How balanced do you think the puzzle types of this test were?	Fairly balanced
What was your opinion of the distribution of easy/hard puzzles?	A bit hard
What did you think about the puzzle quality of the test?	Very nice
How did you feel about the length / time limit for this test?	A bit too many puzzles
Of the puzzles you solved/attempted, how well did their point values reflect the difficulty/time spent?	Most puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

I loved the Little Killer, as well as the Movable Digits, upon re-solving. Within the test, I really liked the Killer. The Tennis theme was put in wonderfully through the test, and it was nice that there were so many places where it kept coming up. Overall, I enjoyed the test quite a lot.

Coming to the general discussion, I'd just like to say, I think the points distribution was fair enough by my experience. By now, I have a fair idea of my own strengths and weaknesses and I believe thats necessary for every solver to learn. Based on my prior knowledge it pretty much went to expectations except for the 2 broken puzzles. And there was a mix of easy-medium-hard which qualifies as balanced for me because it gives enough to all categories of Solvers to have a go at.

However, I think the discussion of whether or not there should be such hard puzzles in a test is a bit more debatable. I think that, while its ok to have hard puzzles on a test, the 2 here in particular had a higher risk involved than the ideal case and thats the cause of most of the players' frustration. I believe its because, there are some variants where the rules themselves make the puzzle difficult to solve, like a Little Killer, Diagonally Consecutive (imagine ones with the converse rule coming into play a lot too), Movable Digits, Frameless, etc. I think these are generally difficult, because they depend heavily on uncommon logic, like diagonal neighbours/lines, transferring to grids, determination of cells for sum and so on, and its easy to make a mistake in these even on easier examples. So, I think, in the general sense, if one of the more normal variants like say the Renban Groups or Extra Regions is made to be difficult, it won't be noticed as much as say a Little Killer or a Frameless being difficult. Its what I'd keep as a base plan anyway. But obviously that is dependent on each author, and thats what makes puzzling fun, its the constructor challenging the solver after giving a rough estimate of the difficulty he will face :)

How balanced do you think the puzzle types of this test were?	Fairly balanced
What was your opinion of the distribution of easy/hard puzzles?	Just right
What did you think about the puzzle quality of the test?	Fairly Nice
How did you feel about the length / time limit for this test?	Just right
Of the puzzles you solved/attempted, how well did their point values reflect the difficulty/time spent?	Most puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

How balanced do you think the puzzle types of this test were?	Fairly balanced
What was your opinion of the distribution of easy/hard puzzles?	Just right
What did you think about the puzzle quality of the test?	Fairly Nice
How did you feel about the length / time limit for this test?	Just right
Of the puzzles you solved/attempted, how well did their point values reflect the difficulty/time spent?	All puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

Assuming the results hold, I'm just thrilled to have finally managed to get more than the participation point from one of the Grand Prix events. That's pretty much been my goal, so mission accomplished! And it's pretty much all thanks to Movable Digits. That was such an interesting variant that I couldn't pass it up, even with the huge points/difficulty, and for once the gamble paid off. I'll admit that I was completely oblivious to the theme while solving, but it is absolutely brilliant in hindsight. Enjoyed all of it I've done so far, and hope to at least approach the remainder later.

Sudoku GP - Serbian round is over. The results will be public in a while (after Branko has a final review).

Here is the link to solution booklet

Loved the theme - though as usual I never fully catch on til after the fact (Canada v. Serbia this fall - we're just happy to be there). With respect to the points variance - my single voice/vote is I'm not a fan but can live with it when, as was the case here all the details are well known before the fact. For those who struggle with errors on killers (hand raised) it made this an easy test to decide what to skip.

How balanced do you think the puzzle types of this test were?	Fairly balanced
What was your opinion of the distribution of easy/hard puzzles?	A bit hard
What did you think about the puzzle quality of the test?	Very nice
How did you feel about the length / time limit for this test?	Just right
Of the puzzles you solved/attempted, how well did their point values reflect the difficulty/time spent?	All puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

Oh, I understood the rule all right. I just would have thought your own posts would be exempt.

