The New Sudoku Players' Forum

[champagne]

::: comment :::
Hi champagne

Nothing new from my night batch

So my 11.x are

Code: Select all

12.3.....4...2.5....6..7...3..1..2...5..8..3...4.....9.4.5..3......9..8......6..7 ED=11.2/11.2/10.8  OR=9042
12.3.....3...1.4....5..6...7..2..3...3..8..7...9.....8.1.4..7......6..9......8..5 ED=11.1/11.1/9.9  OR=166149
12.3.....4...1.2....5..2...3..6..4...4..5..7...8.....9.6.4..1......9..5......7..8 ED=11.2/11.2/10.0  OR=349300
12.3.....4...2.5....6..7...5..4..2...3..8..5...1.....9.1.2..3......9..8......6..7 ED=11.4/11.4/10.7  OR=357441
12.3.....4...2.1....5..6...3..7..4...4..8..5...9.....8.7.4..2......6..9......8..3 ED=11.1/11.1/9.9  OR=523493
12.3.....4...1.2....5..6...3..7..4...4..8..9...9.....8.7.4..1......6..5......8..3 ED=11.1/11.1/9.4  OR=624539
12.3.....4...1.2....5..2...3..6..4...1..7..3...8.....9.6.4..1......8..5......9..7 ED=11.3/11.3/9.4  OR=643121

The last number OR= is the order of the sudoku in my database
The 11.2 came first very rapidly only 9042 sudoku generated
The last found one came 643121
I've now 2 250 000 sudokus in the DB but nothing new with the additionnal 1.6 million sudokus

It seems I've not found the right place to generate new ones : I've just stopped everything !

How many sudokus do you have for this game ?
And I'm interested if you have the equivalent of my OR
Best regards

Hi Patrice, I find better to answer here.

I have currently nearly 200 millions ED puzzles in my data base for that game.

I keep no track of the sequence

Contrary to you, as I had no free time, I launched long runs, so my good results came late, what is not good to win a game.

I used my seed generator to feed most of the areas of the game, but the bundle of other players produced most of the high ratings with a simpler process.
One problem for me if the ratio of high ratings given by that pattern. Consequently, ratings last very long.

I have 5 relatively small batches from that night still at the rating step.

My last findings are a new 11;x (not yet rated with serate) and a 10.8 10.8 10.6

I made a quick analysis, it seems that the main exotic property for that pattern is a big rank 0 logic.

Regarding the number of puzzles, keep in mind that I have no filter at generation time

[m_b_metcalf]

always difficult to answer to such a question out of realized figures, but I have 100 millions ED puzzles in my data base, so the minimum should be 200 millions.
I likely covered more than 80% of the high ratings, but the yield of my search is much lower in the low ratings.

Interestingly, it didn't take champagne long to find his own (canonical form of the) seed for this game:

Code: Select all

98.7.....6...9.5....4..6.9.4....3.8..63.....2..8.1.3...4...8.6....2....5...1....7;10.50;10.50;2.60;GP;13_03;847912;25;*;
.7...5.8.8..7....2..486......1...2...6..5..7...3...4......791..3....6..9.9.5...60  pi  rotational  (180  degrees)

[champagne]

Hi Mike,

you forced me to dig in the bin

the following puzzle is rated by skfr. I did not check again the serate rating

Code: Select all

.1...2.3.
3..1....4
..536....
..7...4..
.6..2..1.
..8...5..
....197..
8....6..9
.9.2...6.;10.5;10.5;2.6

as you can see, it has a symmetry of given, so I got it some minutes after I started the exploration of that pattern.

But generally speaking, with the strategy I apply, the highest rating comes in the first day, at worst in the second day.

In that game, as far as I remember, having a limited number of puzzles passing the dynamic plus level in serate, I used some depth 4 vicinity, but most important, I used all "group diamonds" (solved using the same serate level of eliminations) as seed. That's the way later I got the 2 cats.

[m_b_metcalf]

This is a pattern where I would be tempted to play for rarity, but I have a very old base of past games sent long ago by Mike so it's not for me.

So that everyone has access to a recent data base, I attach my most recent version. It occasionally differs from g.r.emlin's, mainly because of some system hiccoughs around game 40.

Mike

Edit: file deleted as there is a new one on the next page.

[champagne]

Thanks to mike to share that base.

New players should download and use it

[champagne]

Not a full scan, it is slow. Regular +/-.
72,228,251 puzzles in total.
all rated by skfr
all */*/3.0+ rated by serate
all 8.0+/*/2.0+ rated by serate
all enclosed to {-1,+1}
all */*/1.5+ enclosed to {-3,+3}
at 8M puzzles all 9.3/*/* enclosed to {-3,+3}
at 65M puzzles all */1.5+/* enclosed to {-3,+3}
at 67M puzzles all 8.8/*/* enclosed to {-2,+2}
at 56M puzzles all */*/7.2+ enclosed to {-4,+4}
at 56M puzzles all 9.4+/*/* enclosed to {-4,+4}

68 runs, about 300 generations, some of them contributing only few puzzles.

I didn't reach 7.9/7.9/2.6 and few of the others.
Below is the puzzles count by serate.

