Kingofthehill476

SM 13% IcePb writeup

Dec 10th, 2025 (edited)
181
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 11.14 KB | None | 0 0
  1. From Kingofthehill476:
  2. 13% Ice-Pb is made possible by underflowing ammo at draygon with only 10 pbs. However, this requires laying a pb frame perfectly and then collecting a pb drop from goop (a 2-frame window). Thus, it takes both luck and skill to complete a run.
  3.  
  4. I would definitely say that I got lucky when I landed the 10pb underflow because things fell into place. However, my solution aims to mitigate the RNG working against you and increase the skill component. Within the small RNG loop, I didn't grind for a 2 in 87 chance for a pb drop. I grinded for a 4 in 43 chance of a goop pattern that I know will result in a pb drop at the right time. This essentially normalizes goop shot timing and pb placement timing while increasing the odds of encountering the desired RNG (goop patterns).
  5.  
  6. Here is the solution for how I did this:
  7.  
  8. _________________________________
  9. Short answer:
  10. 1) Enter a tiny RNG loop with 87 RNG values compared to the normal 2200+ values. Here's how to do it:
  11. - Save your game outside of Norfair (Not quite anywhere outside of Norfair, but saving in tube, aqueduct or pre-dray will definitely work)
  12. - Get down to 69 health
  13. - Walk into pre-cathedral from the left with suits off
  14. - Don’t move in the room. Sit there and die
  15. - Load your save. As long as you don’t get to the title screen (resetting), then you’ll still be in the loop. This means that you need to die and select continue after any failed attempts
  16. *Subunit health must be at 0x00, so pause abusing and refilling health from reserves will set it to this.
  17.  
  18. 2) Recognize workable patterns. There are 4 that I identified and I'm sure there's other ways to do it. Also, 43 out of the 87 RNG values are goop values, so it's actually 4 in 43 (1 in 10.75, or a 9% chance) of getting a workable pattern. Here are the setups I found before completing the run:
  19. https://youtu.be/Zvv77RTK-z4
  20. https://youtu.be/YK6NYugHajc
  21. https://youtu.be/-0-dlUGRgLs
  22. https://youtu.be/6YsuKHg7IYc
  23. These patterns are characterized by the 4th or the 5th goop being just above samus's arm.
  24.  
  25. However, turrets also advance RNG forward 1 frame every 64 frames (even when destroyed). This may slightly change one goop angle (which might've happened with the 3rd goop in the 2nd setup but I could be wrong). Still, I found during my run that most goop patterns that ended up with the 4th and one pattern with the 5th goop at the right height ended up being workable. Anything else was not a workable pattern (at least with my setup). I don't know much about this, but Strotlog said it could be a 10% chance that this affects the goop angle in a way that throws off what would otherwise be a workable pattern.
  26.  
  27. 3) Time the goop shot accordingly. The 2 RNG values that drop a pb are back-to-back, meaning it's a 2-frame window to shoot the goop. I believe this is based on the timing of when the first goop appears (or when the second one appears in the 3rd setup). When it appears later, it should be shot later. One pattern with the 4th goop at the right height would drop a pb just out of reach (too early), which one could tell because the 1st goop appears significantly earlier. There's also a small chance that turrets could skip one of the pb drop values, essentially making it a frame-perfect goop shot.
  28.  
  29. 4) Time the frame-perfect pb placement accurately. In my opinion, this is the hardest part of the setup, and the part I know the least about. Draygon's position relative to pb placement timing seems to be +/- 1 pixel across the setups. The below images show the frame when the pb needs to be placed in multiple setups. If all the steps are mastered except for this, then all that has to fall into place would be getting a workable pattern and placing the pb at the right frame.
  30. https://imgur.com/a/draygon-underflow-pb-placement-frames-i10NxHa
  31.  
  32. _________________________________
  33. Long answer:
  34. 1) Enter a tiny RNG loop with 87 RNG values compared to the normal 2200+ values. ED found a consistent and easy way to enter this loop:
  35. "Successful confirmed setup to get the RNG stuck/caged in the tiny RNG loop:
  36. NTSC: Having 69 ET energy (no Reserves), and 0x00 (default) sub-unit HP (by not having had heat or acid or lava damage beforehand), and then aligning to the blue door to the right leading into the 1st heated Norfair room with sidehoppers, opening it (with pea shot), and walking into the room to die to only heat (and then continue anywhere else at some previous savestation or ship).
  37. PAL (analogous): Having 69 ET energy (no Reserves), and 0x00 (default) sub-unit HP (by not having had heat or acid or lava damage beforehand), and then aligning to the blue door to the right leading into the 1st heated Norfair room with sidehoppers, opening it (with pea shot), and walking into the room to die to only heat (and then continue anywhere else at some previous savestation or ship)."
  38. *Subunit health must be at 0x00. This can be achieved by taking heat damage, pause abusing at 0 health, turning on Varia suit, and either refilling health from reserves or farming a health drop.
  39.  
  40. More specifically, do not reset to the title screen and do not enter the green or orange rooms shown here if you wish to stay in the RNG loop:
  41. https://imgur.com/a/rooms-not-to-enter-to-keep-rng-loop-LfgTgoe
  42.  
  43. This is the same RNG manip that will make Draygon never slap you, so the fight after the underflow will be significantly easier.
  44.  
  45. I used this tiny RNG loop for 2 categories so far. The first was 14% SuperChargeless in order to increase the probability of desired ammo drops from mochtroids and puyos in Botwoon hallway by 0.73%. This loop fairly increases the odds of super drops. The second was the Kona challenge, where I used it to guarantee that Draygon would never slap me. The manip probably cost me at least 2 minutes, so there's at least a small time cost to doing it in a speedrun.
  46.  
  47. Here are the RNG values of the tiny loop. These values occur in the game sequentially in the order written rather than on shuffle:
  48. 01FF 0B0C 384D 1A92 85EB 9EA8 1A59 84CE 9917 FE84 F9A5 E14A 6783 06A0 2231 AC06 5D2F [D2FC] [1FFD] A102 261B BF98 BF09 BC3E AE47 6874 0B55 39BA 21B3 A990 50E1 9576 EC5F 9EEC 1BAD 8B72 BA4B A488 37B9 17AE 7777 5664 B105 762A 4FE3 9080 D391 22E6 AF8F 6EDC 2B5D D9E2 427B 4D78 8469 971E F4A7 C854 EAB5 969A F213 BB70 AA41 5456 A6BF 42CC 4F0D 8C52 BEAB BA68 A519 3A8E 25D7 BE44 B865 9B0A 0843 2A60 D4F1 29C6 D1EF 1ABC 86BD A2C2 2EDB EB58 99C9
  49. I've bracketed [D2FC] and [1FFD] because those are the two values that result in a pb drop.
  50.  
  51. This means a 2 in 87 (2.3%) chance that shooting a goop will drop a pb. Because the two pb RNG values are back-to-back and always occur in order, this means that a goop shot would be a 2-frame window.
  52.  
  53. 2) One could enter the tiny RNG loop and perform a setup without calculated timing for when to lay the pb and when to shoot the goop. However this would make the odds 2/87 (for the pb drop) times 1/5 (for a hypothetical range of 5 frames for the correct frame to lay the pb. This would make the odds 2/435 or 1/217.5. I could be doing the math wrong, but the odds are no better than this if one blindly goes for the Draygon underflow without prior knowledge on patterns. Having at least one known pattern to look out for would allow a 1/87 chance for the exact series of goop patterns to happen times the probability of the player properly executing the necessary inputs. Theoretically, this underflow could be a 1 in 87 chance with one known working pattern, or 2 in 87 (1 in 43.5) with two known working patterns if execution is perfect.
  54.  
  55. Again, the execution even after a known working pattern is a frame-perfect pb placement followed by a 2-frame lenient goop shot. This will never change.
  56.  
  57. Tewtal found at least one working setup even though he didn't identify the RNG values at which to lay a pb and shoot a goop. However, his setup was in the general 2200+ value loop rather than the 87 value RNG loop. Others may have found working setups, either by luck or through RAM watch, but I'm not aware of them. I found one setup that I happened to encounter two separate times in practice attempts. Here is the setup. I line up with the left wall above the door and morph while facing left, then tap right to face right:
  58. youtube.com/watch?v=Zvv77RTK-z4&themeRefresh=1
  59.  
  60. I used RAM watch after my 2nd encounter and found that when this exact goop pattern occurs, the pb must be placed on 29C6 every time. The player then needs to input shoot on either 01FF or 0B0C for the goop to drop a pb on either D2FC or 1FFD. To be clear, 29C6 is the frame that the pb first appears on the screen, and 01FF/0B0C are the frames in which the player begins to input shoot.
  61.  
  62. I found a second setup that appears to be Draygon entering 1 frame earlier than before. Draygon first enters the screen at D4F1 (Sadly, I didn't get this value for the 1st setup). However, the underflow inputs must be input on the same frames. Lay the pb at D4F1 and begin shot input at 01FF/0B0C:
  63. https://youtu.be/YK6NYugHajc
  64.  
  65. A third setup is shown below. Draygon enters at 0843. Pb must be placed at 2A60 and goop must be shot at 01FF/0B0C:
  66. https://www.youtube.com/watch?v=-0-dlUGRgLs
  67.  
  68. A fourth setup is shown below. Draygon enters at 86BD, pb must be placed at D1EF, and goop must be shot at 01FF/0B0C:
  69. https://www.youtube.com/watch?v=6YsuKHg7IYc
  70.  
  71. In short, the workable patterns appear to require pb placement within a 5-value range: 0843 2A60 D4F1 29C6 D1EF. It's possible that more could work for different setups, but it would likely be near this range. Any RNG values for Draygon appearing on screen that are very different from these setups would likely not be a workable pattern using my setup.
  72.  
  73. Because the 87 RNG values flow in order (and because I was able to encounter some of the same patterns multiple times), any exact pattern should happen 1 in 87 times. The same goes for the other setups. Thus, I believe the final probability is 4/87*probability of correct timing for inputs. This is at least a small improvement over the previous plan to blindly lay a pb and shoot goop hoping for the best.
  74.  
  75. However, turrets advance RNG by a frame every 64 frames. I won't get into the details because I don't understand them, but Strotlog says that this can interfere with the goop angle 6/64 times (10% of the time.
  76.  
  77. The following potential discoveries could improve these odds even more:
  78. 1) Additional working setups. Tewtal's setup shows that there could be more than the four that I found, but he likely tested in the general loop and not the small loop. Another setup (given the player can recognize them in time) can improve the odds to better than 1 in 10.75.
  79. 2) RNG normalization to consistently get the needed pattern. This could potentially be done by observing turret patterns or bubble timing from Samus's helmet. However, I don't even know where to begin to find such a normalization. Knowing the RNG values for the timing of the Draygon underflow could be beneficial for finding a setup.
  80. 3) An even smaller RNG loop with at least one pb drop value and at least one goop value. However, the 87-value loop currently known is the smallest one that I'm aware of.
  81.  
  82. With now 4 RNG setups, the category has finally been completed. The odds of achieving the underflow at any given right-to-left goop pattern is as high as 1 in 10.75.
Advertisement
Add Comment
Please, Sign In to add comment