EDIT: All right, I get the puzzle now (my comment was regarding the hidden posts rule, and I only see the reply today). I was under the impression that each clue was utilized once and only once for a different digit each. Therefore I was reading that, in the example of the 4, that column 1 was not solvable based on my perception of the rule, since no 4 was in the second column that could be moved. I just saw Para's post explaining it further and now understand that not every cell (on the right) has a clue, but it must be determined which clues (on the left) are shared by a cell (on the right). Not sure I can solve that one, but I'll give it a try now that I understand that.



Attitude? It was my first post in the thread, and what I thought was an error. There was no attitude whatsoever. Please accept my apologies if I came across that way at all, as I simply came to the forum to report what I thought was an error.

And now I get what Deb meant by "rule" above (a puzzle rule). Clearly I didn't get that rule, then, at the time. I've given it another read since; see my edit above.

Cyclone


Edited by Cyclone 2013-04-23 10:10 AM

How balanced do you think the puzzle types of this test were?	Perfectly balanced
What was your opinion of the distribution of easy/hard puzzles?	A bit hard
What did you think about the puzzle quality of the test?	Very nice
How did you feel about the length / time limit for this test?	A bit too many puzzles
Of the puzzles you solved/attempted, how well did their point values reflect the difficulty/time spent?	Most puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

How balanced do you think the puzzle types of this test were?	Fairly balanced
What was your opinion of the distribution of easy/hard puzzles?	A bit hard
What did you think about the puzzle quality of the test?	Very nice
How did you feel about the length / time limit for this test?	A bit too many puzzles
Of the puzzles you solved/attempted, how well did their point values reflect the difficulty/time spent?	Many puzzles were worth too much or too little
What was your opinion of the booklet formatting and printing?	Just right

Perhaps participating highly unprepared and tired after a long day working wasn't the best idea, but after I missed the German round (because of being an author to that test) already, I decided to give it a go. Unfortunately I messed up the diagonal consecutive and it took veeeeeery looooooong to finish the irregular and the frameless. But the puzzles as well as the theme were really nice, so I had a good evening in the end.

Thanks for a good quality set of puzzles with good, well worked out, ideas.

I think the main reason that your attitude was questioned is that you basically pointed out something thats been asked and clarified on this forum thread itself already. When there's multiple posts already in the topic its obvious you should read through in case your query's already been asked. Also, you didn't exactly ask or even address the possibility that you might be understanding the rules wrong, instead choosing to think that the authors and everyone else viewing the Instruction Booklet for the past few days missed all the easy-to-see contradictions you pointed out from your interpretation.

Link corrected.

Few months a go, Zoran gave a theme for our round. Tennis!
Nikola didn’t have problem with this theme and created 8 sudokus in a short time. Branko created three more in next weeks.
We had 11 sudokus with some similar patterns, but all looked well.
After some time, Cedomir created 2 sudokus, Zoran one, Nikola added two more.
And Movable Digits from Cedomir came in last days of competition booklet preparation, right on time.
We made a selection of 16 sudokus, tested each other sudokus, and Thomas was so kind to test all in ones.
We wanted to have enough easy sudokus, enough well-known sudokus and some challenging for best solvers. We hope that we achieved most of that.

Congratulations to Jan Novotny for clear victory and to Bastien, Tiit, Kota and Michael for solving 15 sudokus.

Once again, we want to say that we have excellent help from Thomas in testing and reviewing booklets and Deb in preparing contest.

We also want to say that we made manual corrections when we have good reason to do that and we tried to be fair to all players. We are also really sad that we couldn't help to Jan Mrozowski.

Score page: http://logicmastersindia.com/2013/04S/score.asp

Nikola, Zoran, Cedomir and Branko

Edited by skywalker 2013-04-23 1:01 PM

Congratulations to Rishi for making to top 10. I hope this is the worst result for Prasanna and Rakesh.

And I am still waiting for Rohan to participate in this. I think his clock has stopped working.