Hidden Text: Show

72201679 ED=0.0/0.0/0.0
1 ED=10.4/1.2/1.2
1 ED=3.6/3.6/3.0
1 ED=3.6/3.6/3.6
1 ED=4.4/4.2/2.6
1 ED=4.5/4.5/3.4
1 ED=4.6/3.6/2.6
1 ED=4.7/1.2/1.2
1 ED=4.8/1.2/1.2
1 ED=4.9/1.2/1.2
1 ED=5.0/1.2/1.2
1 ED=5.1/1.2/1.2
1 ED=5.2/2.0/2.0
1 ED=5.6/3.6/2.6
1 ED=5.7/3.6/2.6
1 ED=5.8/1.2/1.2
1 ED=5.9/1.2/1.2
1 ED=6.0/1.2/1.2
1 ED=6.1/1.2/1.2
1 ED=6.2/1.2/1.2
1 ED=6.5/3.6/2.6 *
1 ED=6.6/6.6/2.6
1 ED=6.7/6.7/2.6 *
1 ED=6.8/6.7/2.6 *
2 ED=6.9/6.6/2.6 *
1 ED=7.1/3.8/3.0
4 ED=7.1/3.8/3.4
33 ED=7.1/3.8/3.8
3 ED=7.1/4.0/3.8
1 ED=7.1/6.6/2.6
6 ED=7.1/6.6/3.0
18 ED=7.1/7.1/3.0
203 ED=7.1/7.1/3.4
1 ED=7.1/7.1/3.6
273 ED=7.1/7.1/3.8
3 ED=7.1/7.1/4.0
37 ED=7.1/7.1/6.7
77 ED=7.1/7.1/7.1
14 ED=7.2/3.4/3.0
32 ED=7.2/3.8/3.0
8 ED=7.2/3.8/3.4
28 ED=7.2/3.8/3.8
4 ED=7.2/4.0/3.0
1 ED=7.2/4.0/3.8
19 ED=7.2/6.6/3.0
1 ED=7.2/6.6/3.8
1 ED=7.2/7.1/3.0
74 ED=7.2/7.1/3.4
1 ED=7.2/7.1/3.6
80 ED=7.2/7.1/3.8
11 ED=7.2/7.1/6.7
17 ED=7.2/7.1/7.1
66 ED=7.2/7.2/3.0
251 ED=7.2/7.2/3.4
1 ED=7.2/7.2/3.6
221 ED=7.2/7.2/3.8
1 ED=7.2/7.2/4.0
41 ED=7.2/7.2/6.7
66 ED=7.2/7.2/7.1
29 ED=7.2/7.2/7.2
26 ED=7.3/3.4/3.0
1 ED=7.3/3.4/3.4
14 ED=7.3/3.8/3.0
12 ED=7.3/3.8/3.4
57 ED=7.3/3.8/3.8
20 ED=7.3/6.6/3.0
2 ED=7.3/6.6/3.4
1 ED=7.3/6.6/3.8
3 ED=7.3/7.1/3.0
36 ED=7.3/7.1/3.4
54 ED=7.3/7.1/3.8
1 ED=7.3/7.1/4.0
12 ED=7.3/7.1/6.7
11 ED=7.3/7.1/7.1
20 ED=7.3/7.2/3.0
34 ED=7.3/7.2/3.4
16 ED=7.3/7.2/3.8
1 ED=7.3/7.2/4.0
1 ED=7.3/7.2/7.1
1 ED=7.3/7.2/7.2
33 ED=7.3/7.3/3.0
113 ED=7.3/7.3/3.4
5 ED=7.3/7.3/3.6
74 ED=7.3/7.3/3.8
1 ED=7.3/7.3/4.0
1 ED=7.3/7.3/6.7
1 ED=7.3/7.3/7.2 *
12 ED=7.4/3.4/3.0
1 ED=7.4/3.8/3.0
3 ED=7.4/3.8/3.8
3 ED=7.4/7.1/3.4
2 ED=7.4/7.1/3.8
2 ED=7.4/7.1/7.1
1 ED=7.4/7.2/3.0
1 ED=7.4/7.2/3.4
4 ED=7.4/7.2/3.6
1 ED=7.4/7.2/3.8
1 ED=7.4/7.3/3.4
3 ED=7.4/7.3/3.8
2 ED=7.4/7.4/3.4
17 ED=7.4/7.4/3.8
1 ED=7.4/7.4/7.1 *
1 ED=7.5/3.8/3.0
1 ED=7.5/4.0/3.8
1 ED=7.5/6.6/3.0
1 ED=7.5/7.1/3.4
1 ED=7.5/7.2/3.4
1 ED=7.5/7.2/7.1
1 ED=7.5/7.3/3.0
1 ED=7.5/7.3/3.8
3 ED=7.5/7.5/3.8 *
1 ED=7.6/3.4/3.0
6 ED=7.6/3.8/3.0
1 ED=7.6/7.1/7.1
4 ED=7.6/7.2/3.8
1 ED=7.6/7.3/3.0
3 ED=7.6/7.3/3.4
1 ED=7.6/7.6/3.4
1 ED=7.6/7.6/3.8 *
1 ED=7.7/3.4/3.4
1 ED=7.7/3.8/3.0
1 ED=7.7/3.8/3.4
1 ED=7.7/3.8/3.8
4 ED=7.7/7.1/3.4
3 ED=7.7/7.1/6.7
1 ED=7.7/7.2/3.4
1 ED=7.7/7.2/3.8
1 ED=7.7/7.6/3.0
2 ED=7.7/7.7/3.4
1 ED=7.7/7.7/3.8 *
3 ED=7.8/3.4/3.0
1 ED=7.8/3.8/3.0
1 ED=7.8/3.8/3.4
1 ED=7.8/7.1/3.4
1 ED=7.8/7.1/3.8
1 ED=7.8/7.2/3.8
3 ED=7.8/7.3/3.4
2 ED=7.8/7.3/3.8
1 ED=7.8/7.4/3.8
2 ED=7.8/7.8/3.0
1 ED=7.8/7.8/3.4
1 ED=7.8/7.8/3.8 *
1 ED=7.9/3.8/3.0
2 ED=7.9/3.8/3.8
1 ED=7.9/7.1/2.6
1 ED=7.9/7.1/3.8
1 ED=7.9/7.2/3.8 *
3 ED=8.0/1.2/1.2
3 ED=8.0/3.4/3.0
1 ED=8.0/3.8/3.0 *
35 ED=8.1/1.2/1.2
1 ED=8.1/3.8/3.0 *
1 ED=8.2/1.2/1.2
71 ED=8.2/2.0/2.0
1 ED=8.2/2.5/2.5
1 ED=8.2/2.6/2.6
2 ED=8.2/3.4/2.6
1 ED=8.2/3.8/3.0
3 ED=8.2/4.0/2.6
3 ED=8.2/6.6/2.6
2 ED=8.2/6.6/3.0
35 ED=8.2/7.1/2.6
8 ED=8.2/7.1/3.4
2 ED=8.2/7.1/3.8
