ab wrote:Some compilers have a theory that certain distributions of clues, such as diagonal type patterns, lead to more hard patterns. It may be that they are finding these patterns here because that's where they're searching for them, ie if they decided that patterns with pairs of clues next to each other had hard puzzles in them and started concentrating their searches there, they'd also find many hard puzzles.
looking in the
Patterns Game Results (
using Sudoku Explainer 1.2.1 ratings ),
consider these 20-clue patterns
which are very similar
( differing only by shifting the rows by 1
and the columns likewise by 1 )
yet very opposite --
the one being very low in diagonality,
the other being entirely diagonal (
every 2 clues sharing a box never share a row or a column )
gsf wrote:game 49
highest rating = 9.1
- Code: Select all
. . . . . . 1 2 .
. . . 3 9 . . . .
. . 4 . . . 5 6 .
. 7 . . . . . . 4
. 9 . . 8 . . . 6
. . . . . 2 . . .
5 . 9 . . . 7 . .
2 . 1 . . . . . .
. . . 8 4 . . . .
JPF wrote:game 57
highest rating = 10.4
- Code: Select all
. . 1 2 . . . . .
. 3 . . . 4 5 . .
6 . . . . . . 7 .
1 . . 7 . . . 2 .
. . . . 5 . . . .
. 8 . . . 9 . . 3
. 5 . . . . . . 4
. . 7 6 . . . . .
. . . . . 3 9 . .