How balanced do you think the puzzle types of this test were?	Fairly balanced
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
How did you feel about the length / time limit for this test?	Just right
Of the puzzles you solved/attempted, how well did their point values reflect the difficulty/time spent?	Most puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

Thanks to the puzzlemakers! It was fun to discover thematic secrets - after the contest, of course.
Thanks to all the participants! It is a big pleasure to browse through the result list again and again. I haven't seen my name on the top for quite a long time.
Looking forward to the next sudoku competitons and to the next tennis battles between Serbia and Czech Republic in Davis Cup and Fed Cup...
Krtek

How balanced do you think the puzzle types of this test were?	Perfectly balanced
What was your opinion of the distribution of easy/hard puzzles?	A bit hard
What did you think about the puzzle quality of the test?	Very nice
How did you feel about the length / time limit for this test?	Just right
Of the puzzles you solved/attempted, how well did their point values reflect the difficulty/time spent?	Many puzzles were worth too much or too little
What was your opinion of the booklet formatting and printing?	Just right

Thanks to all contestants for the feedback, the praises and criticism alike (it will be silly to expect that one can make all of the people happy all of the time :)).

Nevertheless, I would like to mention two things.
Firstly, there is no such thing as an ideal competition. Secondly, the instruction booklet did read: "Points typically indicate difficulty of the sudokus and time required to solve them. While the organizers have made best efforts to match them, your personal experience and preference may differ."

Whether "To make it worth less than 157 just so the scoring overall looks "balanced" would be to grossly misrepresent the challenge it presents and cause even more frustration from solvers"is the case remains to be seen.In fact i feel that the real frustration creeps when knowing that after attempting 8 or more puzzles with full focus and concentration you get to know that
you are beaten by a person who has completed just 2 puzzles which might be his forte and which might incidentally have carried high points in the test.
Also i dont think the issue of point distribution needs to confused with the spread of difficulty at all.I dont think i have complained anywhere about the difficulty of the puzzles as such so far.The spectrum of our discussion is whether a puzzle deserves so much points because it is that much more difficult.I certainly dont think that just because a puzzle
might be X times more difficult than another puzzle in the same test it should be getting X times more points than that puzzle.This idea
seems absurd to me especially because going that way there is no CEILING LIMIT to the points that a particular puzzle can take.

Let me give a situation where the idea that a puzzle should get points corresponding to the difficulty that it has or corresponding to the time that a test solver takes without a ceiling limit to its total value is true.
Take the case of Puzzle marathon.The above idea might just hold true over here.This is because in such a scenario as this,
i)The Time factor is taken our of question with no fixed time limit under which you are supposed to solve a puzzle
ii)Each puzzle does not have any 'relative' or 'cascading' impact on any other puzzle.
iii)Each puzzle can be started,solved independently on its own and has the right to carry whatever points be it 200 or even 500 ,
depending on the presumed difficulty.

Now does any of these cases arise in a timed test such as the one in discussion?The short answer is a NO.If you get tripped on a puzzle you
are doomed .The time that you lose can never be obtained back.
WHEN the tests have a fixed TIME LIMIT and a DEFINITE TOTAL NUMBER OF POINTS,it just does not seem right to me that the puzzle can carry points in direct proportion to the time that a test solver takes or based on how many times it is more difficult compared to another puzzle without a LIMIT to the total points that it can carry.
Maybe it would not be a bad idea to have a survey asking participants about what they felt regarding the distribution of points in this test,
whether they see it fit to not have any restriction on the total points that a single puzzle or a small group of puzzles might carry.............................

My error was that I had only reviewed the second page of posts. I did not review the first page. As I saw the puzzle addressed on page 2 and I still didn't understand the rule being conveyed (which I see now was on page 1), I took that as the information I used when posting. Clearly I was in error. I'll give both the sample and competition puzzles a shot today knowing the rules better now.

Cyclone


Edited by Cyclone 2013-04-23 8:25 PM

The points distribution was very near to reflect the real difficulty of these puzzles.

Puzzles with 20-30 points was solvable within 2-3 minutes, the harder puzzles needed a few more minutes. On the WSC there are also some harder puzzles and nobody is discussing it, because the goal is to sort out good/better/best solvers from the others.

