Divisions A and B 2020-2021

Information and discussion on Puzzle Competitions
jupilogy
Posts: 33
Joined: Tue 07 Jul, 2020 2:22 pm
Contact:

Re: Divisions A and B 2020-2021

Post by jupilogy » Wed 10 Feb, 2021 11:05 pm

Puzzle_Maestro wrote:
Wed 10 Feb, 2021 9:40 pm
I meant the classic slitherlink - I heard from some people that the first took longer than either of the other two
Hmm, I did find the first slitherlink much easier and faster than the other two. The second two used some neat logic I hadn’t really used much before, though, so it took a while to figure out how to apply the logic. Maybe that’s why they had higher points?
pinkagape wrote:
Wed 10 Feb, 2021 8:38 pm
notable Jeffery DIck who finished in fifth in the UK
Teehee. Thank you :D I do really like loop puzzles, except yajilin...

I never considered needing an even number of cells as part of a loop until seeing some comments after that round.

pinkagape
Posts: 204
Joined: Mon 30 Apr, 2012 4:11 pm

Re: Divisions A and B 2020-2021

Post by pinkagape » Thu 11 Feb, 2021 6:16 am

Jeff, I'm with Freddie - I found the second slitherlink the easiest - what am I missing about the first? Is there a pattern with 1s and 3s that I'm missing with the third? I'm struggling to get started on it.

Regional Yajilin I found difficult too. I got the second one but still struggling with the others. I liked Freddie's hint about a path having an even number of cells, so I've got a start on the first one. I'm not sure what to do with the third Regional Yajilin too.

I can't find a start in the second Masyu All Cells.

They're the ones I'm still looking at every so often.

I really enjoyed the Country Road. That was my main positive to draw from the round.

Puzzle_Maestro
Posts: 110
Joined: Mon 11 Mar, 2019 5:23 pm

Re: Divisions A and B 2020-2021

Post by Puzzle_Maestro » Thu 11 Feb, 2021 7:48 am

For the first slitherlink, the 31 pattern gives two x’s and a segment. That causes some diagonal interactions.
For the Masyu (all cells), after a bit of standard Masyu, consider R6C6. You should be able to determine its orientation.
For the third slitherlink, the key deduction is loop avoidance; you can make some progress in three corners with this.

emma
Posts: 324
Joined: Wed 22 Jun, 2011 4:58 pm

Re: Divisions A and B 2020-2021

Post by emma » Thu 11 Feb, 2021 9:44 am

I did do the Puzzle Ramayan, but I'm in the "country unknown" list rather than UK. I had forgotten my password, their "forgotten password" link didn't work, and nor did the "if this doesn't work, email this address". I eventually re-registered with my work address, but failed to fill in my profile until after the contest.

pinkagape
Posts: 204
Joined: Mon 30 Apr, 2012 4:11 pm

Re: Divisions A and B 2020-2021

Post by pinkagape » Thu 11 Feb, 2021 12:34 pm

Thanks Emma. Here is an updated table.

Image

puzzlemad
Posts: 943
Joined: Mon 21 Jun, 2010 10:20 pm
Location: Oldham

Re: Divisions A and B 2020-2021

Post by puzzlemad » Thu 11 Feb, 2021 1:58 pm

Slitherlink - this guide by Bram de Laat is an excellent guide for some of the patterns within Slitherlink.

http://puzzleparasite.blogspot.com/2011 ... .html#more

puzzlemad
Posts: 943
Joined: Mon 21 Jun, 2010 10:20 pm
Location: Oldham

Re: Divisions A and B 2020-2021

Post by puzzlemad » Thu 11 Feb, 2021 2:10 pm

S&W Slitherlink - one of the things I find useful with those is to use some sort of notation on cells that must be inside/outside the loop - I lightly shade those inside the loop. I find that helps when looking at ends that can be joined up to make sure that they don't finish up with an inside and outside on the same side of the line.
I managed to do both of these in the comp along with the first two standard slitherlink - didn't get the 3rd one done at the time. I have solved all puzzles since.

jupilogy
Posts: 33
Joined: Tue 07 Jul, 2020 2:22 pm
Contact:

Re: Divisions A and B 2020-2021

Post by jupilogy » Thu 11 Feb, 2021 3:23 pm

pinkagape wrote:
Thu 11 Feb, 2021 6:16 am
Is there a pattern with 1s and 3s that I'm missing with the third? I'm struggling to get started on it.
What Freddie said is right, they lead into the 1-2-3-1 diagonals. If that is still difficult, I made a video showing how I did it, on my youtube

pinkagape
Posts: 204
Joined: Mon 30 Apr, 2012 4:11 pm

Re: Divisions A and B 2020-2021

Post by pinkagape » Fri 12 Feb, 2021 8:45 am

Puzzle_Maestro wrote:
Thu 11 Feb, 2021 7:48 am
For the Masyu (all cells), after a bit of standard Masyu, consider R6C6. You should be able to determine its orientation.
I'll leave the slitherlink out of it for now. What standard Masyu? The only things I managed to get on my grid were the four corners and that the black cell in R3C5 had to go up, and the one in R3C3 had to go down as a result, which gave me the white cell under that down line.

Puzzle_Maestro
Posts: 110
Joined: Mon 11 Mar, 2019 5:23 pm

Re: Divisions A and B 2020-2021

Post by Puzzle_Maestro » Fri 12 Feb, 2021 8:52 am

Yep, that’s all the standard Masyu you can do. But you know that R6C6 can’t go up and left or down and right (by normal Masyu), and going up and right will leave a cell empty, so it must go down and left.

Feadoor
Posts: 219
Joined: Thu 17 May, 2018 7:08 am

Re: Divisions A and B 2020-2021

Post by Feadoor » Fri 12 Feb, 2021 9:05 am

If it's helpful, here's how I saw the same deduction that Freddie describes, in picture form: sending R6C6 up and right causes problems for the white pearl marked with an X.
masyuallcells.PNG
masyuallcells.PNG (98.99 KiB) Viewed 2576 times

pinkagape
Posts: 204
Joined: Mon 30 Apr, 2012 4:11 pm

Re: Divisions A and B 2020-2021

Post by pinkagape » Fri 12 Feb, 2021 9:15 am

Thanks both!

pinkagape
Posts: 204
Joined: Mon 30 Apr, 2012 4:11 pm

Re: Divisions A and B 2020-2021

Post by pinkagape » Wed 10 Mar, 2021 8:35 pm

Hi

Round 3 of Divisions A and B takes place next week on www.puzzleduel.club

Each day between Monday 15th March and Saturday 20th March inclusive, the times of the Division A and B competitors will be ranked and added up to decide the positions of Round 3.

In order to solve the possible problem of a competitor failing to solve a puzzle (it's happened to me on Puzzle Duel plenty of times), I'm going to take a system where the second fastest solver in a division gets 100 points for the puzzle, and all other solvers get a number of points equal to the percentage of this time for this puzzle, up to a maximum of 500 points. Anybody failing to start or finish a puzzle will also get 500 points. The lowest scoring competitor over the week will be in first place.

Basically, all you need to know is that you need to solve the puzzles on www.puzzleduel.club for Monday 15th March and Saturday 20th March inclusive, as quickly as you can. If your username on Puzzle Duel is not obvious for me to guess, please let me know what it is. You may solve on computer with or without a mouse or with a mobile device, as per our conversation earlier in this thread. I'm happy that people will have different preferences and all work fine. The final scores will be taken on Tuesday 23rd March, since you can compete in past puzzles each day (e.g. complete Monday's puzzle on Tuesday rather than Monday), but please be aware that puzzles are only "live" for four days after initially being published, so you won't be able to compete in Monday's puzzle on Saturday.

Please ask for any clarification you would like. I intend to publish leaderboards during the week so that you can see how you're getting on.

Puzzle_Maestro
Posts: 110
Joined: Mon 11 Mar, 2019 5:23 pm

Re: Divisions A and B 2020-2021

Post by Puzzle_Maestro » Wed 10 Mar, 2021 8:58 pm

Are you going to penalise errors, and if so, by what amount? Due to the nature of the interface it is very possible to register an error just due to an edge/digit/other object having been missed and not an actual solving mistake; it might be possible to consider an ‘honesty error’ system here.

pinkagape
Posts: 204
Joined: Mon 30 Apr, 2012 4:11 pm

Re: Divisions A and B 2020-2021

Post by pinkagape » Thu 11 Mar, 2021 5:22 am

I hadn’t intended to penalise errors. I’m happy to try this year not to have any penalties for errors and we’ll see how it goes?

detuned
Posts: 2056
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: Divisions A and B 2020-2021

Post by detuned » Sat 13 Mar, 2021 1:51 pm

Maybe the very harsh way in which errors are penalised by puzzle duel will discourage more speculative solving - but theoretically making a mistake and having that pointed out is helpful to solving the puzzle. It also makes it fairer to those solvers who intentionally choose to spend a bit of time checking before submitting. E.g. today's passage puzzle is a newer and semi-intuitive variant. I spent 5-10 seconds (which as it turns out it was a significant % of my total time) checking everything before submitting, whereas I assume Sam didn't!

Perhaps a more balanced way forward would be to add 60 seconds per error, kind of similar to the way offline play-offs work?

Feadoor
Posts: 219
Joined: Thu 17 May, 2018 7:08 am

Re: Divisions A and B 2020-2021

Post by Feadoor » Sat 13 Mar, 2021 3:05 pm

detuned wrote:
Sat 13 Mar, 2021 1:51 pm
E.g. today's passage puzzle is a newer and semi-intuitive variant. I spent 5-10 seconds (which as it turns out it was a significant % of my total time) checking everything before submitting, whereas I assume Sam didn't!

