Posts for primorial_soup

1 2 3 4 5
13 14
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Hmm. Yes, I did miss that. Shit.
primorial#soup wrote:
Here's the game plan. During the first 500 steps, bike back down to Fuschia Gym, fight Koga, and walk back out. This take 192 steps. While there, be sure to get poisoned, and to take full damage from an Explosion from Koga's Weezing. Then Fly to Celedon and do the Gym there. This take 130 steps. The rest of the steps could be used by burning off hit points biking around below the Gym there. In the meantime, bike down to Cinnibar, and then Fly to fight Sabrina, being sure to pick up poison from Sabrina's Venomoth. During the second 500 steps, do the Cinnibar mansion. This takes 222 steps. Then do Cinnibar Gym, 104 steps. Once again, a little time will need to be wasted afterwards, but I think that this is the best usage of time possible.
I plan to test this to make sure that it isn't just Pokémon Yellow specific, but it seems likely that it's not. If it is in fact the case the using Escape Rope cancels the Safari Zone glitch, I think there's only two options. The first would be to proceed as I had planned, but wait to set up the second WTW glitch until after the Cinnibar Gym. This would mean biking around for 500 steps doing absolutely nothing; there's nothing left to be done. The second would be to do the WTW glitch three times, once for Sabrina, once for Blaine, and once for Victory Road. The only time difference between these two is walking through Cinnibar Mansion, and setting up the WTW glitch an additional time. My gut feeling is that walking through the Mansion will still be significantly faster, although I'll need to take a measurement of this. Additionally, I've hit another road block. I had planned to save Bubblebeam for Mew, and complete Cerulean to Vermillion with Gyarados having just Bite, Dragon Rage, and Hydro Pump. However, Bite isn't quite as effective as I thought it would be, and is causing numerous 2HKOs where Bubblebeam didn't even need to crit. Not only does this waste a lot of time, but I won't have enough attack PP to make it the whole way. This means that I need a different TM for Mew. In order to make it through the Elite Four, it needs to be an attack with at least 10 PP, which excludes Fire Blast. Right now Psychic is looking like the best option. I had concluded that it was too far out of the way, but it looks like I'll have more than enough time to burn.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Updated WIP with progress through Mt. Moon, now 937 frames (15.62s) faster than the currently published run at this point. The amazing part is, that as this point the runs are nearly identical, which leads me to believe that even without any major route changes, the current run could be improved by as much as 30 seconds. On Route 3, you may notice that I use a critical hit against the Youngster's Ekans which I didn't in the previous run. This is not a mistake. In SGB mode, whenever the opponents health bar changes color (in this case, from green to yellow), it takes 15 frames. A critical hit, on the otherhand, takes 20. However, this Ekans has 30 hit points, and a max normal hit would only do 15. This would mean needed two max normal hits in a row, which requires a lot of manipulation. In comparison, a weak normal hit (so that the health bar doesn't change color), and a weak critical takes approximately zero manipulation. Long story short, I saved 3 frames over the fastest double max combo I found, and 7 frames over the previous run.
Chamale wrote:
It is in fact faster to fight the trainer with Rattata and Nidoran, then backtrack to the outskirts of Pewter, and then advance again. Here's an unoptimized test run, which is at least 2.5 seconds faster despite the fact that Charmander takes a hit from Nidoran.
This was a big time saver, but not in the way you think. By my tests, this route is in fact 62 frames slower. The comparison you made wasn't quite valid; after fighting the trainer with two pokémon, Charmander is still level 14. After the trainer with four pokémon, he is 15, and has sat through the annoyingly long "level up" and "you just learned Leer" jingles, which are approximately 160 frames each; a total of around 5 seconds that you forgot to take into account. However, this saves a lot of time in Mt. Moon. With less experience points, Charmander does not grow to level 16, saving ~160 frames, and does not attempt to evolve, saving another ~380 frames. Of the 600 frames (10s) saved in this section, approximately 480 (8s) can be attributed to this new route. EDIT: It turns out that switching Dragon Rage to the top was an error. This segment is now 12 frames faster.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
petrie911 wrote:
I'm pretty sure the Cascadebadge gave you the ability to Cut and Thunderbadge (or w/e Lt. Surge's badge is) gave you ability to Fly. Does that change anything?
Err, I meant Cut, not Fly. But it wouldn't change anything, because I need both Cut and Fly to perform the second Snorlax skip.
TheRandomPie_IV wrote:
In that case might it be worth holding off on getting mew untill the second skip and using the uncessary trainer to generate the mew? Or are there no suitable trainers near by for that?
