What was your opinion of the distribution of easy/hard puzzles?	Just right
What did you think about the puzzle quality of the test?	Very nice
Which set of puzzles did you like the most?	Quadro Sudoku
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Most puzzles were worth the right amount

great set of sudokus. thankfuly i had no mistakes in this round. in my opinion i found true or lie and coded sudoku easy, finshed 9x9 for both of them in 10 mins combined. had already decided to go for the hghest pointer 2 odd 2 even first and thankfully it worked for me. i felt that quadro should've been given more points compared to true or lie and coded sudoku.

overall i am glad that i finished the whole set within time

Nice test but rather difficult. Even the Classics were hard.
I basically give up an hour fumbling hopelessly so it was a surprise to see other people found it just as hard.
I thought Coded 2 was too guess-y and the 2O2E was beautifully made.

What was your opinion of the distribution of easy/hard puzzles?	Too many medium difficult puzzles
What did you think about the puzzle quality of the test?	Fairly Nice
Which set of puzzles did you like the most?	Quadro Sudoku
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Most puzzles were worth the right amount

Thanks for this amazing set. I really enjoyed quadro and coded during the contest, and 2 odd 2 even offline.

Edited by harmeet 2019-02-10 5:20 PM

Does this mean the only differences are to do with the presentation, rather than how you might solve the puzzle? That's certainly how it seems to me, so I'd be grateful if you could confirm this or else point out differences I might be missing.

My interest in general is that sometimes Sudoku contests do themselves no favours with being overly complex and not friendly to beginners. So I see an old and relatively well-understood variation masquerading under new rules and new presentation, and I wonder if that is really being helpful. My personal opinion is that a beginner would look at a 12 in a pencilmark grid and be able to grasp what is going on far better than a 1 in a true or lie grid. If that's really the case, I ask myself what, if any, benefit this new presentation is bringing to the table.

The whole "Liar" theme is also something that probably needs some further point - at some point in the WPC and WSC community someone decided that "Liar" should mean differing by exactly one and the precedent has stuck for whatever reason. However, to an outsider this might seem quite arbitrary, and whilst this might be lying in a technical sense, this +/- 1 concept is actually far more restrictive. That, to me, is something more like "fuzzy" clues.

The final point I'd like to make is if this is really just a pencilmark sudoku, it kind of stretches the theme of "substitution" a little far. Granted, I personally don't care that much for themes as I know it's perfectly possible to group together variations that superficially look similar but fundamentally solve using different heuristics, however there may be solvers who end up confused by the theming. I tend to think it should be the puzzles that make the theme, rather than the theme that makes the puzzles.

True or Lie sudoku is not a new idea. I think it was first created by Nikola in sept. 2014 for the tournament sudoku day: http://logicmastersindia.com/lmitests/?test=M201409S, and appeared in a few occasions, like indian sudoku championship 2015.
I've created some similar variants, liar sudoku: http://sudokuvariante.blogspot.com/search/label/Menteur, with only the "liar" idea.

Concerning the points you raised:

Presentation: In my opinion it's not less well presented than pencil marks. Here we have only one clue in each cell, which is in my opinion more readable than pencil marks sudoku. Of course it could have been presented as a pencil marks sudoku, it's only a small twist. I think this variants belongs to a category we could call "coded information" variants, exactly as roman numeral sudoku. Roman numeral sudoku could be presented as pencil marks sudoku, too: "V" means pencil marks 45678, "III" means pencil marks 38, etc... I don't see a problem in having a small twist from pencil marks variant which leads to true or lie sudoku.

Beginner-friendly: I don't think the rule and presentation of true or lie sudoku is too complex for a beginner. I think the idea is quite simple and the presentation very clear. If a player prefer pencil marks, it takes few seconds to write them in all clued cells, I don't think it's a problem for players.

"Liar", perhaps liar is not the right word to use, but who really cares about name of variants?

The only complaint I would make is that playing online was not very easy with this formatting, as mstang noticed.

Fred

In a True or Lie sudoku, each given clue restricts marking to ONLY 3 possibilities which is certainly not the case in a pencilmark sudoku (where you may have 2,3,4 or even more ! given digits/possibilities) in a cell.Yes it is similar in solving style to a pencil marks sudoku but that doesn't mean that it should be taken off from contests.



