You are here

Sudoku GP 2d round

Hi All

At first thank you for all the rounds , it is a lot of work.

Why to choose a "Hybrid-sudoku---> battleship" to give most points in this second round of sudoku ?
There are enough variations of sudokus (without variation of puzzles) to propose one with an equivalent value of points .... No ?

I agree with Distri : In my opinion the battleship sudoku was a non-sudoku puzzle, or a « not-only-sudoku » puzzle. Hence I was not happy to see such a variant in a sudoku competition.

It's quite natural that latin squares have their sudoku version, like kropki sudoku and skyscrapers sudoku. Last year, the sudoku GP contained 2 star battle sudoku, which is an hybrid between sudoku and another well-known puzzle : star battle. Star battle is a cousin of latin square, and I can see for what reasons it was accepted as reasonable variant of sudoku (one can argue that it's just one added symbol that repeat twice in each set and that don't touch itself), despite the fact that I didn't really understand why one was choosen to be the high-pointer of the round (105 points).

Battleship is far to be a latin square, the logic of this puzzle is far from the logic of sudoku.

When, before the tournament, I wrote that I thought it was a bad idea to put such a puzzle in a sudoku tournament, The answer of some experienced players were :

« You can't judge before the tournament, perhaps the actual puzzle will have more sudoku logic than battleship logic. »
I understand the point, but I don't really agree with that. As a player I have to know and master all the logic induced by the rules of a variant to overcome this variant. Some sudoku variant may seem puzzlish because of hard logical step that are far from the logic of classic sudoku. That's the point with variants, it can go far away from classic sudoku, and as an author I also like to go deep in the logic of each variant I create. In my opinion, criteria that make that it is considered as a sudoku variant or not is already contained in the rules.

Another answer was that there are lot of puzzle solving skill that you can use and transpose to solve some variants of sudoku. Practicing both puzzles and sudoku is an advantage in a lot of situation.
I strongly believe that it's true. For example playing kakuro will help you in solving killer sudoku, and conversely. I'm sure lot of examples can be found in that sense. My feeling is that it could help me to practice some puzzles in my sudoku playing, but I don't have to, it is not a PREREQUISITE. As an experienced sudoku variant player, when I meet a new variant, I know that I will be able to use my background to get fast into the new rules. Perhaps for some new variant, it will be a bit longer, or I'll be a bit slower than an experienced puzzle player, but I feel that the difference is not big.
As a sudoku player, I know that I'll be faced to puzzlish variations from time to time. For example in this tournament : the « even passage sudoku » is a variant that is a bit puzzlish for me, and I know I'll be slower than best players on such variants, but I really think I can't complain about that, I just have to practice on that logic.
With the battleship sudoku I strongly felt that not being experienced in solving battleship puzzles was a huge disadvantage for me, because it's not only about the similar logic, it's about the crude superposition of both puzzles, I have to solve 2 different puzzles together.

One other answer was that I should practice more overall puzzle techniques, instead of concentrating on classic sudoku's only.
Well, I'll skip the « play less classic sudoku » (it's my weakness and the GP is very classic oriented with 5-6 classic sudokus in each round...). For me, it is not a question of technique. I was able to solve the battleship sudoku logically (and I found the logic interesting, I want to be clear on that point). It's about the speed, I have no time to really practice puzzles in a competition way, so I know that in puzzle competitions I'll be around 4 times slower than the best players. Since sudoku and puzzle competitions are separated, I don't feel it's a big disadvantage for the sudoku competitions to be a « tourist » in puzzle competition, unless there are direct confrontations with non-sudoku puzzles in sudoku competition.

My experience in this tournament is that I solved all sudokus except the battleship in less than 50 minutes (~10 points/minute) and I spent 18 minutes to solve the battleship sudoku (the « not-only sudoku ») (=5 points/minute). I can't help to think that the presence of a non-sudoku puzzle in this tournament cost me 9 minutes.

Please don't take it as a harsh comment, it is not. I think we just don't share the same vision of what a sudoku tournament should be and we don't have the same limits. I'm quite sure I'm not the only player who think that puzzles like this « battleship sudoku » was not appropriate in this competition.
I found important to share my point of view in this forum.
I understand the philosophy of this round like that : lot's of easy classic sudokus for beginners, some common variants for intermediate players, and a hard one to make difference amongst top players. I also heard the argument that « all top sudoku players are at least decent in solving puzzles », meaning that it's ok to have an hybrid between puzzle and sudoku as high-pointer of a sudoku round. As everybody understand, I don't share this point of view.