This gave me a good idea. Unfortunately, what you suggest won't work, because I will have had to have Mew before then to use Fly. However, just because I have Mew, it doesn't mean I need to use her as the main battler. At this point, Gyarados is plenty strong enough to take care of the trainers along routes 8 and 12. Attack PP won't be a problem either, as we've just healed in Vermillion for the Snorlax skip. Because his battle cry is significantly shorter than Mew's, this should save a good bit of time.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
It's been a while, but I finally have enough free time to TAS again. I present you with the current WIP with progress through Pewter Gym, now 338 frames (5.63s) faster than the currently published run at this point. I was 164 frames ahead before the gym, so this means I picked up an additional 174 frames, 23 frames from better manipulation on the Geodude, and 151 frames from a much better strategy against the Onix. One thing I'm still unsure about is the second Snorlax skip. Ideally, the trigger trainer would be one that needs to be defeated anyway. I can't use Erika or St. Surge, because both require using Cut to get to, and the start button is disabled. I don't see any other good candidates either. Misty can't be used because Cut is required to get the the secret house where you get Fly (and the Cascadebadge is required to use Fly), and all of the other trainers between Vermillion and Celedon have already been fought. I may end up having to fight an unnecessary trainer.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
It turns out that I was wrong about the optimality of this run. I've since managed to improve it by 18 frames (Updated Version). LagDotCom had noticed that by pressing Start at the intro screen instead of A, the next button press comes one frame earlier. With a little experimentation, I found that by pressing Start, A, Start (or also A, Start, A) for the first three screens, one could save two frames instead of just one. This game has very few frame rules: a two frame rule for using A to pick up or access something (such as the PC or starting pokémon), a three (?) frame rule for the last button press of a dialog, and a two frame rule whenever the game forces your character to move (such as following Prof. Oak, and in the Poké Mart retrieving Oak's Parcel). Being two frames ahead seems to have put me in a much better position for these frame rules, and I saved another 6 frames from this alone. The other ten frames come from better manipulation. The current manipulation now looks like this:
5326  | 2 frames | Starting stats
6645  | 2 frames | First max critical
7017  | 2 frames | First max critical + tackle miss
7286  | 1 frame  | Second max critical
7520  | 1 frame  | Tackle miss (lag frame)
7604  | 2 frames | Max normal
8208  | 1 frame  | Random encounter avoidance
Total |12 frames |
This looks like it should only be nine frames faster, but a manipulation technique that I had been using which I thought to be frameless was actually causing one lag frame. The runs looks nearly identical (other than attack order, and a graphical glitch I included in the new one), so I wonder if I should cancel and resubmit, or would it be better for an operator to replace the movie file?
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
LagDotCom wrote:
One thing though - I was often able to manipulate random encounters away just by holding A while moving; was this impossible for yours?
Yes, this works, and the reason why is because it slows you down by two frames per tile whenever you do this. The best way to manipulate away random encounters is to change your route slightly, entering the patch of grass at different times or different locations. However, when the route is invariable, this can't be done. One needs to wait two frames because movement input is only accepted every other frame. Typically, a better alternative is to go back to the last dialog button press and wait one frame (thereby saving a frame), but in this instance, it also caused a random encounter in the same location, so I was forced to wait two frames.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
If I were to submit a run, I'd probably start fresh. The reason is that even if the door had worked two steps earlier, I would have been 8 frames from sub 4:25, which would be a shame, considering it's certainly possible. I also made a two frame mistake early on by turning off battle switching (force of habit, I guess), and I used a lot more frames for manipulation that I would have liked (21 total, compared to your 13 total). If you watch my battle frame by frame, perhaps you could pick up a few manipulation techniques, especially regarding the Tackle misses.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
I've made a test run which finishes 165 frames faster than this submission. The main difference is the rival battle, which manipulates two max crits, a max normal, and two tackle misses. I was about 200 frames ahead, but for some reason I couldn't get the correct warp point as quickly as you did... it took me two extra steps (36 frames). Oddly enough, the first time I tried it, it actually took me four extra steps. By waiting a little bit in the Poké Mart in Veridian I was able to change this. Perhaps if one waited long enough, one could save even more time? I think it's worth testing.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
P.JBoy wrote:
I thought someone said that using the "LDR" glitch takes away the holding down bonus
It would, but I don't ever use LDR while the pieces are falling (with exception to one piece at the very end, where I needed a larger manipulation window). Instead, I use it after the current piece has already landed, which causes it not to flash, as written in the movie description. Also, holding down as the piece lands often (read: usually) causes the next piece to appear one frame earlier. So in general, it should always be done, unless for purposes of manipulation.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
FractalFusion wrote:
On a less serious note, the last 20 posts in this thread have all been from Canadians. :)
You're canadian? I'm so sorry. I didn't know...
