The New Sudoku Players' Forum

[ravel]

Here is my solver's log.

The x-wings in 7 and 9 are there without your xy-wing.

[vidarino]

Oh, bugger, nevermind me... I messed up the order of the techniques while improving the "only generate puzzles that fit feature set X" bits... Sorry about that.

[tarek]

ab,
The puzzle you posted & the one you refer to as x are both beautiful as they have an All-axes & rotational symmetry, however they are not that difficlt to solve to warrant a superior.
The puzzle may have an X-wing but another simpler technique is present to save the day....
I'm sure you can twaek both a bit more. the last of your puzzle is worthy to be a superior

Mage,
welcome to forums, brilliant puzzles....the first is a bit too easy, the second is FANTABULOUS.

Tarek

[henk]

Rating 4352: (4-way symmetry)

Code: Select all

 . 5 . | . . . | . 6 .
 . . . | 7 . 4 | . . .
 . . 8 | . 2 . | 9 . .
-------+-------+------
 3 . . | 4 . 2 | . . 9
 . 2 9 | . . . | 6 5 .
 4 . . | 9 . 5 | . . 7
-------+-------+------
 . . 1 | . 3 . | 2 . .
 . . . | 8 . 7 | . . .
 . 6 . | . . . | . 3 .


[tarek]

henk,
beautiful puzzles, #31 & #36 in the updated list raises the bar considerably, good work.

Vidar, excellent all=Axes symmetry puzzles that are also superior.

Tarek

[Ruud]

Your rating system is very different of mine. It looks like you rate the number of steps higher then the technique. Could you explain this a little?

My rating system calculates a score based on the estimated time that a human solver would require. It takes into account how many unsolved cells are left and how many alternatives there are to the chosen technique, as well as the difficulty of the technique. A more complete explanation can be found here on the programmers forum.

@all: This collection is beautiful!

Now all we need is a publisher

Ruud.

[ab]

ab,

The puzzle may have an X-wing but another simpler technique is present to save the day....

Tarek

Ah I see it now, a box/line interaction. I think they're hard to spot when it's a case of 'if a candidate appears in only one box in a given line then it can't appear in another line in that box'. Still I'm happy to accept your rules.

I didn't really expect the other puzzle to make the list, but I like it.

I agree Mage's 2nd puzzle is interesting. needing pairs and triples before you start.

PS don't give me the credit for puzzle #30, I just took 2 clues out of Henk's puzzle.

[tarek]

I think that the rating is best done after applying these in order:
1. Simplest possible technique
2. if several similar level steps appear in succession, take out the ones that do not contribute to the progression of the puzzle. (this is best done after the puzzle is solved (if removing a step will just add singles to the end, then it is not needed)---- in other words the shortest possible path

then rating should then be carried out to account for:
1. difficulty of technique
2. variey

I did not implemnt this at the moment, so my solver may be over-rating some of your puzzles, but it shouldn't under-rate any.

Tarek

[henk]

Rating 4427: (4-way symmetry, 4 x-wings)

Code: Select all

 . . 2 | 5 . 7 | 4 . .
 6 . . | 9 . 3 | . . 8
 . . . | . 1 . | . . .
-------+-------+------
 9 . 4 | . . . | 5 . 2
 8 . . | . . . | . . 9
 2 . 3 | . . . | 8 . 7
-------+-------+------
 . . . | . 9 . | . . .
 7 . . | 8 . 6 | . . 3
 . . 8 | 7 . 2 | 6 . .


[JPF]

Some more :

Code: Select all

Puzzle #02

6 . . | 3 1 9 | . . 7
. . . | 2 . 4 | . . .
. . . | . . . | . . .
------+-------+------
1 8 . | . . . | . 6 9
7 . . | . 6 . | . . 1
2 4 . | . . . | . 7 5
------+-------+------
. . . | . . . | . . .
. . . | 4 . 8 | . . .
9 . . | 7 5 3 | . . 6


Puzzle #03

. . . | 1 6 3 | . . .
. . 4 | . . . | 9 . .
. 7 . | . . . | . 1 .
------+-------+------
6 . . | . . . | . . 5
9 . . | 5 . 7 | . . 4
8 . . | . . . | . . 9
------+-------+------
. 3 . | . . . | . 7 .
. . 8 | . . . | 2 . .
. . . | 8 9 2 | . . .

 

Puzzle #04

. . . | 4 1 5 | 6 . .
. . 3 | . . . | . 9 .
. 6 . | . . . | . . 4
------+-------+------
7 . 6 | . . . | . . 5
4 . . | 3 5 2 | . . 8
3 . . | . . . | 9 . 1
------+-------+------
8 . . | . . . | . 7 .
. 7 . | . . . | 2 . .
. . 2 | 9 3 7 | . . .


JPF

[henk]

JPF:
My solver used an Empty Rectangle and a XY-Wing to solve your first puzzle. The others are singles only...

[vidarino]

Two more from me;

Code: Select all

+-------+-------+-------+
| . 1 . | 7 . 8 | . 4 . |
| 6 . . | . 1 . | . . 2 |
| . 4 . | . . . | . 8 . |
+-------+-------+-------+
| . 6 . | 5 . 2 | . 3 . |
| 2 . . | . 9 . | . . 4 |
| . 9 . | 3 . 1 | . 2 . |
+-------+-------+-------+
| . 2 . | . . . | . 6 . |
| 9 . . | . 6 . | . . 5 |
| . 3 . | 8 . 5 | . 9 . |
+-------+-------+-------+

+-------+-------+-------+
| . 1 . | . 2 . | . 3 . |
| 5 . . | 6 . 4 | . . 8 |
| . . 2 | . . . | 9 . . |
+-------+-------+-------+
| 1 . . | 2 . 5 | . . 3 |
| . . 4 | . 8 . | 6 . . |
| 8 . . | 4 . 6 | . . 1 |
+-------+-------+-------+
| . . 8 | . . . | 1 . . |
| 6 . . | 7 . 3 | . . 9 |
| . 9 . | . 4 . | . 5 . |
+-------+-------+-------+

Vidar

[henk]

Vidarino, I rated your puzzles:
the first: 4305 (nice one!)
second: 2468

[vidarino]

Thanks, henk.

I really should look into implementing some kind of rating system myself, too. Currently I only keep counters for each strategy (e.g. "17 singles, 3 subsets, 2 fish"), which rarely reflects how difficult a puzzle really is.

It would be really cool if we (= contributors to this thread) could, after deeming the collection "done", provide a rated list of puzzles for comparison. Of course there is no "correct answer" for ratings, but it would be interesting to see how the different rating strategies react to the different puzzles.

[Ruud]

Nice entries Vidar,

these are my ratings, with techniques listed (HS-NS-LB-NP-NT-HP-HT-XW)

62-1740 (48-4-2-2-1-1-0-4)
63-1658 (51-1-4-2-0-2-0-3)

Henk, what makes the first puzzle score 1837 points more than the other?

Ruud.