API tools faq
paste
IhavenonameSDA

nurikabe turing test

IhavenonameSDA
Jun 18th, 2023 (edited)
457
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 4.22 KB | None | 0 0
raw download clone embed print report
  1. myamya (no curation)
  2. C: https://puzz.link/p?nurikabe/7/7/h1h2g3ubi1g2x2
  3. A: https://puzz.link/p?nurikabe/7/7/2o2o3n8j1k5h2k
  4.  
  5. puzzle-team (no curation)
  6. B: https://puzz.link/p?nurikabe/7/7/g7zk1nag3k1j
  7. G: https://puzz.link/p?nurikabe/7/7/k4k1r2t9m4g
  8.  
  9. puzzle-team (curated picks)
  10. E: https://puzz.link/p?nurikabe/7/7/i7m2o6k1o2m5i
  11. I: https://puzz.link/p?nurikabe/7/7/p4m4k1k3m3p
  12.  
  13. pzprrt (arbitrary placement until working puzzle)
  14. H: https://puzz.link/p?nurikabe/7/7/n4g6k5u1k1g1n
  15. J: https://puzz.link/p?nurikabe/7/7/zj7m7j4m4i
  16.  
  17. unassisted (solver only used to check)
  18. D: https://puzz.link/p?nurikabe/7/7/i3r4g7q2g4r5i
  19. F: https://puzz.link/p?nurikabe/7/7/k3k3r6r6k3k
  20.  
  21.  
  22. A: <https://puzz.link/p?nurikabe/7/7/2o2o3n8j1k5h2k>
  23. B: <https://puzz.link/p?nurikabe/7/7/g7zk1nag3k1j>
  24. C: <https://puzz.link/p?nurikabe/7/7/h1h2g3ubi1g2x2>
  25. D: <https://puzz.link/p?nurikabe/7/7/i3r4g7q2g4r5i>
  26. E: <https://puzz.link/p?nurikabe/7/7/i7m2o6k1o2m5i>
  27. F: <https://puzz.link/p?nurikabe/7/7/k3k3r6r6k3k>
  28. G: <https://puzz.link/p?nurikabe/7/7/k4k1r2t9m4g>
  29. H: <https://puzz.link/p?nurikabe/7/7/n4g6k5u1k1g1n>
  30. I: <https://puzz.link/p?nurikabe/7/7/p4m4k1k3m3p>
  31. J: <https://puzz.link/p?nurikabe/7/7/zj7m7j4m4i>
  32.  
  33.  
  34. [9:18 PM]Conflux: not planning to solve all but no way B is generated, A probably is
  35. [9:20 PM]Conflux: F also seems human
  36. [9:21 PM]Conflux: G also human
  37. [9:21 PM]Conflux: all of them have good aesthetics though!
  38. [9:21 PM]Conflux: also the ones I’m not commenting on I basically decided not to solve lol
  39.  
  40. [9:23 PM]dj💤man: lowkey suspecting you wrote one of the nonsymmetric ones just to screw with us
  41.  
  42. [9:29 PM]tluo5458: thoughts:
  43. A - generated
  44. B - curated generated
  45. C - generated
  46. D - generated
  47. E - human
  48. F - human
  49. G - human
  50. H - curated generated
  51. I - human human human
  52. J - generated
  53. [9:29 PM]tluo5458: i don't feel like i'm good at this game
  54. [9:41 PM]Rever: I think I agree with you.
  55. I'm stuck on A, B, C, D and J.
  56. B has some stuff goin on but I'm lost
  57. I feels extremely human lmao
  58.  
  59. [2:07 AM]apiad:
  60. generated: ABCJ
  61. curated: DH
  62. human: EFGI
  63.  
  64. ***************
  65. Design comments
  66. ***************
  67. no curation myamya - A & C: These are definitely the weakest puzzles here and show that naive generation (here is a working puzzle) really can stick out.
  68. puzzleteam - B & G / E & I: The only thing I wanted to make sure of was that I took a non-symmetric puzzle at some point. Symmetry definitely seems to elevate a puzzle, as both curated puzzles were symmetric. However, the biggest one was puzzle I: as shown above everyone thought this was human but no, it was made by a generator (and was, in fact, the 4th puzzle I got so it would have shown up in a larger no curation batch!) This group showed that a good generator, and especially with some filtering, can turn out some nice puzzles under specific circumstances.
  69. pzprrt - H: I wanted a symmetric layout and started with this V shape. Once I put in 4-5-6 as a starting point, it just worked with placeholder 1s so I had a puzzle. A bit lucky, and one that definitely seemed to have a human touch from the consecutives.
  70. pzprrt - J: I just wanted a puzzle with two pairs this took so long to stumble into. And it shows, as this style of construction was specifically scattershot to try to simulate a seeded generator and show that "bad" puzzles that are human-made with an idea can still be indistinguishable or seen as worse than generated puzzles. In particular the comparison with puzzle G is striking.
  71. handmade - D: I really wanted to put in some tricky steps here with the largest number having to cut off the side and thus influence connectivity later, and this kind of hard step presented in this context made the puzzle seem less human, which I found interesting.
  72. handmade - F: At the end of these comments, we have a puzzle that shows that the best human puzzles can still stand out. My point was already proven in that these middle categories are impossible to differentiate without being told creating "schrodingers puzzle", and yet the 3s and 6s with symmetry stood out as human from everyone who answered. This just shows that really good puzzles will always stand out, which was never something I tried to argue. All I wanted to do was prove a point that generated puzzles are puzzles, and I think I did that well.
Advertisement
Add Comment
Please, Sign In to add comment
Public Pastes
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!