3 ED=8.2/7.1/7.1
5 ED=8.2/7.2/2.6
3 ED=8.2/7.2/3.4
1 ED=8.2/7.2/6.7
4 ED=8.2/7.3/2.6
1 ED=8.2/7.3/3.0
1 ED=8.2/7.3/3.6
1 ED=8.2/7.4/2.6
1 ED=8.2/7.6/3.0
1 ED=8.2/7.8/3.0 *
12 ED=8.2/8.2/2.6
2 ED=8.2/8.2/3.0
14 ED=8.2/8.2/3.4
605 ED=8.3/2.0/2.0
2 ED=8.3/2.5/2.5
16 ED=8.3/2.6/2.6
15 ED=8.3/2.8/2.6
2 ED=8.3/3.0/2.6
36 ED=8.3/3.4/2.6
28 ED=8.3/3.4/3.0
1 ED=8.3/3.4/3.4
14 ED=8.3/3.6/2.6
20 ED=8.3/3.8/3.0
35 ED=8.3/3.8/3.4
100 ED=8.3/3.8/3.8
37 ED=8.3/4.0/2.6
1 ED=8.3/4.0/3.0
2 ED=8.3/4.0/3.4
1 ED=8.3/4.0/3.8
5 ED=8.3/4.2/2.6
49 ED=8.3/6.6/2.6
22 ED=8.3/6.6/3.0
12 ED=8.3/6.7/2.6
867 ED=8.3/7.1/2.6
66 ED=8.3/7.1/3.4
1 ED=8.3/7.1/3.6
39 ED=8.3/7.1/3.8
6 ED=8.3/7.1/6.7
11 ED=8.3/7.1/7.1
108 ED=8.3/7.2/2.6
1 ED=8.3/7.2/3.0
29 ED=8.3/7.2/3.4
2 ED=8.3/7.2/3.6
12 ED=8.3/7.2/3.8
3 ED=8.3/7.2/6.7
4 ED=8.3/7.2/7.1
3 ED=8.3/7.2/7.2
34 ED=8.3/7.3/2.6
5 ED=8.3/7.3/3.0
12 ED=8.3/7.3/3.4
2 ED=8.3/7.3/3.6
6 ED=8.3/7.3/3.8
1 ED=8.3/7.3/4.0
8 ED=8.3/7.4/2.6
1 ED=8.3/7.4/3.0
3 ED=8.3/7.5/2.6
4 ED=8.3/7.6/2.6
6 ED=8.3/7.8/2.6
4 ED=8.3/7.8/3.8
2 ED=8.3/7.9/2.6
1 ED=8.3/7.9/3.0
4 ED=8.3/8.2/2.6
2 ED=8.3/8.2/3.0
5 ED=8.3/8.2/3.4
314 ED=8.3/8.3/2.6
158 ED=8.3/8.3/3.0
354 ED=8.3/8.3/3.4
2 ED=8.3/8.3/3.6
312 ED=8.3/8.3/3.8
4 ED=8.3/8.3/4.0
3 ED=8.3/8.3/6.7
1 ED=8.3/8.3/7.2 *
527 ED=8.4/2.0/2.0
1 ED=8.4/2.5/2.5
23 ED=8.4/2.6/2.6
15 ED=8.4/2.8/2.6
1 ED=8.4/3.0/2.6
63 ED=8.4/3.4/2.6
24 ED=8.4/3.4/3.0
6 ED=8.4/3.4/3.4
19 ED=8.4/3.6/2.6
20 ED=8.4/3.8/3.0
25 ED=8.4/3.8/3.4
1 ED=8.4/3.8/3.6
89 ED=8.4/3.8/3.8
57 ED=8.4/4.0/2.6
2 ED=8.4/4.0/3.8
2 ED=8.4/4.2/2.6
1 ED=8.4/4.2/3.4
28 ED=8.4/6.6/2.6
18 ED=8.4/6.6/3.0
25 ED=8.4/6.7/2.6
751 ED=8.4/7.1/2.6
4 ED=8.4/7.1/3.0
50 ED=8.4/7.1/3.4
1 ED=8.4/7.1/3.6
27 ED=8.4/7.1/3.8
15 ED=8.4/7.1/6.7
4 ED=8.4/7.1/7.1
107 ED=8.4/7.2/2.6
10 ED=8.4/7.2/3.0
20 ED=8.4/7.2/3.4
14 ED=8.4/7.2/3.8
3 ED=8.4/7.2/6.7
3 ED=8.4/7.2/7.1
1 ED=8.4/7.2/7.2
42 ED=8.4/7.3/2.6
2 ED=8.4/7.3/3.0
16 ED=8.4/7.3/3.4
14 ED=8.4/7.3/3.8
2 ED=8.4/7.3/6.7
7 ED=8.4/7.4/2.6
1 ED=8.4/7.4/3.4
2 ED=8.4/7.4/3.8
3 ED=8.4/7.5/2.6
2 ED=8.4/7.6/3.8
1 ED=8.4/7.7/2.6
12 ED=8.4/7.8/2.6
6 ED=8.4/7.8/3.0
1 ED=8.4/7.9/2.6
4 ED=8.4/8.2/2.6
1 ED=8.4/8.2/3.0
3 ED=8.4/8.2/3.4
1 ED=8.4/8.2/3.8
1 ED=8.4/8.2/4.0
109 ED=8.4/8.3/2.6
7 ED=8.4/8.3/3.0
49 ED=8.4/8.3/3.4
57 ED=8.4/8.3/3.8
2 ED=8.4/8.3/4.0
1 ED=8.4/8.3/6.7
1 ED=8.4/8.3/7.1
647 ED=8.4/8.4/2.6
47 ED=8.4/8.4/3.0
426 ED=8.4/8.4/3.4
3 ED=8.4/8.4/3.6
150 ED=8.4/8.4/3.8
5 ED=8.4/8.4/4.0
11 ED=8.4/8.4/6.7
1 ED=8.4/8.4/7.2 *
282 ED=8.5/2.0/2.0
10 ED=8.5/2.6/2.6
13 ED=8.5/2.8/2.6
1 ED=8.5/3.0/2.6
30 ED=8.5/3.4/2.6
18 ED=8.5/3.4/3.0
3 ED=8.5/3.4/3.4
10 ED=8.5/3.6/2.6
13 ED=8.5/3.8/3.0
14 ED=8.5/3.8/3.4
67 ED=8.5/3.8/3.8
22 ED=8.5/4.0/2.6
1 ED=8.5/4.0/3.0
1 ED=8.5/4.0/3.4
1 ED=8.5/4.0/3.8
7 ED=8.5/4.2/2.6
15 ED=8.5/6.6/2.6
1 ED=8.5/6.6/3.0
3 ED=8.5/6.7/2.6
415 ED=8.5/7.1/2.6
22 ED=8.5/7.1/3.4
14 ED=8.5/7.1/3.8
1 ED=8.5/7.1/4.0
4 ED=8.5/7.1/6.7