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Chef Stef wrote:
While making this movie, I discovered that holding down as the current block hits the ground will sometimes change the order of blocks. It seems to have very limited application, though (since you can only choose from one extra block). I used it once or twice in this movie.
This is because holding down for at least four frames (at least two frames before the block lands, the frame it lands, and one frame after it lands) causes the next block to appear one frame earlier. Holding down for at least six frames will give you points as well; one point for every two frames beyond the necessary four. There's also a small window for manipulation just after the current block lands and before the next block appears (starting at 11 frames before it appears). This window is five frames long, which allows pressing and releasing left+right+down up to three times, each of which delay the next block by one frame (typically giving one the option between four different blocks, which is usually enough). This has a nice side-effect as well, in that it doesn't cause the current block to flash, but only the next block, which is even less distracting. I know all this because P.JBoy suggested that perhaps Acmlm existing run could be modified to use this form of manipulation instead of pausing. I decided to give it a shot, but unfortunately, it's not possible. Left+right+down interferes with moving pieces left and right, and with pieces stacked very high, the window for manipulation just isn't large enough.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
I enjoyed watching this movie; it is not only faster than your previous submission, it also does not have the pauses to manipulate the pieces. However, I wonder if it could not be faster. One only needs to finish 25 lines to finish a level in B mode... in this movie you complete 26?
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Chamale wrote:
About the run, Primo said (in a PM) that he's still working on luck-manipulating in the first gym. Apparently he doesn't have much spare time lately :(
I'm currently in my exam period, so this is likely to remain the case for the next few weeks. Over the summer months, I should have a lot of time though.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Admiral_Sif wrote:
Also, how does sam weigh more than the meat?
He ate his Wheaties™ for breakfast.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
This is also one of my all-time favorite games. I thought that Sami's run was pure awesome (and it was), but this run takes awesome to a new level. Thank you for your time and devotion to this project; I think the end result was worth it.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Chamale wrote:
"You're wasting my ear capacity" - Austin
I might begin using this.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
adelikat wrote:
Wow, this game actually looks impressive now.
My thoughts exactly.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
FractalFusion wrote:
When running Pokemon Yellow, I tested this, and it is faster to fight the trainer with 4, considering that I have a L100 Nidoking.
Ahh yes, Route 3. Against 4 bug pokémon, Charmander is in a similar situation.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
TheRandomPie_IV wrote:
Wouldn't you just warp from Viridian to Palet, as then you wouldn't need to heal? Or does the time saved by ledge jumps exceed the time it takes to heal?
First, yes, this would make a lot more sense than healing Veridian, and then death warping back there. However, there are no Weedles (or any enemies with poison) on route 1 (the Nidorans don't have poison at that level either), so the only option would be to fight a random encounter, and die slowly. Overhead for a random encounter is around 1500 frames (just counting starting and finishing). Each round of attacks adds about 600 frames. Assuming something around three rounds of attacks to die, this would take an estimated 3300 frames; around 55 seconds. Walking back is quite a bit faster. There might be other places where this could be a possibility, however.
TheRandomPie_IV wrote:
Edit: What about Chamale's old idea of fighting the trainer to the right of Pewter for extra money?
Which trainer? The first fightable trainers after the first rival battle are the second rival battle on Route 22, and the Bug Catchers in Veridian Forest.
Chamale wrote:
Yes... IIRC, no one's ever done testing as to whether it's faster to beat that 2-pokémon trainer and walk back to the mart as compared to beating the 4-pokémon trainer. If the deathwarp adds anything, then it looks even better.
I honestly have no idea what you're talking about.
Mr. Pwnage wrote:
How many Escape Ropes does the current path call for?
The current route calls for 5: 1. Mt. Moon for Gyarados glitching. 2. From Bill's house back to the Poké Center. 3. Saffron Gym 4. Cinnibar Mansion 5. Cinnibar Gym The difference of three is due to the following changes: - Teleport will be used to warp back to Cerulean. I believe that this is faster than having to use Cut twice before and after the Gym. - Silph Co. Office Building will be skipped. - Picking up the Surf TM in Safari Zone will be skipped.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Chamale wrote:
Well, the only part I've had happen on console was the random battle part. The rest was a guess based on incorrect knowledge of what would happen if you fainted in battle (That has never happened to me). Here's another thought, though: Faint from poison in front of a trainer you can trainer-fly against normally. Then you'll have an opportunity to use the glitch when normally you couldn't.
