Well, seeing that I am not as strong as the top competitors, my standard of judgement is rather different. It felt that there are more approachable puzzles on the other 2 tests compared to the IPC, though that might have to do with my particular puzzle experiences as well. Objectively though, in the percentage of points department, I actually did quite a bit worse on the USPC than the IPC, even though it didn't feel that way.

Edited by achan1058 2013-07-09 10:11 AM

Thanks to all authors for some super puzzles. A few notes:

A1 - As I am among the minority who solve puzzle contests using paint, I am aware that a maze can be easily solved using paint by using 'fill colour'. Hence, in my personal opinion, mazes should never be part of online contests. They can be part of offline contests definitely.

A2 - The multiple choices did not have the original clue cells filled in the solutions and this made it a little more difficult in sudoku as well as slitherlink. I would have preferred all cells to be filled in the sudoku solutions and also the boundaries, and the existing clues to be included in the slitherlink solutions. Or, maybe this was intentionally done to make players uncomfortable.

E2 - I solved the skyscrapers sudoku two times, as I made an error the first time. In the end, this solved fairly elegantly, with the 3s first, then all 6s, followed by all 5s.

G2 - Pentomino Kakuro started so well that I was wondering if it really deserved 65 points. In the end, I solved it three times to ultimately end up with zero points. If this puzzle had gone well, I might have finished at my regular (fourth) position, as I wasted a lot of time and lost a lot of points here.

G1, G2 - I liked the use of the boundary rows/columns in both puzzles.

H1 - This was the first surprise and was a little painful. Letters are anytime preferable to circles.

H2 - The LMI theme was prominently at play here, in the clues as well as the gray part.

M1 - I have not solved this, but this design is visually disturbing again, in my opinion. Maybe good for a leisurely solve but not as a competition puzzle.

I have still not tried F2 (B&W loop variant), J2 (Happy dots), M1 (slitherlink). But the other high pointers which I did not touch during the contest solved fairly smoothly later.

Top 3 enjoyed puzzles (so far): J1, K2, L2

As already voiced multiple times already, beginners hardly had anything (except A1, A2, E1) which they could solve realistically.

So, maybe, a section with mini versions (easier kakuro, loop, abc, sudoku, tents, tapa, word search). Plus some other types which come in newspaper like mastermind, hitori can be introduced. This can form a set of 8-10 starters which seasoners can solve quickly and beginners can solve over the test duration.

(Adding to Deb/Prasanna's points)

Test goal? - I think this should be "To select the best possible team who can represent the country at WPC." and not "Only 1 or 2 top Indian player should be able to complete."

For this, these could be some of the factors to be considered -

1. Coverage

Reach out to a larger community than current reach-out. There is always a chance that there are individuals who are talented but are not even aware of IPC/WPC.

2. Test duration/difficulty

I think this needs to align with the objective we set. So, if we set a target for a Top 25 finish in the WPC, the test duration/difficulty can be designed/finetuned based on a benchmark of the 25th WPC rank holder's performance of last year. It should not matter if 0 or 1 or 10 Indian players are able to finish.

3. Puzzle selection

Ideally an offline contest takes care. But since that is unlikely, prepare a test with
(a) 8-10 starters (newspaper types)
(b) 10-12 main course (WPC regular types)
(c) 4-6 challenging ones (WPC tough rounds/new unknown types)

In brief, I am probably suggesting an IPC similar in structure/flow to USPC/UKPC - to cover a larger spectrum of puzzles and have more accessible puzzles. (instead of having multiple similar puzzles, and very few accessible puzzles)

I was aware that players might try to solve the Maze using "Fill Colour". But the image was created such that Fill Colour won't work. I used 3 different software to test that (mspaint, Paint.Net, PicPick) and Fill Colour simply won't work with the maze image. Did you find it otherwise? If yes, using which software?

Well, it was done intentionally (even putting dots instead of digits). Otherwise, both the puzzles should have been only 5 points each.

This year's IPC seems to be very successful as it has selected 4 names who are arguably the Top four in the country. With the experience and talent in this team, we can expect extremely good results this year.

I obviously did not use this during the test
But I just created an image in ms paint, closed the entry points and filled with colour.

Bad testing on my part. Luckily, it was not a whole lot of points. Thank you for bringing it up.

Edited by debmohanty 2013-07-09 12:38 PM

I solved this maze during printing the PB. It was the first page, so I had it still on the screen and it was pretty easy to solve it just by eyes. However, the idea by Rakesh is very smart although it is slower than honest solving.

Edited by greenhorn 2013-07-09 1:35 PM

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 of the booklet formatting and printing?	Just right

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?	Fairly Nice
What was your opinion of the booklet formatting and printing?	Just right

As others have said, too few easy and standard puzzles. Online National Championships and WPC qualifications are the competitions that are most likely to attract new/nonregular solvers, and they should not be frustrated.

Also I did not prepare, and so I stumbeled across some unusually wordings in the puzzle booklet during solving.

For example in Easy as ABC, it is not said, that each letter has to be used in each row/column. Actually, the wording is different from the instruction booklet, which is a no-go for me.

And in Kakuro it is only stated, that digits can not repeat in words with a clue. Experienced solvers might see such wording as a "hint" for a twist in the puzzle, which could cost much time.

