Hi Ivan!

Thank you for your feedback.

I don't think that the lack of instruction means a problem. But the fact that I didn't clarify that the PB wouldn't contain instruction was a real problem, indeed.
Sorry for that.

In R1,R2 the puzzles have a number. You can find it in the tables on the top of the page.
In R3 there is only 1 puzzle, and in R4 it would be hard to add a puzzle number.

I think in R2, R3 and R4 the letters on the arrows can replace the numbering of puzzles.


About the answer keys: R2, R3 and R4 have a similar layout, so we wanted to creat a same answer key. This was the main point. It was important to us to avoid the confusions in such a complicated test structure.
And this was the best one that we could find out. Please let me know if you know any better. Don't forget that the instant grading needs a fix length or a fix sum.

Well, the balance of rounds was less important for us in this case. We didn't want to make a well-balanced competition just provide some practise puzzles. And we had two choices, we just publish these puzzles on the WPC blog or we announce a competition. We chose the 2nd alternative, and I think it was a good decision instead of the mistakes of this test.

I can't say that we know the 4th round is much harder than the others. It can be true.

What was your most favorite round?	Round 4 - Black & White Matrix
How balanced do you think the puzzle types of this test were?	Fairly balanced
What did you think about the puzzle quality of the test?	Fairly Nice
What was your opinion about the answer key extraction?	Perfect answer keys
What was your opinion of the booklet formatting and printing?	Too few pages / too small grids

@chaotic: Indeed, good work. I had a lot of moments where I took a few minutes to solve a grid despite it being dirt simple, so I wasn't too surprised to get beaten on that one.

I'll also be surprised if any of my current 1s hold at the end of the week. I didn't completely blunder any of the rounds, but I didn't exactly hit any home runs either...

Hi

I have a query on Round 4. It states that two puzzles will be identical in the shared edge.

Is this implies that Puzzle 1 and 2 share an edge and Puzzle 1 and 5 share an edge. Can someone clarify this rule for me.

I am not sure what exactly your question is.
So puzzle 1 and puzzle 2 share an edge. Since you already have the pdf, I can say that Column 10 of puzzle 1 and column 1 of puzzle 2 should be identical. Like in the image below.

Does it help? Feel free to post if something is still not clear.








Attachments
----------------
Shared Shading.png (4KB - 0 downloads)

Thanks Admin. So similarly, Row 10 of Puzzle 1 and Row 1 of Puzzle 5 will be identical and so on.

Also I have a question in Paint it Black. I will send a PM to you as I dont feel I can post here.

Yes.
Replied your PM.

@rvarun: Yes.

edit: Ninja'd

Edited by Spheniscine 2013-09-25 12:14 AM

Enjoyed the puzzles this time around, but definitely felt that round 4 was out of step with the first three. Extremely cool idea and execution, but I ended up spending longer on round 4 than I did on the first three rounds combined. It was also a little difficult to figure out how to progress on some of the puzzles, since you had to apply adjoining logic to find a solution that met all the requirements, but I eventually got there. Used a lot more trial and error than I would like, but glad to have had the chance to tackle that beast.

What was your most favorite round?	Round 4 - Black & White Matrix
How balanced do you think the puzzle types of this test were?	Fairly balanced
What did you think about the puzzle quality of the test?	Very nice
What was your opinion about the answer key extraction?	Perfect answer keys
What was your opinion of the booklet formatting and printing?	Just right

I got owned by Pata pretty hard. I did not realize that the key squares are counted as white squares. The difficulty of the test seems to vary widely for me though. Round 1 and 3 took me ~1/2 hour, while Round 2 and 4 took me over an hour, and that's not including the botch I have with Pata.

The rank bonus seems to cause a big discrepancy in bonus values for rounds with fewer finishers (especially round 4 currently). Shouldn't the bonus just top off from 60 (1st 60, 2nd 59 etc.) unless there are more than 60 finishers?

The rounds were definitely of very different difficulty levels -- all were a lot of fun, I think most people will agree that #2 and especially #4 were harder than #1 and #3. This is par for the course in WPC and all the puzzles were really well designed (though I really played #4 poorly and then got close to not having enough time to finish it before a meeting scheduled for 75 minutes after my start time...). But I do tend to agree with Bram and others that this seems to be to make the bonuses for rounds 2 and 4 quite sensitive to small variations in solving time or even rank. Perhaps this will go away once more people complete all rounds. Or else, it might be something to watch for in future such mini-marathons -- allotting different amount of times to the different rounds might have been better. -- Denis

What was your most favorite round?	Round 3 - Great Wall
How balanced do you think the puzzle types of this test were?	Perfectly balanced
What did you think about the puzzle quality of the test?	Fairly Nice
What was your opinion about the answer key extraction?	Mostly perfect answer keys
What was your opinion of the booklet formatting and printing?	Just right

I always hurt my own performance by not sufficiently poring over the rules & examples before I begin.