Perhaps a more balanced way forward would be to add 60 seconds per error, kind of similar to the way offline play-offs work?
I quite obviously didn't check, but my error was just a cell that had become shaded through fat fingers rather than a genuine solving error. I'd be a bit miffed if I ended up with a 60 second penalty because of that.

In the end though, it doesn't particularly matter, as long as the rules are specified clearly in advance and everyone has the same advantages / disadvantages.

Feadoor
Posts: 219
Joined: Thu 17 May, 2018 7:08 am

Re: Divisions A and B 2020-2021

Post by Feadoor » Sat 13 Mar, 2021 3:09 pm

detuned wrote:
Sat 13 Mar, 2021 1:51 pm
Maybe the very harsh way in which errors are penalised by puzzle duel will discourage more speculative solving
In case it changes anyone's views on how our internal competition should be run, I should say that it won't discourage me at all because I only made an account on PuzzleDuel a few days ago and care exactly zero about my rating there.

detuned
Posts: 2056
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: Divisions A and B 2020-2021

Post by detuned » Sat 13 Mar, 2021 4:49 pm

Further thoughts: plenty of competitive online interfaces do not penalise errors at all (nikoli.com and fed-sudoku come to mind). Instead, they don't let you submit any kind of answer to be recorded until you happen to get the correct one. That definitely gives you a strong advantage if you happened to have bifurcated, got to the end of the puzzle and hadn't noticed a contradiction up to that point - rather than having to spend time at the end checking to see that you haven't violated any constraints, you can get instant feedback by submitting your answer. This is part of the reason that you have 60 second checking windows during play-offs. Whilst human checkers do need a bit of time to verify correct or incorrect, it very rarely takes the full 60 seconds (indeed when I do it I tend to watch people solve and when errors are made I have usually identified them even before they submit). It's also worth mentioning that just because an error has been announced, it isn't always straightforward to identify exactly where that error is. That's one difference between something like nikoli.com, which pointed out one example of a broken constraint, and fed-sudoku which simply says that you don't match the expected solution.

To take Freddie's examples, I would also add that when you are solving competitions on paper then you will end up penalised rather strongly if you happen to miss a small bit of loop or leave a cell empty ;) And to take Sam's example, equally, for something online with answer codes (like the GP) if you happen to do something fat-fingered whilst typing out the code, again you end up penalised rather strongly for it. In the latter case, one reason why organisers often give very little sympathy to answer code typos is because some people do choose to spend time checking and double checking before submitting.

I tend to take the view that solving accuracy strictly matters more than solving speed, and so some degree of penalty for inaccuracy is probably appropriate in any context... but yeah, Adam is the boss here!

Feadoor
Posts: 219
Joined: Thu 17 May, 2018 7:08 am

Re: Divisions A and B 2020-2021

Post by Feadoor » Sat 13 Mar, 2021 5:13 pm

All fair points, but, for PuzzleDuel in particular, in the very limited number of solves I've done, I've found the interface to be so clunky that I've barely ever got to the end of a puzzle without a typographical mistake - not something that is typically a concern with paper solving or typing short answer codes on a keyboard.

But maybe it's just that I'm not used to it yet :)

pinkagape
Posts: 204
Joined: Mon 30 Apr, 2012 4:11 pm

Re: Divisions A and B 2020-2021

Post by pinkagape » Tue 16 Mar, 2021 9:11 pm

Most of the way through Tuesday, here is the leaderboard after most people have finished Monday's puzzle (Tom and Neil, let me know if you have an unusual username and I can update the board tomorrow. Freddie is on fire, like he has been all year, but Division B is a lot closer at the moment.

Image

Nilz
Posts: 284
Joined: Fri 02 Sep, 2011 5:21 pm

Re: Divisions A and B 2020-2021

Post by Nilz » Tue 16 Mar, 2021 11:36 pm

I'm nilz on there. I should be appearing there now, I just solved the Mon and Tues puzzles. Not sure if there is/needs to be any penalty for solving a day late, although I guarantee I'll forget again before the end of the week.

pinkagape
Posts: 204
Joined: Mon 30 Apr, 2012 4:11 pm

Re: Divisions A and B 2020-2021

Post by pinkagape » Wed 17 Mar, 2021 5:16 am

No, I’m happy for people to solve a day or two late without penalty. Thanks Neil.

pinkagape
Posts: 204
Joined: Mon 30 Apr, 2012 4:11 pm

Re: Divisions A and B 2020-2021

Post by pinkagape » Wed 17 Mar, 2021 8:42 pm

Image

I thought Tuesday's one was going well but then I broke something and had to restart, but it looks like a few of us have been brought down with the odd puzzle. Some incredibly close times in both divisions so far.

pinkagape
Posts: 204
Joined: Mon 30 Apr, 2012 4:11 pm

Re: Divisions A and B 2020-2021

Post by pinkagape » Thu 18 Mar, 2021 8:02 pm

Image

Division A is reasonably close for positions 3 to 6, though the couple of high scores early in the week for David and Neil show how precarious this round can be.

Division B is a bit more spread out, but it can all change very quickly.

Post Reply