Edited by Realshaggy 2013-07-09 2:34 PM

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?	Fairly Nice
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
What was your opinion of the booklet formatting and printing?	Just right

I agree ideally Hitori and Mastermind makes up more in the newspaper quotient, but I think accessibility beyond newspapers can be defined with two points -
1. Easy to understand rules.
2. Types with many examples available if one is willing to search.
Even assuming there were no dedicated practice sets, Nurikabe, Tapa, Tents, Battleships, Star Battle, Easy as ABC, Hidden Words fall under point 2. Corridors, MCQs, fall under 1. Sudoku, Kakuro, Slitherlink are newspaper puzzles. I'm only leaving Black and White Loop away from category 1, because there's a few tricks implied by the rules that aren't obvious.

So accessibility-wise, the "classic" half of IPC 2013 was very much accessible, at least in my opinion. I also don't think we should stray away from the Classic + Variant formula, because that is in fact suited for beginners, as they mainly need to look at 13 sets of rule-sets with a few tweaks (overwhelming in itself, when the newspaper puzzles can only cover up to 5 distinct types at most), instead of 26 different rule-sets. It is definitely more accessible than the USPC/UKPC I think, at least for an Indian. Puzzle backgrounds may differ and the USPC/UKPC may be more accessible to participants of their respective countries.

I'll say again that, more than accessibility the main problem is the fact that the classic puzzles themselves had twists and were a step above what newcomers were prepared for. The Sudoku was a notch higher than newspaper difficulty, the Kakuro had a few clueless cells, the Hidden words didn't have words, the Easy As ABC was huge, the Slitherlink was a complete twist. These being the first puzzles that any Indian newcomer would attempt (this much I'm pretty certain of), and having a bit more than usual would be the main turn-off.

The problem with Atril (the Evince fork for MATE) persists with the IB as well. However, the problem seems to be localised to specific pages.
IB: 6, 8, 11, 12, 13
PB: 6, 10, 11, 13
I thought I had located the culprit in the form of the labeled row- and column- pointing arrows used for keys. Alas, after looking at the IB, this seems to hold very roughly, so something else must be going on.

How did you come to know about Indian Puzzle Championship?	I am a regular visitor at LMI forum.
Did the Instructions Booklet (IB) provide enough information about participating in IPC?	Yes, mostly.
Did you understand the puzzle rules, examples and answer keys from the IB?	I understood all of them.
Did you find enough easy puzzles to solve during the championship?	Yes, some of them were easy. I would have liked more easy puzzles.
After your experience in IPC, are you going to participate in future championshiops or contests at LMI?	May be

Sudoku : The difficulty of the sudoku was higher than newspapers, but should not players be tested little more than what they solve everyday? And this was a sudoku, that still didn't need anything other than hiddens/singles. Although it had fewer clues compared to newspaper ones, it still had a wide solving path.
Kakuro : I didn't anticipate the problem that beginners might face. Otherwise I might have asked Serkan to add all the clues.
Hidden Words : Not sure if I agree here. A search is a search, whether letters or symbols. Agreed that, with letters it is easy, but the solving methods are pretty much same.
Slitherlink : Didn't we already have a classic Slitherlink in page 2?
Easy As ABC : It was huge, but it was easy. But yes, someone looking at this first time, they may not want to attempt it.

For a newcomer, the Sudoku having fewer givens is in itself intimidating. So is a Hidden Words having symbols. Note that a newcomer is someone who has probably never solved a word search with symbols, but has maybe solved a few word searches with letters. Slitherlink, I hope you're joking with that :P Easy As ABC as well, like Sudoku, intimidates with appearance itself.

The main point I'm trying to make here is that, the IB "features" (whether intentional or not) these puzzles as the ones that are familiar or easy to think about. A newcomer will go in thinking what they know from the newspapers is good enough to get some points, and then get intimidated at the starting steps itself. Thats never good. Make the common types really easy (easy by the definition of a newcomer and not an experienced puzzler), so that they can actually translate what they already know into some amount of points, instead of having to advance their thinking from the start itself.

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?	Average
What was your opinion of the booklet formatting and printing?	Just right

With no more players participating at this time, IPC is over.

Official Results : http://logicmastersindia.com/2013/IPC/score.asp
Complete Results : http://logicmastersindia.com/2013/IPC/score.asp?mode=complete

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 of the booklet formatting and printing?	Just right

My opinion applies to the Turkish GP as well.
If we all cater newcomers then there's no room for "surprise" innovations.
But we must also consider what is fair game and what is too much of a stretch.

The Turkish GP's only flaw I find is with the rotation of the photo in Spy Sudoku.
A player shouldn't be asking "hmm.. maybe this flips" during the competition.
I don't see how changing fonts on the first 3 puzzles is an enough excuse for one to complain about the entire test.

With this year's IPC, the puzzles are fantastic. The symbol word search, unclued Kakuro, few-clues sudoku shouldn't be a problem as the
rules aren't that much deviated. Heck, you might as well criticize the 16-submarine battleships too then.
But when the rules do bend by far: for example having Strange as ABC, the instructions of Easy as ABC - it didn't happen, but it
illustrates what isn't fair.

I thought it was good to think on your feet, since it is required at the WPC.
Remind yourself that this test's goal is to select the best Indian team members, not making newcomers too comfortable.

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?	Fairly Nice
What was your opinion of the booklet formatting and printing?	Just right

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 of the booklet formatting and printing?	Just right

The sudoku I think is actually ok, and the kakuro wasn't too bad neither. The Nurikabe on the other hand, especially for someone who is used to the Nikoli style, was awfully hard for 20 points. The slitherlink I think is ok in a twisted sort of way, since the point value and the geometry made it clear enough that I should skip. (and I would imagine it to do the same for new solvers as well) Coming from someone of my skill level at least, there's nothing more frustrating than diving into a problem that you think you can solve due to type/point value, but it turns out that you can't, or that it takes a large amount of time.

Edited by achan1058 2013-07-09 10:16 PM