My wish for the future : Please keep the sudoku competition free of non-sudoku puzzles.

Fred, A sudoku lover.

Dear, please explain more spesific about non consecutive sudoku as my english not really good to understand the additional information

Avina's picture

Thank you of all 'Authorities and Practitioners'.
In my idea, the points of product and classic Sudoku "6" are not possible. Because product Sudoku takes twice solving and so it takes more time than classic or even than the others. For me the end classic is easier of the other puzzles and it took least the time than the others. Battleship is very hard and I couldn't solve it. Please teach this puzzle very simply for all. I know many lovers of this kind puzzles want to learn it.
A review of previous competitions, I realized that this is harder than them. Why?! In this competitions all participating at different levels, but I think here is for the high level, and this is not important for me, because I am new, and I like but not possible for all of levels. Maybe, it was better this tournament held for 3 levels.
I hope to get more time to practice and speed.

I think I should reply to Fred as an author of this round. At first I want to thank him - for me it is very important to have some feedback. Even if this feedback is kind of negative.
I'm not a good sudoku solver (and not really like sudoku) so for me is very hard to estimate sudoku's difficulty comparing with other puzzles. Usually sudoku seems much more difficult to me. When I created these puzzles I tried to score them and battleship wasn't the toughest puzzle. On my first estimation it was in fifth or sixth place. So at beginning I thought this is an ordinary variation.
From the other hand I really want to extend sudoku competition and make it more "puzzlish". May be this point of view do not agree with many solvers, but I never participated in pure sudoku competition and will not participate in GP if it doesn't contain variations. So I tried to make round which is interesting for me. May be I'm not alone in this. But if the most players think I'm wrong then I probably shouldn't be sudoku author (and I will not at least for one year :) )
And finally I should say that one variation was excluded by Thomas from this round - I think it was more puzzlish and more difficult then battleships. So in some point I couldn't accomplish my mission completely :)

Hi all,

I must say I strongly disagree with Fred’s attitudes about vision of the sudoku competitions, not just GP tournament. I'm worried and a little distressed when I imagine this would be the generally accepted concept for the future sudoku championships.

I don’t want to restart an old discussion how sudoku competition should look like. Of course, making sure that everything is represented. Really, I don’t see why should we throw out sudokuro puzzle, for the example, from all future sudoku contests just because kakuro is originally the puzzle? The same question is for the skyscrapers sudoku, word (or number) search sudoku, mathrax sudoku, and many other familiar variations? Battleships sudoku, yes. By the way, great variation, Andrey!

Last week we had many very easy sudoku puzzles and only 2 or maybe 3 from medium category. This is not good. Every sudoku need to be a puzzle because if there is no thinking, then we compete in who is faster robot machine, or which pencil can withstand more. This is not interesting for me.

Frustrated because at the last WSC did not show a single sudoku novelty (although some of them appeared for the first time at the WSC’s), I decided to prepare a contest with innovative sudoku puzzles only (for the Sudoku Day on LMI) to prove that it is very possible to continue to develop and take care of sudoku. Thereby, I didn’t think whether I would mix some well-known puzzle and put it in Sudoku. I think I didn’t, but why not?

I’m afraid with such backward views on things we can hurt sudoku. I mean boring is a serious disease, and then it might be too late for therapy. Nothing personally Fred, this is my observation of this wonderful game. I know the puzzles are not in the main sphere of your interests. You already had similar comments earlier and I respect them. I just think sudoku does not deserve to be a slave of the uptight rules.


I'm glad to have both Fred and Nikola offering thoughtful opinions here. I do not expect every one of these contests to satisfy all solvers, in part because the different authors try to stretch our rules a bit in their own ways and each one has its own unique feel.

The Sudoku GP guidelines do ensure that ~80% of the test will be classics and familiar variations, but that still leaves some room for designers to make new variations or hybrids where traditional sudoku approaches are a strong element but not the only element needed for solving. There is no obvious line to draw for most of these new styles until after I've tested them, and I do cut a lot of sudoku that I say would fit better in a Puzzle GP or other setting. Other considerations that go into each contest include making sure the set of common/new variations don't focus too much in one area (for instance, I try to not allow too much "math" beyond 1 or 2 sudoku). I also try to ensure a balance of easy, medium, and hard puzzles, but not all rounds end up with as many "fair" but hard classic sudoku as others.

Thanks to Andrey, Nikola and Thomas for your reply.
The answers of Andrey and Thomas helped me to understand the choices that are made for this round, and more generally in the GP. I really appreciate to know your thoughts.

