The New Sudoku Players' Forum

[dobrichev]

Command is

Code: Select all

./pg --pattern --verbose --scanfor ......3....9.....5.4.2...6.8...9.7...1.3.4.2...4.5...6.8...1.7.5.....9....2...... < symmetrical_grids > seed

where

• pg is the gridchecker compiled with patterns game extensions (skfr, etc). For windows it could be gridchecker.exe.
• --scanfor xxx gives the pattern (number positions define the pattern, they could be all ones)
• symmetrical_grids is this file (only leftmost 81 characters of the line matter)
  
  Hidden Text: Show

  Code: Select all

  123456789456789123789123456231564897564897231897231564312645978645978312978312645 # 648 001 964550
123456789456789123789123456231564897564897231897231564312978645645312978978645312 # 108 001 964558
123456789456789123789123456231564897564897231897231564348672915672915348915348672 # 54 001 964568
123456789456789123789123456231564897564897231897231564348915672672348915915672348 # 54 001 964569
123456789456789123789123456231564897564897231978312645312645978645978312897231564 # 36 001 964589
123456789456789123789123456231564978564897312897231645312978564645312897978645231 # 72 001 965096
123456789456789123789123456231564978564978231978231564312645897645897312897312645 # 108 001 965174
123456789456789123789123456231564978564978231978231564312897645645312897897645312 # 36 001 965182
123456789456789123789123456231564978645897231897312564312645897564978312978231645 # 36 001 966422
123456789456789123789123456231597864597864231864231597312678945678945312945312678 # 36 001 985543
123456789456789123789123456231597864597864231864231597312948675675312948948675312 # 36 001 985544
123456789456789123789123456231678945678945231945231678312597864597864312864312597 # 36 001 989378
123456789456789123789123456231678945678945231945231678312894567567312894894567312 # 36 001 989379
123456789456789123789123456231897564564231897897564231348672915672915348915348672 # 27 001 989400
123456789456789123789123456231897564564231897978645312312978645645312978897564231 # 36 001 989404
123456789456789123789123456231897645564312897978564231312978564645231978897645312 # 36 001 989622
123456789456789123789123456234567891567891234891234567318642975642975318975318642 # 54 001 992761
123456789456789123789123456234567891567891234891234567345678912678912345912345678 # 54 001 992769
123456789456789123789123456234567891567891234891234567372615948615948372948372615 # 54 001 992773
123456789456789123789123456235964817817235964964817235392641578578392641641578392 # 54 001 1006748
123456789456789123789123456267591348591834672834267915375618294618942537942375861 # 54 001 1007164
123456789456789123789123456267591834591834267834267591375618942618942375942375618 # 162 001 1007168
123456789456789123789123456267591834591834267834267591375942618618375942942618375 # 54 001 1007169
123456789456789123789132465231564978564978231978213546312645897645897312897321654 # 36 004 50092115
123456789456789123798132465231564897564897231879213546312645978645978312987321654 # 36 009 224283635
123456789456789123897231564231564978564978231789312645312645897645897312978123456 # 54 016 473714886
123456789456789231789123645231564897564897312897231456375618924618942573942375168 # 36 027 1012364857
123456789456789231789123645231564978564897123897231564375942816618375492942618357 # 36 027 1012366524
123456789456789231789312456231564978564978312978123564312645897645897123897231645 # 27 030 1061944621
123456789456789231789312456231645978645978312978123645312564897564897123897231564 # 54 030 1062073983
123456789457189236968372514291738465374265198685941327546813972732694851819527643 # 36 133 4869572229
123456789457189326869372514214965873635847192978213465381624957546798231792531648 # 36 240 5449015463
123456789457289163689173452235964817816735924974812635392641578568397241741528396 # 54 348 5472677656
123456789457289163689173452245968317316745928978312645564897231731524896892631574 # 108 348 5472677695
123456789457893612986217354274538196531964827698721435342685971715349268869172543 # 72 416 5472730538


• file output is
  

  Code: Select all

  ......1....1.....2.2.3...4.5...1.6...7.8.9.2...4.6...9.8...2.7.1.....5....3......


• terminal log is
  

  Code: Select all

  Symmetry Level   2
...................................

Total time 349.116 seconds.


The code is in https://github.com/dobrichev/gridchecker/blob/master/patterns.cpp, function int scanfor(const char *p)
Looking at the loops, the pattern automorphisms matter and grid automorphisms don't. I.e. for this particular pattern the number of checks is halved due to the pattern, and similar timings are expected for any other solution grid for this pattern.

My choice to check against these special solution grids is arbirtary and is inspired by a finding that once Mike played a 10.3/10.3/5.4 puzzle having MC solution.

[champagne]

in other words, if I got it, you use a limited list of solution grids to find seeds
And for each solution grid, you likely test all perms of the game pattern (say 1296 x 2)

It's clear that in the game, you had to find quickly alternative seeds to the given to search in other places.
I mostly used the symmetry of given to produce quickly a wide set of puzzles, some not minimal, some with multiple solutions .

in the last runs of the game, most of the key positions (diamonds) were posted in the first day, and the chances to win the game was very low if you missed these key positions

[dobrichev]

in other words, if I got it, you use a limited list of solution grids to find seeds

Exactly.

in the last runs of the game, most of the key positions (diamonds) were posted in the first day, and the chances to win the game was very low if you missed these key positions

In this direction I am considering a rule change where the player manages a private queue of potential submissions. The server, on each turn, automatically picks the first playable item, skipping w/o penalty the items in the queue that cannot be played due to previous own or opponent's submissions.

[m_b_metcalf]

In this direction I am considering a rule change where ...

Fine.When are you planning to modify g.r.emlin?

[dobrichev]

Replacing the referee is possible. Actions can be performed by a remote agent acting on behalf of a regular user.
Several important unresolved issues remain:
- Forum reliability.
- The game's loss of popularity even before the forum problems arose.
- Any rule change would destroy the global rarity bonuses.

For a completely new game, I would prefer the following:
- Keeping the concept of ratings, trumping, uniqueness, providing the initial easy puzzle that defines the pattern, submission frequency, as well as the penalties for an invalid puzzle.
- Replacing the rating tool, for example with a fixed version of SKFR, accessible for use by scripts from different operating systems.
- Making the template symmetry optional.
- Making the minimality requirement optional and chosen by the game creator.
- "More" puzzles to be limited to only one per ER and a rating higher than the best corresponding ER rating at game close.
- Removing rarity bonuses until the game is eventually stabilized.
- Modify the "turns" concept with the introduction of a queue for each player. The server cycles through the players' queues and for each player chooses what to play, ignoring currently disallowed puzzles without penalty and feeding back to the player's queue the reason the puzzle was ignored (does not improve, self-trump, morph of an already played puzzle).
- Existence of a machine-friendly interface backed by a user-friendly web interface.

[champagne]

In my current work:

Finding a set of "not easy puzzles" some minutes
Rating the file : tens of hours with skfr...

Where I am, a new clone of Serate 10 times fatster than SKFR seems possible and of interest...