either that, or the explanation of what the mastermind clues mean (like in puzzles 17-18) is missing

this contest looks exciting !!!

Is the explanation of the Minesweeper answer key wrong ? Looking at the example, I feel it is simply "the number of mines in each row".

The name of the puzzle will be reworded and will not include "mastermind"

The answer key will be to enter contents of two rows/columns. This will be corrected in the next version of IB.

The contest does include some classic/evergreen puzzles like minesweeper, fillomino, skyscraper and tapa.

The contest name will be corrected before the start of the test.

For the linked minesweeper/nanro, are numbered cells always blank, or can there be a mine on those spaces? The instructions were not clear on that front.

For the minesweeper, numbered cells cannot contain mines. The instructions mention this indirectly "Place mines in some empty cells"

This is corrected now.

The test has started now.

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?	Linked Maxi Loop
What was your opinion about the answer key extraction?	Perfect answer keys
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Most puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

Great concept, and pretty good test, thanks Prasanna!

I tripped up a bit on the first Twopa, thinking I couldn't get out under the 13s in both puzzles. I feel like they were quite tricky anyway, an easier starting type would have felt more friendly.

The Fillomino relay types are really fun ideas. Some of the types were a bit low on interaction, with the connection rule mostly coming in towards the end of the solve. But keeping them easy seems appropriate for the PR.

* The Twopa disambiguation via shaded cells late in the solve were a bit of a pain. If you want to do it like this, at least ask for the same rows/columns in both grids, so it's easy to read off if you've solved the other way around? Or ideally just accept both combinations.
* The code order and direction on the right grids in the 4-by-4 types is confusing. Intuitively, I'd want to fill the "top" codes before the "right" codes, and read the "right" codes right-to-left. Improvement (but not perfect): point the arrows on the right and below the grid right and down, respectively.

Valid point. If a solver ended up solving the other way around, answer keys for 1 go into 2, and answer keys for 2 go into 1. Of course, you would have to select the correct rows/columns. But that should not be a big constraint (we thought so). We had a discussion on this exact point before the contest and then decided to go ahead with the current approach. If it was an offline contest, there would be no issue but this adjustment was only because the contest is "online."


If someone enters top code before right code (that is intuitive), it will still be marked correct (although it is not mentioned anywhere). We have emphasized left-to-right and top-to-bottom wherever possible. And the IB examples also illustrate that. We just wanted to stick to the "rows before columns" mentioned under the generic section ("If one row and one column is marked, enter the row first and then the column"). Your idea about the arrows' direction is also a good one, perhaps better than the current depiction. We did not want the arrow to point outside the grid, as that may have also introduced some confusion.

Thanks for the useful suggestions!

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
Which set of puzzles did you like the most?	Fillomino-Norinori Relay
What was your opinion about the answer key extraction?	Perfect answer keys
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Most puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

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?	Linked Maxi Loop
What was your opinion about the answer key extraction?	Mostly perfect answer keys
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Most puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	I have a different complaint

Regarding answer keys I think backward arrows are confusing.

I would also prefer if Twopa did not have shaded squares to force each solution to go in a specific grid, since that gives a potentially undesired break-in, and it makes it hard to plan answer drawing. I don't know if the scoring setup allows for multiple possible answers though.

I also don't like that the answer keys were separate for each grid, since if you guess to finish one piece of the puzzle I don't think you should be able to confirm that it's right until you finish the other grids.


Otherwise, I loved this contest though! I think connected puzzles are always really fun, and it's cool to see a full contest of them.


EDIT: Looks like rob already said a lot of the same things.

Edited by ghirsch 2019-01-25 11:14 PM

Your point may be valid. But we were very clear that we wanted to encourage partial points, especially for more beginner kind of solvers.

Very! In some of the puzzles the answer key instructions made it clear that the reading order was still left-to-right and top-to-bottom but not in all, and I happened to start with Norinori which did not include those instructions so I lost points because of this. I hope admins will consider retroactively awarding points for backwards answers on those - I'm sure I'm not the only one!

EDIT: other than that this was great :D

Edited by boing 2019-01-26 12:33 AM

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?	Linked Minesweeper-Nanro
What was your opinion about the answer key extraction?	Answer keys could have been better
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Most puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

Answer key arrrows poiinting left were slightly confusing.

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?	Linked Skyscrapers-Easy as 123
What was your opinion about the answer key extraction?	Perfect answer keys
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Most puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

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?	Linked Skyscrapers-Easy as 123
What was your opinion about the answer key extraction?	Answer keys could have been better
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Most puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

Mostly the same comments as ghirsch. I got tripped up by the left arrow in the norinori (like boing) and found the twopas confusing because of the one shaded cell.

In the twopa rules: "This means, in a multi-digit clue, some of the digits can have the same behavior, but not all," I would have liked the example puzzle to clarify whether a single-digit clue could have identical behaviour between two grids (I guessed not but couldn't solve the puzzle).

With partial marks I noticed one of my mistakes by submitting an incorrect subpart which you wouldn't usually be able to notice, so I felt like having separate answer keys was bad - perhaps you could still submit separate answer keys but only mark them individually at the end of the contest / when all other subgrids have also been submitted.

A very nice test though - the logic was nicer than I was expecting in the puzzles I did solve (especially the linked skyscrapers-easy as 123).

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
Which set of puzzles did you like the most?	Twopa
What was your opinion about the answer key extraction?	Answer keys could have been better
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Most puzzles were worth the right amount
What was your opinion of the booklet formatting and printing?	Just right

I also got tripped up by some backwards arrows, and I couldn't finish the harder Linked Maxi Loop! Still, nice puzzles overall.

Personally, I think the Fillomino/Star Battle relay was worth (a lot) more than 2+2 points.

Edited by mstang 2019-01-26 10:48 AM

@2 above:

Yes, the rules for Twopa as I would state them are: "for any pair of corresponding clues in the two grids, the arrangements of shaded/unshaded cells in the (usually 8) neighboring squares cannot be exactly the same." So that applies to single-digit clues equally well.

Edited by mstang 2019-01-26 10:47 AM

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?	Below Average
Which set of puzzles did you like the most?	Twopa
What was your opinion about the answer key extraction?	Mostly perfect answer keys
Of the puzzles you solved/attempted, how well did the point values reflect the difficulty?	Most puzzles were worth too much or too little
What was your opinion of the booklet formatting and printing?	Just right