Calcudoku puzzle forum
http://www.calcudoku.org/forum/

Simple curiosities
http://www.calcudoku.org/forum/viewtopic.php?f=2&t=78
Page 1 of 1

Author:  jomapil  [ Mon Oct 10, 2011 3:16 pm ]
Post subject:  Simple curiosities

There are much people such as me that have curiosities and so I ask them to put the questions here. For my part I'd like anyone answers the 2 following questions:

1 - 6 or 7 years ago I bought a program named kakuroworks about of course kakuro. One of the qualities is the uniqueness of the solution as in Calcudoku. How are sure the programmers of the uniqueness of the solution?

2 - How is the criterion the programmers use to attribute the degree of difficulty ( easy, medium or difficult )?

Author:  pnm  [ Mon Oct 10, 2011 10:55 pm ]
Post subject:  Re: Simple curiosities

jomapil wrote:
1 - 6 or 7 years ago I bought a program named kakuroworks about of course kakuro. One of the qualities is the uniqueness of the solution as in Calcudoku. How are sure the programmers of the uniqueness of the solution?

You may also want to ask this question in the "describe your solver" thread.
In my case, my solver simply finds all solutions to a puzzle, and if there is more than one,
it is rejected.
jomapil wrote:
2 - How is the criterion the programmers use to attribute the degree of difficulty ( easy, medium or difficult )?

That is a bit more involved, and a trade secret :)

Patrick

Author:  clm  [ Tue Oct 11, 2011 7:40 pm ]
Post subject:  Re: Simple curiosities

pnm wrote:
jomapil wrote:
1 - 6 or 7 years ago I bought a program named kakuroworks about of course kakuro. One of the qualities is the uniqueness of the solution as in Calcudoku. How are sure the programmers of the uniqueness of the solution?

You may also want to ask this question in the "describe your solver" thread.
In my case, my solver simply finds all solutions to a puzzle, and if there is more than one,
it is rejected.
jomapil wrote:
2 - How is the criterion the programmers use to attribute the degree of difficulty ( easy, medium or difficult )?

That is a bit more involved, and a trade secret :)

Patrick


Sometimes I have found “sudokus” with more than one solution though never “kakuros”. I suppose each programmer has his own methods to guarantee the uniqueness of the solution (as the Patrick’s one for the calcudokus). However, if a solution is correct, although not the “official” solution, nothing is wrong if not rejected and the points assigned to the solver.

The degree of difficulty… that is a very interesting subject. First of all, I am sure there is a lot of subjectivity on that, let’s say that some people is better with the 5’s and some with the 7’s … and that’s why I suggested some time ago to ask (i.e., in the green screen, in the final "OK" window or similar): “Assign a value from 1 to 100 percent to the puzzle just solved” in order to collect some information that may be compared with “My solver rating for this puzzle:” . This last rating sometimes is difficult to understand, i.e., what is the meaning of 130.1?, or 2.7? and, what are the lower and upper limits of this parameter?.

In my opinion, to be objective, a good system for measuring the difficulty is assigning a number based on the time consumed by the computer in finding a solution to that puzzle (an average time in three attempts, for instance). I have played chess with computer programs having, for instance, 16 levels of difficulty and, in fact, apparently, the first 8 levels were practically the same, then the question arised: How is the criterion? (I suppose a good chess program must have extensive libraries of apertures, finals, “master” games, etc., and the depth of the analysis proportional to the level selected).

Now, how to create a difficult puzzle?. As in chess, I would put a time limit to the computer to “think”. Other criterion, in the case of calcudokus, could be to configure the puzzle with higher cages (less number of them in the puzzle) (I suggested this a little bit in my post “Colouring the cages”). Finally, with a really high percentage of the info of the calcudoku being “essential” (I was talking the other day about this in “The essential info in a puzzle”) (jomapil, you have probably observed that the more difficult kakuros have "wider" intersections, and in the easy kakuros the intersection is just a cell).

Page 1 of 1 All times are UTC + 1 hour [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/