How balanced do you think the puzzle types of this test were?	Fairly balanced
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
How did you feel about the length / time limit for this test?	Way too many puzzles (too little time)
Of the puzzles you solved/attempted, how well did their point values reflect the difficulty/time spent?	Many puzzles were worth too much or too little
What was your opinion of the booklet formatting and printing?	Just right

I don't think you're considering everyone here, you are only considering "I completed 8 puzzles with full focus and concentration", which is completely disregarding the population of solvers who will feel "I completed this Little Killer by taking 20-30 minutes on it, within which time I could've easily solved 6-7 puzzles" or even "I took 30-40 minutes and couldn't solve this Little Killer and couldn't get time to attempt others either". All 3 situations have merit, and there will always be people falling in numerous categories of things going wrong in a long test. The solution is not to make things suit one group or the other, but just make sure that practically, the points are distributed fairly. And the testers' timings method is the most practical way. So thats not going to change to suit one side of the argument. So again, I'd say the topic of whether such difficult puzzles should be in a competition format is debatable, but the way the points are distributed is as accurate as can be.

Also, from the last line I can derive that you might have the opinion that people strong at some variants can solve any instance of that variant in a short time. This is not the case. speaking from my perspective, I'd say Little Killer falls roughly in my strong variants. I still took 25 minutes on it before breaking it (I've now found that a typo mid solve cost me and I was solving the right way). I think I solved about 6-7 of the lower pointers in the same time span early on in the test, but this 25 minute expenditure not only cost me directly, I also felt the effects of it while going forward in the test. Now say, I'd solved it right, but the points total was just 120 or so, would I not feel even more frustrated? My point here is, please take a second and consider the enormous risk a person takes attempting these puzzles and the amount of focus and concentration needed to solve these.

So I don't think anyone addressed my post directly, although I think my thoughts are broadly represented by what motris said. The issue is not points allocation, which can only be done sensibly in the one way everyone has already outlined. The issue is the appropriate inclusion of puzzles. And I think it is an issue that hasn't really been discussed seriously before - and I believe this is certainly something that has been got wrong at previous WSC's.

The issue comes down to variation in puzzle difficulty. Shorter, easier puzzles will have much less variation in solving times than longer, more difficult puzzles. Basically this comes down to a couple of factors. First is the fact that you can go wrong in far more exotic and long-winded fashion on a harder puzzle than with an easier one, so often you have no idea that you've gone wrong if you've made a typo or whatever. Second is the fact that a harder puzzle might require one specific deduction that you have to get, and if you miss it for a while then you are going nowhere. I have lots and lots of data to show this is an observable phenomenon with harder puzzles.

Managing the risk of taking on a big pointer is a complex issue. For example, my last correct submission was 35 minutes before the end, at which point I decided to take on the 118 point moveable digits. I broke it after about 15 minutes and then couldn't quite get it finished before the end. In that same time there were 146 points spread across 3 puzzles I'd have got without any problem if I'd attempted them, with perhaps a long shot at the higher variance 108 point killer to boot. Put in context, that decision cost me certainly 33% of what I could have got, and about 30-40 places in the table.

The solution seems obvious right - stick to the safer low variance puzzles for your points right? The problem, especially for a one round format where there is a good chance to finish most if not all of the puzzles on a good day, is that ultimately you'll have to have a go at the harder puzzles, and given they are harder puzzles you need to make sure that you've left sufficient time to have a go at them otherwise again you are left with the problem of having say 10-15 minutes of solving time where you run out of time trying to finish a puzzle. This brings me to the point I was trying to get across, in terms of the difficulty of the test being massively skewed. When you have relatively few puzzles commanding a relatively large proportion of the points, then the lesser pointed puzzles start to lose relevance and the test is exposed more than usual to the elements of chance with the inherently higher associated variance.

Perhaps the proof is in the pudding. Looking at the results there are more solvers (I'm not going to name names) than I'd expect who are way below where they "usually" are. Applying the risk variance thesis gives an explanation: a test with more "risk" than normal is inevitably going to throw up more casualties to variance.

Of course that's not to say managing the harder puzzles is not a skill in itself. It is also no coincidence that there are certain solvers who seem to fare consistently well with more difficult puzzles...