The point of view of Nikola made me think and ask myself why I have a conservative point of view on some points. I'll try to explain more precisely my point of view. I don't try to convince you, as I think that our visions are unfortunately irreconcilable.

First, I want to clarify that I only speak for myself, and even if I know that some players have the same opinion as me on some points, it doesn't mean that we are a majority of players and that it's the generally accepted vision.

Secondly, I think I agree with you concerning the difficulty of puzzles. I don't like competitions including only very easy puzzles, unless it's designed for beginners. I think the GP is designed to be fairly accessible for the beginners, but still satisfying for experienced players, but I agree that the last round was very “sprint-y”, with only 1-2 interesting puzzles for experienced players. I'll perhaps surprise you by writing that the battleship sudoku was the most interesting puzzle of this set. I really think it's true, and that is a real problem for me: the best puzzle of this set was not a sudoku, in my opinion.
Like you, I don't want to be bored at sudoku competitions, I want to solve interesting and sometimes challenging sudokus. What makes me distressed is when I have the feeling that authors or organizers of sudoku tournaments (I speak in general, not specifically this round of GP) think that “classic sudoku is boring, variants are acceptables, but only puzzlish variants are fun”, and finally it's the case: if someone is bored to create classic sudokus, players will be bored to solve them, because he'll only be able to create boring ones ! (I even had this impression at a WSC, that didn't contain a single classic sudoku). In my opinion, sudoku players will never be bored to solve good classic sudokus, but it's true that it's challenging to create good classic sudokus, and we still have lot of things to learn from Japanese authors.

Sudoku, including variants, is a subset of the huge set that I will call “puzzles” (even if puzzles is a larger set than what is generally accepted in puzzle competition).
Classic sudoku is a subset of the set of sudoku.
While it's easy to define clearly what is a classic sudoku, to establish rules to define what is a sudoku, including it's variants, is very hard and perhaps impossible. The consequences are that every player can have different feeling about which puzzles can be considered as sudoku, or how big should be this subset.
I imagine that if I had to write rules for this subset, it'll be impossible for me to include consecutive sudoku and exclude kropki sudoku, I also strongly think that skyscrapers sudoku and sudokuro (to use your examples) would be included in this subset. On the other side, my conception of sudoku stops when you have to find location of ships, when you have to draw loops, shade some cells, place mines or tents, or other puzzle things. I don't say these are not interesting puzzles, or we don't have to create them, I only think that these ones belong to the set of “puzzles” and not the subset of “sudoku”, so I think they should appear in puzzle competition and not sudoku competition.

Where my point of view really differs from yours is that I think it's not good to try to stretch this subset, by including gradually puzzlish variations or hybrids between sudoku and non-sudoku puzzles. Why? I imagine then that in 5 years, you've made this subset so big that the WSC players will spend 2 days finding the location of tents, ships, mines, shading some cells, drawing loops... and finding the digits will become a detail. Then WSC will be a clone of WPC. And then the only outcome will be to hybrid WSC and WPC, because having twice the same competition makes no sense, and this will be the dead of sudoku competitions... This is my fear.
In my vision, WSC and WPC should remain distinct competitions, WSC should keep its specificities, especially we have to accept that the field of sudoku is more narrow than the field of puzzles.

I think there is still lot of room for innovations without extending the subset of sudoku to puzzles. Nikola, you amazed me by creating lot of new types in some of the tournament you organized (Serbian championships, sudoku day, ...). There are also so many great sudokus we can create by hybridation of 2 existing variants (I think about the advent calendar by Dutch authors on LMD). I understand that for you it's important to see new variants in competition. For me, that's not the main point, as I think one can also be innovative in creating existing variant, but treating it in an unusual way, and I'm very happy when I'm amazed by a variant about which I though I knew everything.

I hope everybody understand my point of view.


Just a few words. I’m glad that we have almost uniformly thoughts in many cases. I wanted to point out that the nowadays problem is that most of the puzzle authors from the start excludes any sudoku variation (even classics) because there is a separately sudoku championship, or GP or other sudoku contests. The organizers also insist on strict splitting in the mood “why to put sudoku here when we have a special sudoku competition”. Reasonably or not? I think both answers are correct.

Each country has two national championships, sudoku and puzzle, and very rarely or never sudoku appear in puzzle competition. Also, several last WPC’s have had sudoku at the margins. Now, if sudoku authors are avoided to combine a sudoku and logic variations, there will be a huge hole between nothing in which sudoku can easily fall in.

