or not; I don't understand StrmCkr's note about that). So, I don't think you need to mention that strong link at all, and it's in fact clearer if you don't. Otherwise it can be interpreted that you mean a strong link between the two patterns, which is inaccurate.

As far as ERI presentation is concern, need more precise presentation to be thinking/suggesting.

A couple of possibilities:

1) Using a '*' instead of '+' to mark the ER candidates:

Code: Select all
` +---------------+-------------+--------------+ | 37   2   6    | 4     5  8  | 9   1   37   | | 78   34  5    | 6     1  9  | 24  23  78   | | 89   49  1    | 2     7  3  | 5   48  6    | +---------------+-------------+--------------+ | 9*3  7   29*3 | 5     8  1  | 24  6   (34) | | 6    1   4    | 7     3  2  | 8   5   9    | | 5    8   2*3  | 9     4  6  | 7   23  1    | +---------------+-------------+--------------+ | 4    39  379  | 38    6  5  | 1   78  2    | | 2    6   8    | 1     9  47 | 3   47  5    | | 1    5   7-3  | (38)  2  47 | 6   9   (48) | +---------------+-------------+--------------+`

2) Using '*' to mark the ER cells (pretty standard for any pattern)

Code: Select all
` +-----------------+-------------+--------------+ |  37   2    6    | 4     5  8  | 9   1   37   | |  78   34   5    | 6     1  9  | 24  23  78   | |  89   49   1    | 2     7  3  | 5   48  6    | +-----------------+-------------+--------------+ | *39   7   *239  | 5     8  1  | 24  6   (34) | |  6    1    4    | 7     3  2  | 8   5   9    | |  5    8   *23   | 9     4  6  | 7   23  1    | +-----------------+-------------+--------------+ |  4    39   379  | 38    6  5  | 1   78  2    | |  2    6    8    | 1     9  47 | 3   47  5    | |  1    5    7-3  | (38)  2  47 | 6   9   (48) | +-----------------+-------------+--------------+`

3) Using {} to mark the ER candidates (similar to your use of () for the XY-Wing candidates):

Code: Select all
`+-----------------+-------------+--------------+ |  37   2   6     | 4     5  8  | 9   1   37   | |  78   34  5     | 6     1  9  | 24  23  78   | |  89   49  1     | 2     7  3  | 5   48  6    | +-----------------+-------------+--------------+ | {3}9  7   2{3}9 | 5     8  1  | 24  6   (34) | |  6    1   4     | 7     3  2  | 8   5   9    | |  5    8   2{3}  | 9     4  6  | 7   23  1    | +-----------------+-------------+--------------+ |  4    39  379   | 38    6  5  | 1   78  2    | |  2    6   8     | 1     9  47 | 3   47  5    | |  1    5   7-3   | (38)  2  47 | 6   9   (48) | +-----------------+-------------+--------------+`

4) Using chain markers and Eureka (my preference)

Code: Select all
`.---------------.--------------.-------------.|  37  2   6    |   4    5  8  | 9   1    37 ||  78  34  5    |   6    1  9  | 24  23   78 ||  89  49  1    |   2    7  3  | 5   48   6  |:---------------+--------------+-------------:| d39  7  d239  |   5    8  1  | 24  6   c34 ||  6   1   4    |   7    3  2  | 8   5    9  ||  5   8  e2(3) |   9    4  6  | 7   23   1  |:---------------+--------------+-------------:|  4   39  379  |   38   6  5  | 1   78   2  ||  2   6   8    |   1    9  47 | 3   47   5  ||  1   5   7-3  | a(3)8  2  47 | 6   9   b48 |'---------------'--------------'-------------'`

XY-Wing with transport (ER):

(3=8)r9c4 - (8=4)r9c9 - (4=3)r4c9 - r4c13 = (3)r6c3 => -3 r9c3; stte

or shorter: Show
(3=84)r9c49 - (4=3)r4c9 - (3)b4p[13=9] => -3 r9c3; stte

Or whatever. Just don't use something that already has a well-established very different meaning like the '+' or <>, and you're on the safe side.

