jeudi 1 septembre 2016

Tic-Tac Sudoku

Rules of sudoku state that a set of N different symbols has to be placed in a N*N grid which is divided into N regions (squared or irregular), such that each symbol appears exactly once in each row, column and region.
Variants of sudoku normally keep this basic rules, while adding new ones which act as additional constraints on where can be placed the symbols. These rules have also been adapted for geometrical variations : cube sudoku, sudokurve, star sudoku, etc.. where the notion of rows and columns have been generalized.
In the past, these rules have also been slightly changed, for example the set of symbols contains one symbol which is repeated twice, and adding the non-touching rule for that symbol makes the link with star battle puzzle. Surplus and deficit sudoku, with region of size N+1 or N-1 appeared. In both cases, even if the puzzles are not strictly obeying the sudoku rules, some strong arguments can be found that these are appropriate in sudoku tournaments, provided that they constitute only a tiny part of the competition. That's a way to put a foot outside the limit of sudoku field while keeping the other inside, what someone may call the « grey zone ».
Some authors and organizers understand this sign as « It's not important that WSC puzzles obey the basic rules of sudoku ». After all, if these « grey zone » puzzles are accepted, why should we care about the fact that sudoku has quite clear rules. They felt allowed to present puzzles where the set of symbols contains lot of repetitions (1234 in round 2 of WSC 2015), or which don't have regions (only a few cells having local constraint seems sufficient to distinguish from latin square).

In the past, when the basic sudoku rules were clearly not followed (here I don't speak about the « grey zone »), there was a clear reaction from the sudoku community. Here I want to take the example of the variant presented in the final of WSC 2009 (read my previous post « Big brother's revenge »). Or some latin squares in the UK sudoku qualifiers 2011. I think the dominant thinking of the sudoku community was that they are not sudoku, hence should not appear in sudoku competitions. « Getting a non-sudoku in the finals of a Sudoku championship is unacceptable. » The 4th round of WSC 2015 was constituted by a puzzle which has big similarities with the puzzle that appears in WSC 2009 finals, particularly they share the same lack of regions (more than 50% of cells don't even have a local constraint in both cases). But this year, I strongly feel that the dominant thinking of sudoku community is that these puzzles are appropriate in sudoku competition or at least in WSC, contradicting the 2009 opinion. « it is still just within the limits of what I think is possible for a WSC. » or « the puzzle in the Round 4 at WSC 2015 was appropriate for the WSC. » is what I read most of the time. However nobody clearly states or has strong arguments that these are sudoku. If we now accept these kind of puzzles, the question we have to clarify now is : Either they are sudoku, then we should explain why some puzzles that weren't sudoku in 2009 suddenly entered in the sudoku category of puzzles. Either they are not sudoku, then we should explain why it is good to accept non-sudoku in WSC.

Now I would like to invite you to think about a new sudoku variant, following the new thinking tendency that basic sudoku rules are not important. We can now accept that symbols can be repeated, and regions are not important, the important is that there is some kind of local constraint, right ? I suggest a square grid containing N*N cells, without region (we'll add a local constraint afterwards). Let's just have 2 symbols, 0 and 1 for example, each repeating N/2 (thus N has to be even) times in each rows and columns. Now about the local constraint, let me think about it.... mmmh, I have an idea : we can add the rule : There are no more than two consecutive 0's or 1's in a row or a column. Nice, isn't it ?
Wait a minute, I know this variant. Oh Yes, if we add the rule « All rows and all columns are unique. » it's the conceptis game Tic-tac-logic (also named binoxxo in some countries).
I stupidly thought for years that Tic-tac-logic was not a sudoku. It seems that in contrary it is !

Become serious again during a minute. Of course, non-sudoku puzzles ideas can be introduce in sudoku. If we take the example of Tic-tac-logic, we can create a variant where the idea « no more than two consecutive ... in row and column » is present. In fact, this variant already exists, it is the well known « 2 even 2 odd sudoku ».
Playing Tic-tac-logic is very similar to determine the parity of digits in a 2 even 2 odd sudoku, but, frankly speaking, if you had to say which of the 2 puzzles is a sudoku variant, would you have some hesitations ?

Aucun commentaire: