Posts for ledauphinbenoit


Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
Hmmm. So many factors to be considered! Great job so far dave_dwfm. Your run is shaping up quite nicely so far. Some thoughts: adelikat suggested a list of 5 items to get using the glitch. Another was suggested afterward. At 1:30 each, that's between 7:30 and 9:00 of real time. That's quite a bit. Are there any other movies on this site that have 7-9 minutes of absolute exact repetitiveness? None that I know of. I love the glitch idea, but it will need to be used sparingly to avoid putting the unfamiliar audience to sleep. Using dave_dwfm's run, we should be able to gauge the amount of time things take. The cave west of Noaniels would take the place of the Pyramid. Go in, get the ruby, die, pick up party member #2, warp to Romaly. This is almost sad, because the Pyramid is the one "truly unique" dungeon in the whole game. Oh well.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
That's useful information. Here are the links:
Bestiary http://www.geocities.com/arcanelore2001/articles/dw3bestiary.txt Enemy spell resistances: http://www.geocities.com/arcanelore2001/articles/dw3monspell.txt Enemy action tables: http://www.geocities.com/arcanelore2001/articles/dw3monact.txt Enemy encounter tables: http://www.geocities.com/arcanelore2001/articles/dw3monenc.txt
Looks like a Thor's Sword from the Boss Troll is 1/256, which is about as expected. Not quite as bad as those Stilleto Earrings Adelikat was trying for in DW4 (1/2048).
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
I've done some more testing on my theory about needing to only kill the Orochi and the Boss Troll to get the NewTown to grow so you can get the Yellow Orb. Edit: Update. Here is what the logic is for NewTown Advancement Level 0: drop off merchant Level 1: collect any 1 orb or kill the boss troll. There may be other things to move to level 1 as well. Level 2: collect the purple orb. If you have only the purple orb, you can advance from level 1 to level 2 by exiting and entering. Level 3: have the purple orb and kill the boss troll. collecting the change staff is not required to get the town to advance (although you would always need it anyway). Level 4 and yellow orb: have the purple orb, kill the boss troll, and have at least 2 other orbs (red, green, blue, silver) Edit 2: Aparently you must also cast return or use a wing of wyvern to acheive the final growth This leaves a little flexibility in ordering of orb collection. If it works out that silver is best kept till last, then at least one other orb can be put off until after getting yellow.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
I dug up the route I had roughly mapped out last summer. It was really just a rough draft. Here is the pertaining section:
Return to Aliahan
Sail to Lancel
Get invis herbs
Sail to Edinburg from Portoga? or Noaniels?
Get Vase of Drought
Sail West
Fight metal babbles (enough to kill Orochi and Boss Troll)
Notes: Edinbear is closest town to metal babble location

Return to Baharata 
Go to Jipang
Kill Orochi
Get purple orb
Sail to reef 
Get final key
Notes: No keys needed to fight Orochi, only have to visit Jipang once

Return to ?
go to Samanao cave
get mirror of ra
fight boss troll
get staff of change

Return to Aliahan? and get merchant
Sail to house of pirates
Get red orb

Return to Edinbear
Sail to new town
Get yellow orb
sail to greenlad
get sailor's thighbone
(level more?)

Return to Romaly
get locket of love off of phantom ship

Return to Noaniels?
go to jail
get sword of gaia

Return to Baharata
Go to volcano
Go to cave of necrogond
Fight more metal babbles (enough to kill Baramos)
Get silver orb

Go/Return to baharata
get green orb from tedanki
Note: Tedanki is closest to Liamland, so visit on the way there