I think you are reading too much into an issue which I think is very trivial.The opinion that you put forth regarding a beginner grasping idea of a pencilmark grid more than a TOL grid is very dubious and I don't quite agree with that.On the contrary, any person grasping anything boils down to whether that person in question understands the rules clearly(which I think can definitely be taken for granted for TOL rules) and once that is done , I think TOL is a far better and beginner friendly version than a pencil mark sudoku.



As I said before, the issue is too trivial .We didn't even think of it being compared to a pencil mark sudoku.Perhaps you might have had a bad experience solving a TOL in the past which bothers you when you see it and your opinon on it is largely framed based on you being very subjective ?!



If a beginner or any person for that matter had a question regarding rules of TOL, that would have been raised in the forum and besides TOL doesn't find its debut here .It has occured previously in other websites' contests too (argio logic for one which also presents pencil marks sudokus separately !) and even in last year's SM round (where you didn't seem to have an issue) and where no one had posted any problems regarding comprehension of rules.

I was quite disheartened after my performance today and felt I was not fit for competitive sudokus. Then I decided to try and solve the puzzles offline. I had got stuck on classics 6-7 points and wasted lots of time in the contest. Offline, I finished the 10 points true-lies quite fast and also the odd-even count. I think the 10 pointers were easier than the classics. I take back my comments that it was extremely tough for a beginner. The variants were actually very interesting. I just chose the wrong puzzles. And forgot to upload the result of the first classic thinking i will do it later . Anyways I continue to learn from my mistakes. Well done, Kumars.

Do you have any suggestions for improving the formatting for this variant?

You may have noticed that points for classics were higher as compared to most SM tests. So classics were expected to be harder.

There is nothing like a wrong puzzle but individual preferences can differ. And with the clock running, even easy grids can look formidable sometimes.

What was your opinion of the distribution of easy/hard puzzles?	Too many hard puzzles
What did you think about the puzzle quality of the test?	Very nice
Which set of puzzles did you like the most?	2 Odd 2 Even Sudoku
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Most puzzles were worth too much or too little

Extremely hard test,as expected.
Just like last time,literally every sudoku was difficult,which is now certainly a theme of Ashish's creations without doubt.
I felt a few sudokus as a result were grossly undervalued,examples being coded and odd even count,both of which were worth more points.
Classics should have been given lesser points to compensate.
Didnt enjoy the test much,but thanks to Ashish and Kishore for the sudokus.

Slightly off-topic, but regarding +-1, my suggestion would be to go with "knapp daneben".

In terms of presentation I think it's fine to not present the True or Lie as pencil marks. Besides the mentioned Roman Numerals, there's also e.g. last GP's Braille Sudoku or just Odd/Even Sudoku. Certainly they can all be expressed as pencilmarks, but I feel that the presentation does direct the logic in a way. The kinds of arguments that are possible are somewhat restricted, and there are different patterns to be recognized. I did end up writing out candidates in the hard competition puzzle after missing some logic, but in the end I think the messier grid rather solved me down extra (on top of the work of writing out candidates).

The puzzles were great, and quite difficult. Thanks to the authors and organizers!

What was your opinion of the distribution of easy/hard puzzles?	Too many hard puzzles
What did you think about the puzzle quality of the test?	Fairly Nice
Which set of puzzles did you like the most?	2 Odd 2 Even Sudoku

What did you think about the puzzle quality of the test?	Very nice
Which set of puzzles did you like the most?	Classic Sudoku
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Many puzzles were worth too much or too little

@kishy72: Ok, I feel there has been a lot of playing the man not the ball here. Comments like:

"Perhaps you might have had a bad experience solving a TOL in the past which bothers you when you see it and your opinon on it is largely framed based on you being very subjective ?!"

are not helpful. I have had no such bad experience, and indeed I've already outlined where I am coming from with my thinking. I don't know why you are trying to falsely attribute further words and motives to me, but I sincerely hope this is unfortunate behaviour you might hold back from in any further discussion. You might think what I've raised is trivial, but that's also just your opinion, and it's an opinion with which I disagree. I also doubt I'm the only one who disagrees.

You are correct in saying that not all pencilmark sudoku can be represented as true or lie, but it is very much the case that all true or lie sudoku can be represented as pencilmark sudoku. So we certainly have some kind of equivalence going on here, and so I think the question as to which presentation is more helpful is a legitimate question to raise.

Re beginner friendly - I've framed things in this way because presentational issues are more likely to affect less experienced solvers than more experienced solvers - although perhaps this was not the best way to make my point. I take the point that as stated true or lie isn't unclear in the statement of its rules, but ultimately my point is that doesn't mean that things couldn't be improved. In general I think contests should aim to present puzzles as clearly as possible.