5 ED=8.5/7.1/7.1
70 ED=8.5/7.2/2.6
8 ED=8.5/7.2/3.4
1 ED=8.5/7.2/3.6
7 ED=8.5/7.2/3.8
1 ED=8.5/7.2/6.7
1 ED=8.5/7.2/7.1
2 ED=8.5/7.2/7.2
50 ED=8.5/7.3/2.6
3 ED=8.5/7.3/3.0
4 ED=8.5/7.3/3.4
3 ED=8.5/7.3/3.8
1 ED=8.5/7.3/4.0
1 ED=8.5/7.3/6.7
1 ED=8.5/7.4/2.6
2 ED=8.5/7.4/3.8
1 ED=8.5/7.5/2.6
1 ED=8.5/7.7/3.8
14 ED=8.5/7.8/2.6
3 ED=8.5/7.8/3.0
2 ED=8.5/7.9/2.6
3 ED=8.5/8.2/2.6
1 ED=8.5/8.2/3.0
1 ED=8.5/8.2/3.4
66 ED=8.5/8.3/2.6
3 ED=8.5/8.3/3.0
35 ED=8.5/8.3/3.4
28 ED=8.5/8.3/3.8
2 ED=8.5/8.3/6.7
2 ED=8.5/8.3/7.1
96 ED=8.5/8.4/2.6
4 ED=8.5/8.4/3.0
48 ED=8.5/8.4/3.4
38 ED=8.5/8.4/3.8
4 ED=8.5/8.4/4.0
1 ED=8.5/8.4/8.3
383 ED=8.5/8.5/2.6
21 ED=8.5/8.5/3.0
201 ED=8.5/8.5/3.4
82 ED=8.5/8.5/3.8
5 ED=8.5/8.5/4.0
5 ED=8.5/8.5/6.7
1 ED=8.5/8.5/7.1 *
32 ED=8.6/2.0/2.0
2 ED=8.6/2.6/2.6
1 ED=8.6/3.0/2.6
4 ED=8.6/3.4/2.6
3 ED=8.6/3.6/2.6
1 ED=8.6/3.8/3.0
3 ED=8.6/3.8/3.4
12 ED=8.6/3.8/3.8
2 ED=8.6/4.0/2.6
61 ED=8.6/7.1/2.6
5 ED=8.6/7.1/3.4
1 ED=8.6/7.1/3.8
7 ED=8.6/7.2/2.6
1 ED=8.6/7.2/3.4
2 ED=8.6/7.2/3.8
1 ED=8.6/7.2/4.0
3 ED=8.6/7.3/2.6
1 ED=8.6/7.3/3.8
1 ED=8.6/8.2/3.4
10 ED=8.6/8.3/2.6
5 ED=8.6/8.3/3.4
1 ED=8.6/8.3/3.8
7 ED=8.6/8.4/2.6
1 ED=8.6/8.4/3.0
6 ED=8.6/8.4/3.4
11 ED=8.6/8.5/2.6
6 ED=8.6/8.5/3.4
1 ED=8.6/8.5/3.8
47 ED=8.6/8.6/2.6
7 ED=8.6/8.6/3.0
29 ED=8.6/8.6/3.4
13 ED=8.6/8.6/3.8
1 ED=8.6/8.6/6.7 *
8 ED=8.7/2.0/2.0
1 ED=8.7/2.6/2.6
1 ED=8.7/3.8/3.4
1 ED=8.7/4.0/3.4
6 ED=8.7/7.1/2.6
1 ED=8.7/7.2/2.6
1 ED=8.7/8.5/2.6
1 ED=8.7/8.6/2.6
33 ED=8.7/8.7/2.6
5 ED=8.7/8.7/3.4
1 ED=8.7/8.7/3.8
75 ED=8.8/2.0/2.0
7 ED=8.8/2.6/2.6
4 ED=8.8/2.8/2.6
4 ED=8.8/3.4/2.6
1 ED=8.8/3.4/3.0
1 ED=8.8/3.4/3.4
5 ED=8.8/3.6/2.6
4 ED=8.8/3.8/3.4
20 ED=8.8/3.8/3.8
4 ED=8.8/4.0/2.6
2 ED=8.8/6.6/2.6
1 ED=8.8/6.7/2.6
73 ED=8.8/7.1/2.6
8 ED=8.8/7.1/3.4
2 ED=8.8/7.1/3.8
1 ED=8.8/7.1/6.7
1 ED=8.8/7.1/7.1
14 ED=8.8/7.2/2.6
1 ED=8.8/7.2/3.0
2 ED=8.8/7.2/3.4
1 ED=8.8/7.2/3.8
3 ED=8.8/7.3/2.6
4 ED=8.8/7.3/3.4
1 ED=8.8/7.3/3.8
1 ED=8.8/7.4/2.6
1 ED=8.8/7.4/3.8
1 ED=8.8/7.5/2.6
2 ED=8.8/7.8/2.6
1 ED=8.8/8.2/2.6
1 ED=8.8/8.2/3.4
10 ED=8.8/8.3/2.6
2 ED=8.8/8.3/3.4
16 ED=8.8/8.4/2.6
1 ED=8.8/8.4/3.0
15 ED=8.8/8.4/3.4
11 ED=8.8/8.4/3.8
1 ED=8.8/8.4/4.0
12 ED=8.8/8.5/2.6
1 ED=8.8/8.5/3.0
10 ED=8.8/8.5/3.4
2 ED=8.8/8.5/3.8
4 ED=8.8/8.6/2.6
2 ED=8.8/8.7/2.6
158 ED=8.8/8.8/2.6
13 ED=8.8/8.8/3.0
66 ED=8.8/8.8/3.4
2 ED=8.8/8.8/3.6
39 ED=8.8/8.8/3.8
1 ED=8.8/8.8/6.7
1 ED=8.8/8.8/7.3 *
219 ED=8.9/2.0/2.0
2 ED=8.9/2.5/2.5
24 ED=8.9/2.6/2.6
6 ED=8.9/2.8/2.6
28 ED=8.9/3.4/2.6
10 ED=8.9/3.4/3.0
5 ED=8.9/3.4/3.4
7 ED=8.9/3.6/2.6
21 ED=8.9/3.8/3.0
17 ED=8.9/3.8/3.4
54 ED=8.9/3.8/3.8
15 ED=8.9/4.0/2.6
2 ED=8.9/4.0/3.4
1 ED=8.9/4.0/3.8
2 ED=8.9/4.2/2.6
5 ED=8.9/6.6/2.6
3 ED=8.9/6.6/3.0
6 ED=8.9/6.7/2.6
220 ED=8.9/7.1/2.6
31 ED=8.9/7.1/3.4
9 ED=8.9/7.1/3.8
3 ED=8.9/7.1/6.7
2 ED=8.9/7.1/7.1
52 ED=8.9/7.2/2.6
2 ED=8.9/7.2/3.0
11 ED=8.9/7.2/3.4
5 ED=8.9/7.2/3.8
21 ED=8.9/7.3/2.6
1 ED=8.9/7.3/3.0