SpAce

Posts: 1903
Joined: 22 May 2017

### Re: February 5, 2019

SpAce wrote:
Ngisa wrote:It is important to complete the basics, otherwise you can have an advantage of using candidates which were not supposed to be in that cell to make eliminations. Fortunately I did not use them in my chain in this puzzle.

How exactly could that ever be an advantage? I can't imagine any scenario where leaving free basic eliminations and even placements on the table would actually help anyone in these challenges. It can definitely make things more difficult but not easier. (Not really sure if you were joking or serious, though.)

The only situation i am aware of, that it could be harder (to spot) with more eliminations is, when a UR digit is eliminated, see here. But that is not possible with basics.
eleven

Posts: 2178
Joined: 10 February 2008

### Re: February 5, 2019

eleven wrote:The only situation i am aware of, that it could be harder (to spot) with more eliminations is, when a UR digit is eliminated, see here. But that is not possible with basics.

Yes, that came to my mind too. I hadn't actually realized that basics can't kill UR digits (good to know!), but I just thought that even if it happened, it wouldn't make the UR unusable (just harder to spot). On the other hand, adding more clues can really break URs and make puzzles harder, as demonstrated here.

SpAce

Posts: 1903
Joined: 22 May 2017

### Re: February 5, 2019

Hi SpAce,

How about "XY-Wing Transport: 348 @ r4c9 r9c49 ERI 3 @ b4r4c3 => -3 @ r9c3; stte" by just dropping "Strong Link" wordings and not adding "with" between "XY-Wing" and "Transport" words?

As far as pencil mark marker for empty rectangle is concern, let consider the markers as per Andrew's online solver.
As per February 7, 2019 puzzle, after basics and skipping several moves, found empty rectangles and pm shown as follows:
ERI.png (24.06 KiB) Viewed 100 times

For pure text base pencil mark grid representation instead of coloring: place an asterisk "*" before ERI cell (not digit) and "/" before empty rectangle cells.

Similarly, I have tested following collection of puzzles with and without including ERI cell check (i.e., let for example ERI cell is C9, checked C789 and ABC9; and checked C78 and AB9). Found no difference in searching patterns that contain ERI formation as strong/weak link(s):

Ruud specialty puzzles: 50,000
Tarek's Pearly: 6,000
Mike Metcalf patterns games puzzles: 26,855
17 clue puzzles: 49,157

R. Jamil
rjamil

Posts: 418
Joined: 15 October 2014
Location: Karachi, Pakistan

### Re: February 5, 2019

rjamil wrote:How about "XY-Wing Transport: 348 @ r4c9 r9c49 ERI 3 @ b4r4c3 => -3 @ r9c3; stte" by just dropping "Strong Link" wordings and not adding "with" between "XY-Wing" and "Transport" words?

Whatever you like. I'm not exactly motivated to think about these pattern-specific notations more, because I'm never going to use them myself. I can usually read them if people insist on using them, but personally I'd very much prefer something closer to Eureka, because it's a generic solution that works with almost anything. Your style works with these simple examples, but I'm pretty sure it would get really messy or outright impossible with anything more complicated.

PS. You don't ever see me use even more generally accepted ALS XZ or Sue de Coq notations either, because I think they're just unnecessary and ugly complications. Eureka works for those cases just fine and is easier to understand, at least for me. When a generic solution works nicely, why waste time learning specific ones? Besides Eureka (+some extensions), the only sudoku notations I've found useful are the UFG fish notation and Allan Barker's set notation. Anything else is unnecessary, at least with any techniques I'm familiar with, and (like Frank Underwood) I have no patience for useless things

SpAce

Posts: 1903
Joined: 22 May 2017

Previous