I have tested this, as stated above. The result is that you are returned to the last Poké Center without activating the glitch.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Chamale wrote:
BRILLIANT IDEA! Start with Squirtle instead of Charmander. In Viridian Forest, run into a wild Pikachu in front of a trainer (I've had this happen on console). Get owned by Pikachu. You white out and return to the last pokemon centre, thus triggering a trainer-fly glitch.
Hmm. I experimented earlier with this, fainting from poison directly in front of a trainer (result, teleported directly without triggering the glitch). I'll experiment with this as well.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Phallosvogel wrote:
uhm might just be me not understanding what ypoure saying but: fight animations? arent they switched off at all??
kirbymuncher wrote:
Some moves take longer to work even with animations off.
This is exactly right. There seems to be two different types of "animation-off animations." In the faster of the two, the whole screen moves quickly right and then left. In the slower, the enemy flashes for something like a second and a half. The difference in time between the two is, from my measurements, 55 frames (nearly a second penalty for every use). In general, attacks with a fast animation-on animation, such as most starting attacks, also use the slower animation-off animation. Bubblebeam and Thunderbolt are both fast, and obtained early on, which makes them rather ideal.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Chamale wrote:
I still think it's a bad option. Keep in mind the message is longer than "Critical hit!" The only reason it should be discussed would be if Bubblebeam or Cut couldn't take out a fairly large number of enemy pokémon.
The plan I had was to keep Pound, and in addition learn Bubblebeam and Thunderbolt. Charmander will have already learned Cut, so to teach it to Mew would mean teaching it an additional time, which seems unnecessary. The only problem with this plan is that Pound has a slow animation (Cut also does). I shouldn't need to use it very often though, so this shouldn't be a big deal. I've also decided against inventory stuffing; because so much of the game is being skipped, there just aren't enough (quick) opportunities to pick up extra items. As for progess, I'm currently in Pewter Gym, attempting to roll the second of three max critical hits against Geodude. When I finish with the Gym I'll post another WIP.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Mr. Pwnage wrote:
...set up wall clipping, then do Glitch City again...
The only problem with this is that entering or exiting a building (such as the entrance to Safari Zone, or the house with the Fly HM) deactivates the ability to walk through walls.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Chamale wrote:
I think the first option might be better, because Gyarados will be at just level 7, so the glitch might not be worth executing.
You may not have seen the currently published run in a while, but even at L7, Gyarados is much stronger than Charmander at L16, mainly due to having Dragon Rage (40 points damage guaranteed). I've done some rough estimates from my previous two runs. Not including the time Gyarados spend at the Gym, Charmander wins across Nugget Bridge by about 1688 frames. However, at this point Mew hasn't been glitched yet. In the Gyarados path, Mew is glitched as a by-product of performing the Snorlax skip, so the only thing accoutable is the 800 or so frames for seeing a random encounter, and running away (for comparison, the 6 times Growl is used is the same for both, so it isn't weighted against one or the other). I'm measuring from when the battle begins with the first Hiker on the way to Bill's. At this point, even though he's finished the entire Gym already, Gyarados is already ahead by 658 frames. However, an Abra has not yet been caught, and there's still the 800 frames from the extra random encounter later. Factoring out Gym time, and adding a gracious 1800 frames for catching an Abra (1600 would probably be sufficient), and the 800 frames for the random encounter, the Gyarados route wins by around 6000 frames (over a minute and a half). Then there is the fact that Mew has a much longer cry than Gyarados, so I think this makes Gyarados the clear winner. So, I just had an awesome Idea. The glitch only needs to be performed twice, and here's why. After performing the glitch, bike down to Cinnibar. Once you're there, you're there, and can Fly back at any time. Now, Fly to Saffron to fight Sabrina. You'd still have to go through the Pokémon Mansion, but this is still years faster than performing the glitch yet another time, and plus, making your way through the Mansion give the appearance that you're actually doing something useful, rather than just biking around for 500 steps. EDIT: Here's the game plan. During the first 500 steps, bike back down to Fuschia Gym, fight Koga, and walk back out. This take 192 steps. While there, be sure to get poisoned, and to take full damage from an Explosion from Koga's Weezing. Then Fly to Celedon and do the Gym there. This take 130 steps. The rest of the steps could be used by burning off hit points biking around below the Gym there. In the meantime, bike down to Cinnibar, and then Fly to fight Sabrina, being sure to pick up poison from Sabrina's Venomoth. During the second 500 steps, do the Cinnibar mansion. This takes 222 steps. Then do Cinnibar Gym, 104 steps. Once again, a little time will need to be wasted afterwards, but I think that this is the best usage of time possible.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
1 2 3 4 5
13 14