6 ED=8.9/7.3/3.4
4 ED=8.9/7.3/3.8
1 ED=8.9/7.3/4.0
1 ED=8.9/7.4/2.6
1 ED=8.9/7.5/2.6
1 ED=8.9/7.6/2.6
5 ED=8.9/7.8/2.6
2 ED=8.9/7.8/3.0
33 ED=8.9/8.3/2.6
2 ED=8.9/8.3/3.0
21 ED=8.9/8.3/3.4
17 ED=8.9/8.3/3.8
61 ED=8.9/8.4/2.6
6 ED=8.9/8.4/3.0
50 ED=8.9/8.4/3.4
21 ED=8.9/8.4/3.8
1 ED=8.9/8.4/4.0
1 ED=8.9/8.4/6.7
61 ED=8.9/8.5/2.6
3 ED=8.9/8.5/3.0
17 ED=8.9/8.5/3.4
13 ED=8.9/8.5/3.8
6 ED=8.9/8.6/2.6
1 ED=8.9/8.6/3.0
1 ED=8.9/8.6/3.8
3 ED=8.9/8.7/2.6
1 ED=8.9/8.7/3.0
1 ED=8.9/8.7/3.4
1 ED=8.9/8.7/3.8
11 ED=8.9/8.8/2.6
1 ED=8.9/8.8/3.0
2 ED=8.9/8.8/3.4
2 ED=8.9/8.8/3.8
604 ED=8.9/8.9/2.6
51 ED=8.9/8.9/3.0
451 ED=8.9/8.9/3.4
168 ED=8.9/8.9/3.8
4 ED=8.9/8.9/4.0
26 ED=8.9/8.9/6.7 *
450 ED=9.0/2.0/2.0
1 ED=9.0/2.5/2.5
22 ED=9.0/2.6/2.6
7 ED=9.0/2.8/2.6
1 ED=9.0/3.0/2.6
42 ED=9.0/3.4/2.6
16 ED=9.0/3.4/3.0
12 ED=9.0/3.4/3.4
14 ED=9.0/3.6/2.6
24 ED=9.0/3.8/3.0
38 ED=9.0/3.8/3.4
142 ED=9.0/3.8/3.8
34 ED=9.0/4.0/2.6
1 ED=9.0/4.0/3.0
1 ED=9.0/4.0/3.4
2 ED=9.0/4.0/3.8
5 ED=9.0/4.2/2.6
2 ED=9.0/4.2/3.4
7 ED=9.0/6.6/2.6
11 ED=9.0/6.6/3.0
9 ED=9.0/6.7/2.6
1 ED=9.0/7.0/2.6
499 ED=9.0/7.1/2.6
35 ED=9.0/7.1/3.4
17 ED=9.0/7.1/3.8
5 ED=9.0/7.1/6.7
3 ED=9.0/7.1/7.1
76 ED=9.0/7.2/2.6
5 ED=9.0/7.2/3.0
22 ED=9.0/7.2/3.4
15 ED=9.0/7.2/3.8
1 ED=9.0/7.2/7.1
78 ED=9.0/7.3/2.6
2 ED=9.0/7.3/3.0
21 ED=9.0/7.3/3.4
7 ED=9.0/7.3/3.8
2 ED=9.0/7.3/4.0
2 ED=9.0/7.3/6.7
1 ED=9.0/7.3/7.2
2 ED=9.0/7.4/2.6
1 ED=9.0/7.4/3.0
1 ED=9.0/7.4/3.8
6 ED=9.0/7.5/2.6
1 ED=9.0/7.6/3.0
1 ED=9.0/7.6/3.8
10 ED=9.0/7.8/2.6
6 ED=9.0/7.8/3.0
1 ED=9.0/7.9/2.6
2 ED=9.0/8.2/2.6
3 ED=9.0/8.2/3.4
83 ED=9.0/8.3/2.6
30 ED=9.0/8.3/3.4
28 ED=9.0/8.3/3.8
1 ED=9.0/8.3/7.1
84 ED=9.0/8.4/2.6
3 ED=9.0/8.4/3.0
61 ED=9.0/8.4/3.4
30 ED=9.0/8.4/3.8
1 ED=9.0/8.4/4.0
5 ED=9.0/8.4/6.7
1 ED=9.0/8.4/7.2
1 ED=9.0/8.4/8.3
104 ED=9.0/8.5/2.6
5 ED=9.0/8.5/3.0
29 ED=9.0/8.5/3.4
15 ED=9.0/8.5/3.8
1 ED=9.0/8.5/4.0
1 ED=9.0/8.5/6.7
1 ED=9.0/8.5/7.1
8 ED=9.0/8.6/2.6
3 ED=9.0/8.6/3.0
5 ED=9.0/8.6/3.4
8 ED=9.0/8.6/3.8
1 ED=9.0/8.6/6.7
3 ED=9.0/8.7/2.6
2 ED=9.0/8.7/3.4
3 ED=9.0/8.7/3.8
19 ED=9.0/8.8/2.6
5 ED=9.0/8.8/3.4
3 ED=9.0/8.8/3.8
127 ED=9.0/8.9/2.6
7 ED=9.0/8.9/3.0
59 ED=9.0/8.9/3.4
33 ED=9.0/8.9/3.8
1 ED=9.0/8.9/6.7
1 ED=9.0/8.9/7.1
2189 ED=9.0/9.0/2.6
97 ED=9.0/9.0/3.0
1894 ED=9.0/9.0/3.4
328 ED=9.0/9.0/3.8
7 ED=9.0/9.0/4.0
56 ED=9.0/9.0/6.7
2 ED=9.0/9.0/7.1
7 ED=9.0/9.0/7.2
4 ED=9.0/9.0/8.3
1 ED=9.0/9.0/8.5 *
125 ED=9.1/2.0/2.0
6 ED=9.1/2.6/2.6
3 ED=9.1/2.8/2.6
1 ED=9.1/3.0/2.6
16 ED=9.1/3.4/2.6
3 ED=9.1/3.4/3.0
5 ED=9.1/3.4/3.4
6 ED=9.1/3.6/2.6
1 ED=9.1/3.6/3.6
8 ED=9.1/3.8/3.0
19 ED=9.1/3.8/3.4
48 ED=9.1/3.8/3.8
8 ED=9.1/4.0/2.6
1 ED=9.1/4.0/3.4
1 ED=9.1/4.0/3.8
1 ED=9.1/4.0/4.0
2 ED=9.1/4.2/4.0
3 ED=9.1/6.6/2.6
1 ED=9.1/6.6/3.0
1 ED=9.1/6.7/2.6
123 ED=9.1/7.1/2.6
9 ED=9.1/7.1/3.4
1 ED=9.1/7.1/3.8
1 ED=9.1/7.1/6.7
1 ED=9.1/7.1/7.1
36 ED=9.1/7.2/2.6
4 ED=9.1/7.2/3.4
1 ED=9.1/7.2/3.8
48 ED=9.1/7.3/2.6
1 ED=9.1/7.3/3.0
