Indeed , the so called random grids made from random puzzles are biased to a degree in the number of U4 unavoidables.
A lot of work was done
hereI did a few calculations some time ago [in the min clues thread]!!!
coloin wrote:have run a solver on a few grids of the following pattern [with different B2-B9]:
- Code: Select all
123 --- ---
456 1-2 ---
789 --- -12-
--- -2- --1
-1- --- 2--
--2 --1 ---
--1 --- -2-
2-- -1- ---
--- 2-- 1--
coloin wrote: The analysis take 9 hours each! [Hopefully the B1 filling doesnt introduce bias - I cant see any]
Of the five I have done I got solution rates of approx.
2100,000,000
1220,000,000
1200,000,000
1080,000,000
989,000,000 {this is canonical - 1/2 in different row and column}
I dont have the actual grids pertaining here ! - But it is not a difficult "stunt" - very possible with suexk2.exe
I will run it with a few selected 2-rookeries [now we know what they are !]
Is there is a significant difference between the sol counts....?
I stumbled on this some time ago but didnt have the insight that we do now
hereon page 4 of this thread Viggo wrote
- Code: Select all
Grid name: top1 top2 top3 ran1 ran2 ran3 SF SFB
Number of 2-permutable: 20 19 26 18 23 25 28 36
Number of 4-permutable: 13 13 9 16 9 8 8 0
Number of 8-permutable: 2 4 1 2 3 3 0 0
Number of 16-permutable: 1 0 0 0 1 0 0 0
Sum of solutions all 2-rookeries: 124 122 96 116 122 106 88 72
Sum of solutions all 3-rookeries: 3414 3384 2322 3252 3222 2574 1818 1194
Max numb. solutions of 3-rookeries: 360 258 252 264 264 120 102 54
Number of minimal 3-rookeries: 7 5 10 2 5 11 15 39
MCN: 12 10 10 10 12 10 9 8
I think he means
Sum of solutions all 2-rookeries: ~ Sum of solutions all 7-rookeries:
Sum of solutions all 3-rookeries: ~ Sum of solutions all 6-rookeries:
- Code: Select all
...749568456.8.9.77896.54....8567.4956.498.7.947...68587495...66.5.7489..9.8.6754 48 sol.
...849567456..78.9789.56..464.5987..97.46..855.8.7.946.9768.45.8657.4.9...49.5678 60 sol.
.46..8975785.96.4...945786.49.56.7.8..897.654657.84..95648.9..78..7.549697.64.58. 6 sol.
.4798..5658.476.9..695..874896.45..7754..896....7694859.5..764847869.5..6..8547.9 72 sol.
..568.97464..79.587894.56....49687.596.5.784.857.4..964.6.9.587.7.854.695987.64.. 12 sol.
4.6..5987.756894..98..476.5.9.7.85468475.6.9.56..947.8.5987..646.495.87.7.846..59 30 sol.
8569..7.47..456.9849.8.7.56.48.9.6759.5764.8..67.854.9.7.54896.684.795..5.96..847 12 sol.
86.95..74.95.746.84.768.59..498.7.56..64.5789578.964..6..7498.575..6894.9845...67 18 sol.
..89.4576596.7.8.4.74586..97856..94..4..5968796.847.5.4..768.956594..7.88.7.9546. 12 sol.
.4.58976.675.4.98.8.9.675.44678...5959.6.4.78.8.975.467.8..64959.645.8.7.547986.. 96 sol.
these are 10 random 6-rookeries, with the sol. count which is the perm. count of the missing 3-rookery.
To note each completion has 6 isomorphic clue combinations.
And the 6 sol. must be a minimal 3-rookery, equivalent to an unentwined 3-rookery.
RW wrote:According to the numbers, random picks of 36 rookeries should give:
which numbers ?.......[not mine hopefully !]
my random 2-rookeries were taken from 2700 random 17-puzzles [one from each puzzle]
C