Monday, 26th June 2017
Puzzles Solved Yesterday: 83
Home | Register | Login | Current Puzzle | Archives | Leaderboard | Forum | Tutorial | FAQ
Forum Index
 
How to solve the puzzle from Friday, 01.07.16?
MondSemmel
Kwon-Tom Obsessive
Puzzles: 3684
Best Total: 7m 47s
Posted - 2016.07.09 11:54:32
Everyone occasionally struggles with a puzzle. In my case, my solution to the puzzle from Friday, 01.07.16 took 14m 56s. But I still don't know how to solve the puzzle significantly more effectively than via 10+ trial & error explorations.

Since several people managed to solve the puzzle in less than, say, 6 minutes, I must have been doing something wrong, or overlooked something.

So I'd appreciate any tips on how to solve that puzzle effectively. Any suggestions?
Darklady
Kwon-Tom Obsessive
Puzzles: 3375
Best Total: 9m 37s
Posted - 2016.07.10 01:29:11
I can't help you there. I kept making mistakes and had to reset twice; took me close to an hour. I don't know why but I've never had that much trouble on a puzzle before.
Tilps
Kwon-Tom Obsessive
Puzzles: 4024
Best Total: 20m 22s
Posted - 2016.07.10 07:42:49
I can't look up the scoreboard anymore to know if I'm remembering the right puzzle, but I believe I just made some extremely lucky nested guesses after messing about with a bunch of trials - 3 levels deep I think.
MondSemmel
Kwon-Tom Obsessive
Puzzles: 3684
Best Total: 7m 47s
Posted - 2016.07.10 17:40:29
Quote:
Originally Posted by Tilps
I can't look up the scoreboard anymore to know if I'm remembering the right puzzle
FYI, you can see your solution time in the archives.

Quote:
Originally Posted by
I believe I just made some extremely lucky nested guesses after messing about with a bunch of trials - 3 levels deep I think.
Ah, that's possible. Whenever I try to do a nested guess, and the guess isn't correct, I have a hard time with recalling which changes to revert. So I try not to use those if at all possible.
Tilps
Kwon-Tom Obsessive
Puzzles: 4024
Best Total: 20m 22s
Posted - 2016.07.11 09:06:30
I never fix position for nested guess - if I fail, I just start again from before the first guess.

I try to avoid it usually, but if I have a first level trial which gives me a lot of progress but no contradiction, I just keep going in case I'm lucky...

Archives say I got 4:19 - so yeah I think that was the one I did the very lucky guesses.
Jankonyex
Kwon-Tom Obsessive
Puzzles: 3523
Best Total: 9m 35s
Posted - 2016.07.11 14:15:15
It's possible to solve it without nested guess:
solution

I made this solution but I didn't solve it in this way when I play. I solve it with about 30% of trial and error, analysis and maybe highlander, and 70% of luck I think
Jankonyex
Kwon-Tom Obsessive
Puzzles: 3523
Best Total: 9m 35s
Posted - 2016.07.11 14:25:11
Quote:
Originally Posted by Jankonyex
It's possible to solve it without nested guess:
solution

I made this solution but I didn't solve it in this way when I play. I solve it with about 30% of trial and error, analysis and maybe highlander, and 70% of luck I think
JohnAtEcs
Kwon-Tom Obsessive
Puzzles: 972
Best Total: 14h 54m 58s
Posted - 2016.07.29 10:13:45
While not directly related to this puzzle, I did find the discussion here very interesting, specifically the use of the 'Fix current position' feature. I refuse to use this feature as the initial help guide suggests that this should only really be used when starting out as the real challenge of solving it is to look ahead and deduce the correct path.

Given the record times that some of these puzzles now solved in and the skill already exhibited by the top performers in visualising the way ahead, I was wondering whether it would be more interesting to define that a timed score could only be entered onto the leaderboard if the 'Fix current position' feature was not used or would some of the Friday solutions become too hard to visualise in the mind without providing some sort of backup.
arbor8
Kwon-Tom Obsessive
Puzzles: 2978
Best Total: 14m 2s
Posted - 2016.07.29 16:18:36
The hard Friday puzzles wouldn't become insolvable, but the times would be much worse. In the longer excursions it would take much time to erase your path and it would cause errors in trying to find the correct original point.
It's a very useful tool. I use it automatically, if it gets complicated. It's faster to "think" with fix-button than to try visualize the paths.
Zyntax
Kwon-Tom Obsessive
Puzzles: 3897
Best Total: 13m 6s
Posted - 2016.07.30 20:29:28
I don't think it's a handicap in any way. All it does is save you time from having to erase every segment by hand when you have to backpedal.

But this made me think of an interesting idea - maybe a separate daily puzzle area/tournament where that feature is removed. It would certainly make the non-trivial "Big" puzzles a lot more difficult, not to mention Friday puzzles.
Last edited by Zyntax - 2016.07.30 20:29:57
pqg
Kwon-Tom Obsessive
Puzzles: 3687
Best Total: 15m 37s
Posted - 2016.07.31 12:29:13
Personally I'd far rather see FP introduced for the Wrongs than removed for the regular puzzles!
JohnAtEcs
Kwon-Tom Obsessive
Puzzles: 972
Best Total: 14h 54m 58s
Posted - 2016.08.01 17:08:15
> All it does is save you time from having to erase every segment by hand when you have to backpedal.

But that is my point. Before I commit to an 'x' or |, I first have to visualise in my mind that going that route will fail and so eventually leave me with the one and only solution that completes the loop.

This is what makes these puzzles so absorbing [and sometimes time consuming]. I also know that once I have solved the puzzle, I have also proved that this is a unique solution [even though we are told that there is only one solution, I never use this fact when solving the puzzle].

If the cost of making a mistake is that you have to 'start again', which is what I have to do if I get a wrong deduction, then it focuses the mind so that you think twice before you make that commitment.

Sometimes, I have found that I have had to keep half the puzzles route in my memory trying to look for a conflict. If I go too far, then I normally assume that this is looking like the correct solution and so abandon it and try a different approach to find conflict, bearing in mind that I now want to avoid this route in other parts of the puzzle as likely to be right. It is at this point that I am normally tempted to hit the 'Fix' button but then that feels like I have 'got lucky' rather than proven the uniqueness of the puzzle.

My times have normally got quicker as I play more as like the rest of us, can start to feel where the key sections that need to be solved are. I assume that the advanced players can target these decision points very efficiently.

My original post was more out of curiosity in trying to visualise how some of the more devious ones could be solved so quickly unless the 'trial and error' approach was used to safeguard the time and focus on the most probable route to the solution.
JohnAtEcs
Kwon-Tom Obsessive
Puzzles: 972
Best Total: 14h 54m 58s
Posted - 2016.08.01 17:13:44
Also, to quote from the FAQ [which I have just noticed as I was looking for the help guide to the new Javescript version]

What does the 'Fix Position' button do?

"This is a bit of a cheat to help people whose brains can't cope with working out what the consequences of putting a line or cross might be."

So I am not sure that 'Foilman' intended the use of this feature once you became experienced in solving the puzzles.
Zyntax
Kwon-Tom Obsessive
Puzzles: 3897
Best Total: 13m 6s
Posted - 2016.08.02 03:48:18
Considering everyone gets the option to "Fix Position", I don't consider it a cheat.
foilman
Kwon-Tom Admin
Puzzles: 1705
Best Total: 24m 8s
Posted - 2016.08.02 08:09:06
I guess back when the site first started (and I wrote the FAQ which has remained mostly unchanged!) the puzzles were substantially easier, and perhaps I did consider it 'cheating'. These days, however, the puzzles are harder and I definitely couldn't solve a lot of them without using it - so I've made a slight revision to the FAQ wording...
JohnAtEcs
Kwon-Tom Obsessive
Puzzles: 972
Best Total: 14h 54m 58s
Posted - 2016.08.02 14:28:25
Regarding my previous suggestion, I wasn't advocating removing the feature, simply that in order to register on the leaderboard you had to avoid its use which would still allow everyone to enjoy the puzzles.

Perhaps Foilman could add an additional column to the leaderboard that recorded the number of times 'Fix position' was used so that those that were interested could focus on keeping that zero as well as recording a fast time?

By the way, I have no personal gain here, I have never managed to complete any puzzle in under 3-4 minutes - I can't draw quick enough and still need to fill in most of the 'x's to visualise the path but I would find it interesting to compare my times [not using the feature] to others that have the same philosophy.
arbor8
Kwon-Tom Obsessive
Puzzles: 2978
Best Total: 14m 2s
Posted - 2016.08.02 15:44:08
Recording the number of uses of "fix" would be a nice thing. It would quite accurately rate the level of difficulty of a puzzle (supposing no guesses were made )
   If it doesn't require too much work for Foilman, I support it
JohnAtEcs
Kwon-Tom Obsessive
Puzzles: 972
Best Total: 14h 54m 58s
Posted - 2016.08.11 10:49:34
A good example was the puzzle for Thursday 11 Aug 2016. This took me about an hour to solve [with a couple of gaps as elapsed time was 2.5 hours] as I got to a point where the look ahead took a while to visualise in my mind to eventually reveal that one of the two routes would lead to a trapped end that got blocked. I then took a few iterations for me to believe that this was true and not made a mistake before committing the 'x' but very satisfying once completed.

Quite often the look ahead is quite local [especially if proving that the loop will close prematurely] and I find these easier to visualise but in the case above it was quite long and I find spotting an orphaned / trapped end harder to prove.

However, if I had remembered the position I could quickly have seen which of the two routes would have failed but that, for me, removes the satisfaction of solving the puzzle.

However, I haven't done many Friday ones yet and so I expect that, at some point, I will find a Friday one where I almost have to visualise the whole puzzle before committing and I am not sure that I can do that without making a mistake in the logic [not to mention the time it will take!].

Forum Index