6 ED=9.1/7.3/3.4
1 ED=9.1/7.3/3.8
2 ED=9.1/7.3/6.7
1 ED=9.1/7.3/7.2
1 ED=9.1/7.4/2.6
2 ED=9.1/7.5/2.6
4 ED=9.1/7.8/2.6
2 ED=9.1/7.8/3.0
1 ED=9.1/8.2/2.6
2 ED=9.1/8.2/3.4
20 ED=9.1/8.3/2.6
11 ED=9.1/8.3/3.4
8 ED=9.1/8.3/3.8
1 ED=9.1/8.3/6.7
24 ED=9.1/8.4/2.6
27 ED=9.1/8.4/3.4
11 ED=9.1/8.4/3.8
2 ED=9.1/8.4/6.7
1 ED=9.1/8.4/7.1
2 ED=9.1/8.4/7.2
16 ED=9.1/8.5/2.6
1 ED=9.1/8.5/3.0
6 ED=9.1/8.5/3.4
3 ED=9.1/8.5/3.8
1 ED=9.1/8.5/4.0
4 ED=9.1/8.6/2.6
1 ED=9.1/8.6/3.0
1 ED=9.1/8.6/3.4
2 ED=9.1/8.6/3.8
3 ED=9.1/8.7/2.6
2 ED=9.1/8.8/2.6
1 ED=9.1/8.8/3.4
38 ED=9.1/8.9/2.6
3 ED=9.1/8.9/3.0
4 ED=9.1/8.9/3.4
9 ED=9.1/8.9/3.8
1 ED=9.1/8.9/4.0
1 ED=9.1/8.9/6.7
119 ED=9.1/9.0/2.6
5 ED=9.1/9.0/3.0
42 ED=9.1/9.0/3.4
30 ED=9.1/9.0/3.8
1 ED=9.1/9.0/4.0
4 ED=9.1/9.0/6.7
1 ED=9.1/9.0/8.8
1 ED=9.1/9.0/9.0
815 ED=9.1/9.1/2.6
30 ED=9.1/9.1/3.0
512 ED=9.1/9.1/3.4
193 ED=9.1/9.1/3.8
4 ED=9.1/9.1/4.0
28 ED=9.1/9.1/6.7
1 ED=9.1/9.1/8.4 *
55 ED=9.2/2.0/2.0
1 ED=9.2/2.6/2.6
1 ED=9.2/2.8/2.6
8 ED=9.2/3.4/2.6
1 ED=9.2/3.4/3.0
8 ED=9.2/3.4/3.4
1 ED=9.2/3.6/2.6
1 ED=9.2/3.8/3.0
8 ED=9.2/3.8/3.4
10 ED=9.2/3.8/3.8
2 ED=9.2/4.0/2.6
3 ED=9.2/4.0/3.4
1 ED=9.2/4.0/3.8
1 ED=9.2/4.2/2.6
1 ED=9.2/6.7/2.6
46 ED=9.2/7.1/2.6
7 ED=9.2/7.1/3.4
2 ED=9.2/7.1/3.8
2 ED=9.2/7.1/6.7
16 ED=9.2/7.2/2.6
6 ED=9.2/7.2/3.4
1 ED=9.2/7.2/3.8
22 ED=9.2/7.3/2.6
4 ED=9.2/7.3/3.4
1 ED=9.2/7.3/3.8
1 ED=9.2/7.3/7.2
2 ED=9.2/7.8/2.6
1 ED=9.2/7.8/3.0
6 ED=9.2/8.3/2.6
2 ED=9.2/8.3/3.4
4 ED=9.2/8.4/2.6
10 ED=9.2/8.4/3.4
1 ED=9.2/8.4/3.8
2 ED=9.2/8.4/4.0
1 ED=9.2/8.4/6.7
2 ED=9.2/8.4/7.2
11 ED=9.2/8.5/2.6
1 ED=9.2/8.5/3.4
1 ED=9.2/8.5/3.8
1 ED=9.2/8.6/2.6
1 ED=9.2/8.7/3.4
1 ED=9.2/8.8/2.6
1 ED=9.2/8.8/3.4
6 ED=9.2/8.9/2.6
1 ED=9.2/8.9/3.0
3 ED=9.2/8.9/3.4
4 ED=9.2/8.9/3.8
33 ED=9.2/9.0/2.6
1 ED=9.2/9.0/3.0
10 ED=9.2/9.0/3.4
2 ED=9.2/9.0/3.8
3 ED=9.2/9.0/6.7
39 ED=9.2/9.1/2.6
4 ED=9.2/9.1/3.0
9 ED=9.2/9.1/3.4
9 ED=9.2/9.1/3.8
1 ED=9.2/9.1/4.0
1 ED=9.2/9.1/6.7
606 ED=9.2/9.2/2.6
17 ED=9.2/9.2/3.0
335 ED=9.2/9.2/3.4
95 ED=9.2/9.2/3.8
4 ED=9.2/9.2/4.0
30 ED=9.2/9.2/6.7
1 ED=9.2/9.2/7.8
1 ED=9.2/9.2/8.2
1 ED=9.2/9.2/8.9 *
5 ED=9.3/2.0/2.0
1 ED=9.3/3.8/3.8
1 ED=9.3/4.0/4.0
4 ED=9.3/7.1/2.6
1 ED=9.3/7.1/6.7
5 ED=9.3/7.2/2.6
7 ED=9.3/7.3/2.6
2 ED=9.3/8.3/2.6
1 ED=9.3/8.5/3.4
1 ED=9.3/8.5/6.7
1 ED=9.3/8.9/2.6
2 ED=9.3/9.0/2.6
3 ED=9.3/9.1/2.6
1 ED=9.3/9.1/6.7
9 ED=9.3/9.2/2.6
3 ED=9.3/9.2/3.4
1 ED=9.3/9.2/3.8
1 ED=9.3/9.2/6.7
186 ED=9.3/9.3/2.6
2 ED=9.3/9.3/3.0
79 ED=9.3/9.3/3.4
11 ED=9.3/9.3/3.8
1 ED=9.3/9.3/4.0
9 ED=9.3/9.3/6.7 *
1 ED=9.4/2.0/2.0
1 ED=9.4/3.4/2.6
1 ED=9.4/7.1/2.6
1 ED=9.4/7.2/2.6
1 ED=9.4/8.3/3.4
1 ED=9.4/8.4/7.2
1 ED=9.4/9.0/2.6
1 ED=9.4/9.1/2.6
1 ED=9.4/9.2/2.6
1 ED=9.4/9.2/3.8
58 ED=9.4/9.4/2.6
15 ED=9.4/9.4/3.4
3 ED=9.4/9.4/3.8
1 ED=9.4/9.4/4.0
5 ED=9.4/9.4/6.7 *
1 ED=9.5/9.2/3.8
7 ED=9.5/9.5/2.6
3 ED=9.5/9.5/3.4 *
1 ED=9.6/9.3/2.6
2 ED=9.7/9.7/2.6
2 ED=9.7/9.7/3.4 *
1 ED=9.8/9.8/2.6
1 ED=9.9/9.9/2.6

