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?	Just right
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
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 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

"Prime Domino" is actually based on the German(?) puzzle type "Domino Hunt" (e.g. http://www.janko.at/Raetsel/Dominos/index.htm, also it was round 1 at 2012 WPC), which is a much different sort of puzzle to the domino joining types.
It has its own rules that speed up solving. Quickly, the best ones are:
1): There are two ways that two dominoes can fill a 2x2 square. If both ways would use the same two dominoes (i.e. when a digit repeats on one of the two diagonals) then there would not be a unique solution, and so the 2x2 square must be covered in another way.
e.g. If the top-left corner is:
13
36
There are two ways to draw a 1-3 domino and a 3-6 domino in these 4 cells, therefore if we expect there is a unique solution then neither can be used - the 6-domino must go down or right. Draw in two edges between the 6 and the 3s.
2) If you locate a (e.g.) 4-5 domino, then you can draw lines between all other 4-5 pairs. For uniqueness reasons, this also works for blanks. If you find a 2-blank domino, then you can rule out all other 2-blanks (as long as they have the same grey/not-grey status), even before you know what the blank will contain.

If anyone is interested, I also wrote down some ideas and diagrams for how to solve three of the other types in the attached apreadsheet:

Regional Yajilin
Blackout Domino
Domino Loop

Edited by kiwijam 2013-09-10 4:12 AM




Attachments
----------------
DTGT - LMI Sep 2013 Theory.xls (42KB - 9 downloads)

I thought Capsules puzzles always had all pentomino regions. At least the ones I've seen and made always do. Maybe I've just seen many of Richard's.


Edited by Para 2013-09-10 5:47 AM

Password removed from puzzle booklet.

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
What was your opinion about the answer key extraction?	Mostly perfect answer keys
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 the point values reflect the difficulty?	Many puzzles were worth too much or too little
What was your opinion of the booklet formatting and printing?	Too few pages / too small grids

My first puzzle test for LMI is now over. It was an exciting weekend for me to watch players from all over the world being busy solving at lightning speed.
Congratulations to deu, MellowMelon and uvo for taking the podium positions!

Later I will write some notes about the test and puzzle selection.

Thank you all for participating and the (mostly) positive feed back so far!

Unfortunately I missed the test, but am looking forward to doing it in my own time. Thanks for the puzzles.

I discovered Capsules a long while back. Maybe around 2007, when I first started
solving WPF-style puzzles.
It was under the name "Capsules", so it must be from the same origin.
I can't remember exactly where but I'm suspecting the 24HPC.
I'm certain there were 1-cell, 2-cell regions and what not.

Perhaps someone can enlighten us.

This was a test with most of my weaknesses. I'm not too great with puzzles that use numbers without using the row/column restrictions, and so kept making silly errors in Capsules. I'm not that good with Loop/placement puzzles with empty grids and outside clues either. Dominos is one of those broad genres that I've always struggled at, and even my Japanese Sums base isn't great.

In short, the majority of this test contained types I usually skip during other tests, and some of them, 5 months or so ago, I wouldn't even have attempted after the test. But I pretty much had to "face my fears" for this one, and then I liked the puzzles so much that I solved them all afterwards. Thanks Richard, both for making me face my fears and for the beautiful puzzles I got to solve because of that

My example was WPC Belarus 2008. Puzzle 13 in round 2 were two Capsules puzzles of the type that was also in DTGT test; pento-shapes to fill with digits 1~5 and no touching.
Few years after that WPC the editor of Breinbrekers asked me to make a couple of these puzzles. At the same time I made one for the puzzle portal with minimal clues (10); it's linked in the forum as practice material.
These puzzles are not really hard to construct if you use a lot of P-pentominos. For the test I used the complete pentomino set once in the small grid and twice in the large grid (of course with exception of the U and V pentomino. )

The puzzle portal in Germany is a breeding ground for developing puzzle variants. Every author can publish any kind of puzzle that he likes. By creating and publishing a new invention, he/she is inspiring other authors as well. Those other authors may, in turn, inspire still other authors and so on. This makes the portal a vivid place.
A lot of puzzles that are included in DTGT are the product of such a process; they First appeared in the portal because I was inspired by other authors. (Not only portal authors by the way.)

Some funny notes about the test:
The grids for Capsules and Regional Yajilin are the same. Did anyone notice that?
In the IB, the grid of the Regional Yajilin is the same as the grid of the Low sudoku in Shapes and Sizes. I liked to link my First puzzle test to my First sudoku test.
Three puzzles have a 100% accuracy submitting range. Close to the end of the test it were still four puzzles.

Some notes about the individual puzzles then.

Killer Skyscrapers
A few years ago there was a whole series of skyscraper variants in the portal. I realised that a lot of variations that are common in sudoku can also be applied to other puzzles. Also on skyscrapers. Quite a few kropki-like skyscrapers appeared. I really liked the interplay between the skyscraper clue and the First digit in the grid when there was a kropki dot on the border of the grid. I played around a bit to make a killer variant to this, with cages going over the borders. The First few puzzles had repeating digits in the cages, so I didn’t link to those for extra practise material. Later I made a few without repeating digits.

Top grid: start with placing 1 in R5C1 and 1 in R1C5. Both cages (sum 6 and sum 8) must contain a 1 and that 1 can’t be outside the grid because of the non-repeating constraint.
Bottom grid: Start with both cages of sum 4. The cage on the right side has the 1 inside the grid and the 3 outside. In the cage of sum 21 bottom right the 1 is not outside the grid, and because of the two 1's in the cages of sum 4, the 1 in the large cage goes in R5C5. Then the 1 in the cage of 3 cells with sum 7 can also be placed in R3C4. Counting the sums in C3 + C4 leads to a 6 in R4C3.