In this case, I was ultimately able to make the correct interpretation of all the puzzle rules, but in a few instances there were some important subtleties that I didn't notice at first. For instance, in round four, it took me a minute to notice the given ship sizes on the far side of the sheet from puzzle 1. I also got very for down a wrong path because I presumed that pentomino types could not be repeated in puzzle 8.

The bonus system seems to be a good attempt to try and normalise the scores of otherwise incomparable scores to the rather arbitrarily chosen value of 60 minutes. It doesn't really make any account for going over 60 minutes so I suppose the puzzles should be so that the majority of solvers can actually finish them within the arbitrarily chosen time limit.

I would suggest a fairer thing to do would be to let this arbitrary time value vary for each puzzle (I guess something in the ratio of roughly 1:2:1:3 for this set), except this seems unnecessarily complicated, and you'd still have to pick an arbitrary time value. I suppose having a set of target times to try and beat is quite fun, but fairly working out what these target times should be a priori would appear to be a challenge to get right (although I'd hope that testing for this set would at least have suggested that 60 minutes for round 4 was going to cause problems...)

And if you can't have the target time a priori, I'm not sure why you would choose to normalise scores based upon an arbitrary time when you could more simply normalise the rank-time curves. You could still give people an (arbitrary and effectively nominal) points score for completing the puzzles and then base bonuses on the normalised rank-time curve.

As a separate issue, I would also argue that any sort of normalisation that uses #1 as a pivot point introduces unnecessary volatility and inaccuracy - see http://wpc-2013.blogspot.co.uk/2013/08/around-world-in-80-puzzles-s... for an analysis.

Finally, the penalty multiplier for going over 60 minutes seems completely unnecessary, and worse quite demoralising, given that you are already rewarding people for going under 60 minutes. There is already a penalty, relatively speaking! I'd highly recommend getting rid of this multiplier.

The puzzle set was quite entertaining!


Edited by detuned 2013-09-26 3:14 AM

also, round 4 would have been better served as 2 lots of 3x2 arrays. You already had linked puzzles over multiple pages for round 2!

I see now that pentominoes can not be repeated, so perhaps my confusion was in regard to not having to use all the types. Anyway, I was definitely confused, and now I'm meta-confused. I also just went back and tackled the combination puzzle for Round 2... it was very difficult for me but I'm happy to have finished the whole test, however slowly.

Actually on part 4 the penalty multiplier for submitting over 60 minutes works reasonably well, and arguably it could have been stricter. Given that almost no one completes this round in under 60 minutes, if there were no penalty multiplier everyone who's persistent enough would get the same score (120 points) (except for the happy few who finished under 60 minutes and got a bonus). Finishing the whole set in 61 minutes or in 5 hours would get you the same score. Instead, right now someone who finishes 8 of the 12 puzzles under 60 minutes does get ever slightly more points than someone who only finishes 4 of the 12 under 60 minutes, and in turn that person gets more than someone who didn't finish a single puzzle in 60 minutes.

Denis

Where I could claim for points?
In the Great Wall puzzle answer key says "enter the numbers from left to right". So did I for row "I" (I started from this corner) and got "wrong mark". Then I changed answer according to arrow direction and it was accepted.
Now the penalty mark is shown in the table. I don't know how it really affects on points, but think there is some lack of them.

In any case I like the construction of this puzzle, but it seems to be too easy.

Will check what is missing, and remove penalty if something is wrong at our end.
You can "Claim No penalty" from Submissions tab.

ETA : In your first answer, "I" is as "Left to right", but "J" is "bottom to top, in the direction of arrow". This is not listed as a valid answer.
In your second try, you entered both I&J as per the direction of the arrow. We allow that.

If you enter both I&J, as per the IB you get points. If you enter both as per direction of arrow, you still get points. If you swap I and J, you also get points.
Other combinations are not allowed.

Edited by debmohanty 2013-09-26 11:22 AM

Thanks a lot for clarification. I've got this.

What was your most favorite round?	Round 3 - Great Wall
What did you think about the puzzle quality of the test?	Very nice
What was your opinion about the answer key extraction?	Mostly perfect answer keys
What was your opinion of the booklet formatting and printing?	Just right

For the moment, I greatly prefer the TVC scoring system to the one here :) Failing to mark off one clue left a much much harder puzzle than I wanted to finish.

Thanks for the puzzles.

Edited by motris 2013-09-27 3:59 AM

At the moment, Para is 780.8 and I am 780.6. Seems like we're going for a wild race now. *opens betting*

But seriously. Either not all the masters have participated yet, or I somehow shot much higher now, with me still fighting for Top 5.

@motris: I can't see why leaving a clue unmarked makes the puzzle difficult?

Edited by chaotic_iak 2013-09-27 7:55 AM

@chaotic_iak: I mean I didn't put its logic into the puzzle by not crossing out one segment (row 3 middle) and then never went back to seeing it as useful. Effectively I was blind to some info, like I was solving a puzzle with fewer clues but it was totally my fault.

Edited by motris 2013-09-27 8:10 AM

I like the format of the test most. Usually once the test time is over I used to forget about it. Here the extended time made me to solve more puzzles. I hope this kind of a format will make more people try to solve the puzzles. Any how a great experience.