late comment on that final status of mladen's generation for the game 262

I transposed some old code for a scan in my current working code.

I checked the order of magnitude of the task for a full scan for that pattern.

I filled the box 1, then expanded band 1 stack 1 and solved the redundancy using
gridchecker --pattern --patcanon < in.txt > out.txt

I got 721 ED partial puzzle

I then made a second step extending the cells marked 'x' here below.
At that point, cells marked '1' are not expanded

Code: Select all

.98.....7
7.6..9...
543......
...1...1.
....x.x.x
.8...1...
....x.1..
...1...1.
9...x...1

after grid checker reduction, we have 13 239 986 ED partial puzzles.

This is a very short primary step (says some minutes);
At that point, residual symmetry is peanut
so the scan can be expressed as expand 7 cells on a little more than 13 millions seeds with 16 cells known.

I am not very optimistic on the result, but I'll test a sample to see how long it could take.

EDIT wrong reading of my screen for the count : 13 millions, not 132 millions

[champagne]

Not a full scan, it is slow. Regular +/-.
72,228,251 puzzles in total.
68 runs, about 300 generations, some of them contributing only few puzzles.

missing here to have a wider view the rough power consumption ( cores*hours and the clock speed)
I on my side used an i7 about 3 days to generate 7 million puzzles, only 10% of mladen's generation

Not a full scan, it is slow. Regular +/-.
Below is the puzzles count by serate.

interesting, but in terms of strategy, I would be interested to have the count of sub puzzles (ideally with my 16 clues giving 12 millions ID partial puzzles)

My idea is the following :

a) A full scan remains out of any realistic approach
b) most of the 12 millions partial puzzles will give no valid puzzle (what we experienced with a small number of clues)

a possible "mixed strategy" would be
_ to expand only sub puzzles(once and only once) where one puzzle has been seen using other methods (vicinity for example)
_ to look specifically for new seeds limiting the vicinity to the sub puzzles cells

Slightly more complex to organize, but likely not more expensive that the wide vicinity made by mladen

[dobrichev]

A pattern with 7 givens in a house and small number of total clues is suitable for full generation of course, because you can fix significant part of the givens.
Unfortunately I have no tool that efficiently generates puzzles from a fixed pattern, exploits the symmetries, and allows parallelism. Have plans to do it but recently there is no time for hobbies.

The large number of puzzles in the game came from the {-1,+1} enclosure, which is proven to give garbage. Maybe except these strange rare puzzles with medium rating.
The only 10+ puzzle came in the first hours, before the brute expansion, while still reasonable seed filters were used and the total population was in the order of one million.

During the generation much operations are unnecessary repeated but usually this is not critical because even then the skfr rating consumes > 80% of the resources. Focus is rating not to be repeated for the same puzzle.

One lesson to me from the last game was that with 8 GB of RAM, 64-bit platform, and 70 millions puzzles in cache, the tool (gridchecker) consumes < 4GB at the start of the pass but close to 8 GB at the end and this is a sign of unnecessary large temporary storage and bad synchronization of the results between threads.