Regional Yajilin
When I first noticed these puzzles in Prasanna's Borders & Beyond test, I was a great fan of this type immediately. I have made some normal Yajilins for Breinbrekers, but the area constraint really added something to the puzzles.
Since a good mixed puzzle test must include a few loop puzzles to gain speed, I included RY's too. But since I also wanted a puzzle where T&E is necessary, I made the second puzzle very hard. Most of the puzzle can be solved by logic, except for the bottom left part. Unthough a few cells in the 3x3-block can be eliminated for blackening, a few remain, and only TE brings the solution.

Capsules
In my opinion an interesting puzzle type, although most of them solve in the same way. Next to the T and Z pentomino are two spots that are forcing the letter placed in that cell to go in exactly one possible cell of the T and Z pento. The F, L, N and Y pento have one such spot. Those spots are powerful in the solving process.

Masyu Battleships
I don't know where, but I have seen a masyu of someone else that had a constraint on the unused cells. Either they had to be connected, or the unused cells must be isolated, I don't know anymore. But I combined this idea with normal battleships to make a nice crossover. Masyu clues bring the start to the puzzle, placement of the ships is necessary to make the loop unique. These puzzles are among the ones with the most wrong answer submissions; I think in the speed of finishing the puzzle lots of mistakes are made.

Easy as Chaos ABC and - No Touch
The editor of Breinbrekers is a great fan of Easy as ... puzzles. As I wrote earlier under the Killer Skyscrapers, also Easy as... puzzles can have additional sudoku constraints like extra areas as in chaos sudokus and the no touch rule. For me these are almost standard puzzles.

City Construction
I have made a whole pile of these puzzles for Breinbrekers during 2005 - 2008. Basically it's battleships, but with other shapes. The loop constraints make it more interesting, and adds to the solving process. Main rules are along the edges where there have to be at least two cells between two buildings. In this puzzle the two bottom rows provide the start. Column 4 with 7 segments is also important later on. Three cells remain empty, so there's on single cell empty and the other two are connected, since the loop passes once from left to right and once back.

Spiral City Construction
One of the most interesting puzzle types in Serkan's akil oyunlari test was the spiral battleships. The spiral can also be applied on other placement puzzles, like City Construction. The puzzles of this type in the portal are much harder, and mainly suitable for leisure solving. By giving all outside clues, I kept it relatively easy, although it is one of the harder ones in the set.

Filled Loop
In the beginning years of the puzzle portal one of the authors there introduced one variant to slitherlink after another. Most of these variants were really good. My contribution to the ever growing list of variants was the combination with pentos, in the way of filling the interior of the loop. This constraint provides additional info on the loop: take care that no pentos appear more than once and avoid smaller ends of the loop with less than 5 connected cell.

Pentomino in the box
This is my variant of the starbattle puzzles. When I did a theme month pentominos in january 2012, this type was the highest rated one, so I picked this type for DTGT. Small shapes in general provide good starting points. In this puzzle the middle area and top left area give a quick start.

Japanese sum pentominos
In one of the older USPC's there was a japanese pentomino puzzle. Outside the grid were the number of connected cells per pentominos. I liked it a lot. By filling the grid with digits a japanese sum variant was born. Try to find complete or almost complete filled rows/columns to get started, followed by ruling out cells that can't be part of any pentomino.

Sudokakuro
More than 50% of all my puzzles in the portal are sudoku variants. For that reason I wanted to include one or two sudoku variants in the set too. But since not all players are sudoku fans, I decided to take a variant that is 'far off' the normal variants. This sudokakuro uses a lot of 'Law of leftovers', one of my favorite techniques. The chaos sudoku part is more important than the kakuro part.

X-Sums sudoku
My favorite puzzle in the LMI Sudoku 8x8-contest of 2012. As 9x9 puzzle it can be quite tough, the one in DTGT was some kind of snack; a good puzzle to save for the last minutes.

Domino Loop
Although I am not a big fan of snake puzzles, combined with domino-counting I like it a lot. Seen it once at a very old Dutch Championship, and never after. The snake-part is handy to know what happens at the border; 3 to 5 cells at the border have to be connected. A 2 elsewhere as clue tells there are two narrow passings. The puzzle in the test can best be started with the grey cell bottom right.

Prime Domino
The dissection domino puzzles don't provide much possibilities to play around with variations. One that I found was this primes constraint. In the portal it was appreciated much better than the plain domino dissection puzzles. Knowing that there are only 11 possible prime domino tiles, you can conclude that in this case, every prime sum has to have a grey cell. That makes it easy to do a lot of eliminations.

Blackout Domino
In the Breinbrekers magazines there have been a lot of puzzles of this type. Only difference is that all black cells are already placed. By leaving the black cells out, the puzzle is getting more interesting. In the Original puzzle, the black cells could be orthogonal adjacent. In this Blackout variant, that would be too hard. So I made the extra rule that black cells may not touch and also don't touch the border orthogonally.

Cells sticking out the grid provide a good start. The two cells in the leftmost column are both part of a domino, and since all tiles appear only once, these two cells form one domino. The same goes for the two (x 2) cells in row 1 and the two cells in the bottom row. Those cells sticking out are either 'doubles' or one of the cells connected to them is black.


If there are any specific questions about individual puzzles: don't hesitate to ask! I'll be happy to answer.
I also like to read alternative starts in the puzzles. ;)

How balanced do you think the puzzle types of this test were?	A bit skewed
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
What was your opinion about the answer key extraction?	Perfect answer keys
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 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