Sail to Liamland
Get Ramia
As Adelikat mentioned, it might be faster to sail from Edinbear directly to the shoal shrine (maybe getting metal babbles on the way) and then hit up Jipang later. I tried to minimize the amount of revisits to recurring locations, but things like the stupid house of pirates are in the middle of nowhere. Metal babbles in the negrogond cave also add a way to break up the amount of leveling needed if that is desirable. Edit: I just realized that I left the Navel of the Earth off of the route, but that's probably because you can do it anytime after you have the final key. It's pretty independent.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
Poison needle and bomb crags is a great idea! Plus, it shows off the "super leveling" area east of Kanave. Nice work. Sounds like your patience with the battling system is working pretty well so far. Keep up the good work!
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
Well, those are good reasons for the extra steps. Always good to check in advance! Unfortunately stuff like that makes things look sloppier than in other TASs, but there's not much we can do about that in this game.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
Hi Dave, I'm glad you've taken an interest in this game. It's a toughie, but you seem to have done some good testing/research so far. Some notes on your WIP: 1) From the start you take an interesting strategy: getting a party right away. You use this well to hold lots of items and make use of the pilgrim's heal. You've convinced me it might be worthwhile. The real question is, how worthwhile is it? I know it doesn't cost much to revive level 1 chars, but right where your movie ends is where you will need to spend some agonizing time with the priest listening to the revival tune 3 times. How did you decide that it was better to make it to the golden claw with a party? 2) It seems like you don't always take the optimal path in some areas. For example, in the cave of enticement you take a few extra steps. Also, when entering the Pyramid and Isis for the first time. Both of those are pretty minor though, compared to the main difficulties ahead, i.e. fighting. I'm still worried that it's all going to have to be pure brute force. I've become reinterested in this by your progress though. So I'll try to test some things when I get a chance.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
Yeah, I was going to say it's not impossible. I don't think I ever used Drumstick when I got him. I actually thought Drumstick stunk pretty bad for a character you can unlock.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
I love how every boss battle is different and entertaining. Watching the boss of Venom Surface really was confusing/amusing. He just kept floating around in circles and you randomly pegged him. Then he just sort of disappears. No big explosion. Nothing. Just whimpers away. It was a good ending to the stage considering that was probably the least "lively" one yet. I know Venom 2 is crazy busy, so it should have plenty of action.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
Wow, you toasted the Dodora. It happened so fast it seemed like glitch exploitation. Is it really just accurate shooting that takes it down? Also, is where you pick up the 2nd twin blasters really the first point you can do so on route 3? I never played route 3 much since it was so darn hard. It's nice to see Fox McCloud live up to the hype the storylines gives him about being an ace pilot. I always laughed when I played it because they would talk so big about him and then I would take control and crash into all sorts of stuff. Then again, this game seems infatuated with large damage causing objects randomly appearing in front of you out of thin air.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
yeah, that would be the difficulty in adding things like invisible doors (i.e. Zelda on NES). If you can get inside of the wall, you can stay inside the wall and move around.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
The Casino Box and whatever the other box you can buy is called only work for coins. Everything will depend on luck manipulation. Can it be done?
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
I agree with the decision for route: Find Cheato books, but don't enter codes No Stop and Swap items This way having to backtrack and pound out all the codes in Treasure Trove Cove is avoided. This will make things a bit more fast-paced from level to level.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
I'm no pro myself, but having been around for a while, I can say that a TAS must be of a well known game or really entertaining to make it. Since this game isn't really all that well known, it would need to be really entertaining to watch, especially if it's going to be long. That is the real challenge to overcome for long games, keeping it lively and entertaining. Some games just aren't setup to make entertaining movies.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
Hmm, I don't know. I think you could make a valid argument either way for both mumbo tokens and cheats since they are both recorded by the game as permanent. There is a fixed number of mumbo tokens in the game, although you don't need them all, so I think that could slide. I would consider that showing all 3 Cheato locations would be a cool feature to exhibit 100% of the game. Then again, this disagrees with the SDA convention of what merits 100%.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
It is probably worth mentioning that SDA has a single segment run of this game. It is just under 3 hours and quite an achievement. Don't forget to reference it when route planning!
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
The original topic is here.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
I think an any % run should be done first as these platformers are a pretty massive undertaking. Besides, then transformations can be avoided altogether.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
Hmmm, after Shadow Temple it might be faster to exit to Hyrule Field. If you could get a straight trajectory, you could slide to nearby the Kokiri Forest that might be faster than the Goron City warp.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
Too bad you can't get more bombchus. It would be easier to bomb the rolling goron for a bomb bag instead of having to manipulate bombchu bowling.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
I would look for big sequence breakers. In both Mario 64 and Ocarina of Time they have contributed to making the run both faster and more entertaining. Glitching into a later area of the game early is not only cool, but often allows you to not even complete earlier levels (as in OoT).
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
mwl wrote:
No desynchs for me running this .m64 from the very beginning. Can anyone else confirm?
No desyncs here. It ran fine from the beginning.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
For those who don't want to go searching: Route (Mirror)
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
AKA wrote:
It desynced during the fishing game.
It did for me as well using a savestate I had made during the Spirit Temple. Playing it back from the beginning, I had no problems though, so you might try that.
Experienced Forum User
Joined: 8/29/2005
Posts: 148
Location: Dayton, OH
An update on a few things: I think we can pretty much rule out a TAS of this game until we find out more about the game internals, specifically the RNG for random encounters. I did a short test a month or so ago, which I have now uploaded. It's 10 minutes long. Consider it an example of how hard a full TAS would be. (Note the route has changed a bit since, see above posts) I've done some low-level investigation into the assembly, but I am nowhere near good enough to find a RNG. I have browsed some material on the web for finding things in the ram and assembly, but nothing this complex. Since it seems to base encounters on number of steps, we need to know when the amount of steps before a battle gets set and manipulate it. Then we can proceed only when that number is very high (maybe using a bot if necessary). The same goes for being able to know that a battle will yield critical hits or appropriate behavior from enemies. So...things to figure out: random number of steps to a battle (both outside and in caves) random numbers for battle effects (criticals, dodges, spells working, item drops, etc.) random number for enemy formations (for getting metal slimes, etc.)