[champagne]

Unfortunately I have no tool that efficiently generates puzzles from a fixed pattern, exploits the symmetries, and allows parallelism.

What I described to reach the point where I had 12 millions partial puzzles in hand is surely something you can do.

I used there gridchecker to reduce the count through symmetries.

Parallelism is then possible just splitting the partial puzzles in lots, but again, I don't think that even with such a pattern, the full scan is realistic(I have 2 chunks running to make a rough forecast of the overall power need for a full scan)

I hope to have within 2 days the results of that test.

[champagne]

the results of my test are not too far from what I had in mind.

It took more than on day to scan 1% of the 12 millions partial patterns using 2 cores.
A little better is the ratio of active partial patterns: about 15% of the total
active seeds gave in average 24 puzzles.

Nothing that could push in that way at the start of the game, the vicinity looks better, but some possibilities to explore to find new areas of vicinity.

[champagne]

Hi Champagne,

Congratulations for these rare ratings.
How they appeared? Are you doing massive rating of all puzzles in your cache at some stage?

I stopped using the partial rating - those good filters you implemented in skfr - because I found that almost always at some stage I am doing complete rating.
But now I have 38 millions of puzzles living on the shortest paths between the hardest two, and rating them is estimated to take at least 200 hours on my machine.

Hi malden,

let me answer here.

I start always with a quick "group" rating splitting the file in several lots

a)ER rating below 4.5
b)ER ratings >=4.5 <6.2
c)ratings passing serate dynamic (somewhere 9.2 and above, but dynamic can reach 9.6)

in between, the rating is done by groups (example all XY chains are rated 7.0) and the split is

- diamonds (group diamond)
- pearls
- others

my experience is that the lot >=4.5 <6.2 is very small not much more than 1% of the total.
That lot is fully rated using skfr and then submitted to serate for puzzles of interest.
There are located ratings 5.1 and 6.1, very rares

Just another remark for recent patterns.
The jellyfish appears 66 times in mike's file for rarity;
I don't remember having seen it since we started pattern with usually 23_26 clues

[dobrichev]

I start always with a quick "group" rating splitting the file in several lots

What is the magic that places the puzzle in the appropriate "group" w/o complete rating?
I can imagine that when ER reaches hard criterion (~9.2) you put the puzzle in the "dynamic" group w/o performing complete rating, but what are you doing with this puzzle later?
For the lower groups you must do complete rating for the classification, right?
Or do you mean partial rating by ED? Then the question what are you doing later remains.

[champagne]

I can imagine that when ER reaches hard criterion (~9.2) you put the puzzle in the "dynamic" group w/o performing complete rating,
but what are you doing with this puzzle later?

Simply rating it using skfr as a first step.

What is the magic that places the puzzle in the appropriate "group" w/o complete rating?

May be the rating sequence can help

Code: Select all

if (myd.Std_HiddenSingles(3))continue;
if (work1.LockedSet20_25(2))  goto DoElims;
if (work1.LockedSet20_25(3))  goto DoElims;
if (LockedSetNaked(2))    goto DoElims;
if (XW(2))    goto DoElims;
if (LockedSetHidden(2))  goto DoElims;
if (LockedSetNaked(3))  goto DoElims;
if (XW(3)) goto DoElims;
if (LockedSetHidden(3)) goto DoElims;
if (wpaires.XYWing())    goto DoElims;
if (wpaires.XYZWing())    goto DoElims;
if (wpaires.UR_Serate()){}
if (wpaires.UL()){}
if (LockedSetNaked(4)) { Quickrate(50); goto DoElims; }
if (XW(4)){ Quickrate(52); goto DoElims; }
if (LockedSetHidden(4)) { Quickrate(54); goto DoElims; }
if (wpaires.DoBUG())   {}
if (AlignedPair()) { Quickrate(62); .. }


if (XChainfast()) { Quickrate(65); .. }
xysearch.Init(1);    if (hint.rating<100) { Quickrate(70)... }
if (AlignedTriplet_Alt())  { Quickrate(75); .. }
xsearch.GoNishiofast(); if (hint.rating <100) { Quickrate(76);.. }.
xysearch.MAICs(1); if (hint.rating <100) { Quickrate(80); .. }
xysearch.DynamicChainsfast(); if (hint.rating <100)  { Quickrate(85); .. }


As you can se, below "Xchain", the rating is not too far from the skfr sequence except that it is simplified for UR ULs and BUGs
Later on, only 6 ratings are recognized corresponding to serate families

65 Xchains level (including XY loops
70 XY chains level
75 aligned triplets
76 Nishio
80 multi chains (kraken)
85 XY dynamic


The diamond/pearl classification refers to theses ratings

but what are you doing with this puzzle later?
For the lower groups you must do complete rating for the classification, right?
Or do you mean partial rating by ED? Then the question what are you doing later remains.

For the 2 lower groups ( {below 4.5} and {>=4.5<6.2}) a kind of skfr final rating except that I just keep in the lower group the highest rating for each ER.

For higher ratings, I first concentrate on diamonds, later on pearls.

Usually, after 2 or 3 cycles all ER are touched except sometimes 8.1 and 6.5, maybe 6.5
Then, I have to analyze the "65" and "76" group, the smallest part of the rest.

[Pat]

Once again, we can see that you search in another area than me

so i found some oddities

( 5.2/3.6/3.6 , 5.9/3.2/3.0 , 6.5/3.8/2.8 , 6.5/3.8/3.4 )

but you found the amazing 5.0/5.0/2.6

in the vicinity of 10.x ??

[champagne]

Once again, we can see that you search in another area than me

so i found some oddities

( 5.2/3.6/3.6 , 5.9/3.2/3.0 , 6.5/3.8/2.8 , 6.5/3.8/3.4 )

but you found the amazing 5.0/5.0/2.6

in the vicinity of 10.x ??

I don't track the source of puzzles, so I can't answer to your question.

In fact, I just noticed that in several recent games, in the "low ER's ", you got several ratings not in my cache, despite the huge number of puzzles generated on my side.

I suspect that, if you look for high ratings, you have a lower chance to catch such puzzles.