@Fred76, @rob: Solving this competition puzzle reminded me of solving a Roman Numerals/S for Sudoku/Morse etc. In general I'm not the biggest fan of such puzzles either - exactly the same arguments can be raised when it comes to the required bookkeeping required to solve these puzzles, and whether they add to or hinder the solving experience. Certainly if you are transposing the given information back into candidates there's a risk you might make transpose something in error and not realise until much later, which I can imagine being very frustrating. On the other hand I accept that not everyone will be solving the puzzle quite like that. Maybe there's something to be said when the information is encoded in completely different sets of symbols, as opposed to numbers 1-9 abbreviating the candidates.

Anyhow all that aside, I thought the puzzles in this round (although quite hard) were of very good quality. In particular it's nice to see some very well crafted classic puzzles, which often get left as an afterthought in many online contests.

What was your opinion of the distribution of easy/hard puzzles?	Too many hard puzzles
What did you think about the puzzle quality of the test?	Very nice
Which set of puzzles did you like the most?	2 Odd 2 Even Sudoku

Puzzle quality was excellent, even though some puzzles seemed quite hard to me, but maybe thats because I did the test late in the evening. Quadro 2 and Classic 4 seemed undervalued to me. Thanks for the nice set of puzzles!

Right.You started this even before you took the test by stating that TOL is a less well presented version of a pencil mark sudoku when it clearly felt to us that you were taking a jibe and belittling the efforts that go behind in putting forth each episode.You probed further asking me what the differences were when you knew yourself what differentiates the two.When the matter is insignificant for a discussion, it's natural to think that your past experience might have had a say and moreover I didn't say that you had one.I said "perhaps you might have ..." .Both have a difference.If I still stick to my point and say now that you are letting your judgement clouded by your past encounter with TOL, then that's totally different.If you think that "unfortunate behaviour" is to withheld then you should be more polite when you addressed the issue.

I am 100% certain that if it weren't for you, it wouldn't even have occured to someone to equate the 2 and then sacrifice one for the other just supposedly because of an alternative version with better presentation based on some preconceived notion.

There might be numerous cases where one variant is very similar in solving style/presentation to another and can be sacrificed for the other but it would just be an unfruitful thing and a waste of time and efforts to search for such.They just add variety.


You seem to be contradicting when you agree that TOL isn't unclear in the statement of its rules and yet you put forth that contests should aim to be as clear as possible.If that's the case here, then why this pointless discussion ?!



Thank you and indeed the puzzles were on the harder side and I will add a little to this after the contest.We discussed this right after sinchai gave the test and it became very clear to us when kwak took 77m to finish the set which is much more than what he usually takes for an episode.

Is it possible to extend the test for one more day i.e. till Wednesday midnight IST.

An idea I had was to put the numbers (faintly, but larger font) over the whole square so the pencil marks would be more visible on their own. Not sure, though.

Thank you. This is noted for future contests.

@kishy72: I’m done here, this isn’t a school playground, and I won’t waste my time further with childish taunts of “you started it”. You’ve twice dissected my posts line by line and remararkably you have failed to come close to addressing my point. I don’t think there’s anything more I can say to such a person.

That said I’m sorry if you thought I was initially making a jibe or belittling the contest. Perhaps you are not aware, but I have been organising online contests for nearly 10 years and I have an interest in all aspects of running them. I certainly know that running a contest is no easy thing to do, that it’s certainly possible to make mistakes, and I am grateful to all those who give up their time to do so, you included.

If you feel you have run the perfect contest, then maybe you have the right to be so touchy about any kind of criticism. If not, then I urge you to react with a little more maturity in the future, and take the positives rather than inventing some straw man negatives to get so upset and offended over.

I may as well try and put all this bad blood behind me and take something useful out of the other “discussion” - given the similarities to other symbolic type variants then some of the same principles apply. I thought larger, centred and greyed out clues worked nicely for Morse on the recent GP round, and perhaps something similar might work for this. Another idea is to apply Pinocchio style outlined clues rather than solid clues.

Thanks for the suggestions. We will definitely consider this approach when we come across such a scenario in future.

Please, no digits that occupy the whole cell and are not part of the solution, even if they are grey or outlined. This is my (personal) advice for formatting. See here for a past discussion on the subject:
http://gp.worldpuzzle.org/content/sudoku-round-6-liar-sudoku

Fred