Avina's picture

Nobody want answer me?! or tell me how to solve it?!

Hi Avina, sorry for skipping your message. I'll try to write the explanation, but it needs some time because it very difficult to do without pictures.

Avina's picture

Hi AndreyBogdanov, take your time.
Thank you./Avina

Avina's picture

Hi AndreyBogdanov,
I almost realized. By solving some examples, I will dominate.
Thank you very much

So I'm a little late to this discussion, but I nevertheless find it fascinating.

I have some sympathy with Fred's point of view. I suspect the fact the competition director has a published book (of albeit slightly different) battleships sudoku at least subconsciously affected just how novel this variant is.

Now, I won't claim to be an expert on battleships, but in more than a few of these puzzles I would suggest the sort of logic used is guided by intuitive speculation about placements, and that's a very different style of solving approach then most sudoku where placements made are generally deterministic. In theory, the combination of the logic is not necessarily harmonious.

At this stage I will pause to say that the competition puzzle itself was very nice, and the logic from each style complimented each other well in harmony, if not quite synergy.

I suppose the main criticism of this type from my point of view was simply how "different" it was from the rest of the test. In which case you could argue that it wasn't necessarily this puzzle that was at fault, but perhaps that some of the other puzzles were too easy. It's hard to say which is which, but as a designed of competitions in the past one of the things I look for in the competition is a certain homogeneity of the solving experience. That's not to say all the puzzles should be all the same, but they should roughly be of the same difficulty, and roughly have roughly the same levels of innovation. A contest after all is a way of comparing competitors, and for there to be any sort of meaningful result then you have to roughly be comparing apples to apples, rather than apples to oranges.

With regards to variation in Sudoku, I also feel I have to be brutally honest and say that I have seen many poor ideas and hybrid combinations of rules that simply do not work and produce puzzles that are boring, inelegant and not a pleasure to solve. This probably comes down in least in part to my own personal tastes, but innovation for the sake of innovation's sake does not guarantee a good puzzle - there should be thought and design behind the idea. Of course I accept the fact that many many different ideas have now been tried, and so any new good variations will probably come at the cost of dozens of terrible variations. That's just the nature of innovation: 99% perspiration and 1% inspiration. But that is not to say that there isn't a crucial editorial role to play.

I'd like to finish by addressing Nikola's frustration of the supposed lack of innovation at the WSC - something which I dispute in the strongest possible terms by the way. I will agree that most of the competition was familiar, but this was a compromise. Just indeed as there were complaints about not having enough harder classics! On the one hand it is my opinion that you cannot have too great a weighting to really innovative puzzles because then people will (rightly!) complain that the puzzles do not resemble sudoku as they know it and the argument about having a separate WSC becomes diminished - because you effectively have a sub-competition of a WPC.

Equally I took a decision to include only easier classics, because I didn't want too many classic puzzles within the competition! Having just easier puzzles seemed like the best choice to make, because 1) it allows everyone taking part the best possible choice of solving at least something, and 2) it makes the most efficient use of time at the end of what were necessarily shorter rounds. I decided the best way to have harder puzzles within the championship was to do so via "classic variants"- which is to say variants that have been tried and tested over time but which hopefully offered something slightly different to what people had seen before.

Now to wrap this all up, I think it's important to recognise that innovation comes not only from trying new and perhaps contrived sets of rules, it also comes in applying the classic rules in ways that you maybe haven't seen before. I must have solved thousands or even tens of thousands of classic sudoku by now - but as Fred says, it's still perfectly possible to come across a classic sudoku that surprises and pleases you and is like nothing else you have seen before.

My book does not use anything like Andrey's addition-heavy version of Battleship Sudoku. I was actually subconsciously driven to reject it, because it is not what I think of when I think of Battleship Sudoku.

This sudoku was more of a triple hybrid, including lots of summation constraints, and I felt that reasoning about numbers in the form of minimal and maximal sums was much more necessary (and new for most solvers) than any real Battleship experience. I judged this as a hard, novel sudoku variation much more than a puzzle.

As GP competition director, I don't really take on the role of editor and pick favorite puzzles. I review what is received, cut or ask for adjustments on sudoku that are too hard, but basically try to merge my goal of a fair competition with what the constructors' themselves have created. I really enjoyed aspects of Andrey's test and again will tell you to wait for future contests that will probably challenge you for entirely different reasons. There is no possible contest that will satisfy everyone. Which is why allowing some variability between contests is the approach I will continue to encourage.