I also thought about this deathwarp.
How would the route look like at the moment? I have some questions about that:
- What element will be used to kill Solvaring?
- Do we go to Glencoe Forest? I think it is be required for fast Zelse and Nepty battles, there are just too much spirits in there.
- Would using the Blue Wings really be faster? The way from the forest to the town isn't that long.
- I guess you know that talking to the girl in Larapool isn't required to lower the water? Just enter and exit the room. In their testruns, Animeowzerz and Peaches didn't seem to know that, so I told told them about it.
- Theres that spirit, number 51 in the gamefaqs guide. It's somewehre on the ship, but I don't remember it exactly. Can this be collected on the first trip to West Carmagh?
- Will we go to Limelin? If we do, getting the Red Wings should save time.
Current project: Gex 3 any%
Paused: Gex 64 any%
There are no N64 emulators. Just SM64 emulators with hacky support for all the other games.
On peaches_ latest speedrun attempt, he found himself missing a lot against Shilf and Guilty. After that he tested what affects your shots missing, and it turns out your own agility matters, not just the opponent's.
We need someone to test if the Agi+ from the Evade spell affects hit%, or if it's just dodge%. Dumping the 7 points into wind after Nepty might actually be worth it if it means hitting more often on the harder bosses without needing to Agi grind.
- What element will be used to kill Solvaring?
Water and Earth seem to be the most profitable in the long run. Wind and Fire happen to be the best against solvering though, which sucks.
- Do we go to Glencoe Forest? I think it is be required for fast Zelse and Nepty battles, there are just too much spirits in there.
In a TAS, avalanche can 2 shot Zelse and 3 shot Nepty if earth is high enough. It's a lot of time, we would have to measure a fair gain somewhere for it to be worthwhile.
- Would using the Blue Wings really be faster? The way from the forest to the town isn't that long.
It depends whether you get exit or not I think.
- Theres that spirit, number 51 in the gamefaqs guide. It's somewehre on the ship, but I don't remember it exactly. Can this be collected on the first trip to West Carmagh?
GameFAQs: "Now, before you get off in Limelin, there's another weird spirit element here. I'm not convinced it's not a glitch. However, to get this element, get on the boat at West Carmagh and go below. Come back up and see you're at Limelin. Get off the boat, get back on and go below again. Now look in the crate to the right of the door and there's a... SPIRIT ELEMENT? No, it wasn't there before and no, I don't get it either. Anyway, travel back to Limelin again and get off the boat." --You have to ride the boat in a full cycle to get back to Limelin. It's dumb, but it doesn't cost a terrible amount of time compared to some other spirits.
- Will we go to Limelin? If we do, getting the Red Wings should save time.
Hard to tell right now. I am aware of the potential of the wings. I really don't see any reason to do anything but get magic barrier, avalanche, and a few MP regen items. Dealing with random battles is a huge factor too though. I have a couple theories on this that I will look into in the coming days. Instant casting during the battle intro is a good start.
- We need someone to test if the Agi+ from the Evade spell affects hit%, or if it's just dodge%
I can maybe look into that. The thing is that Agi+ costs a turn, and avalanche is already random enough. I can't see any gain being worth the loss of a turn and the investment into wind.
--
I have a working document going for non-TAS, presumably the things necessary in a TAS are almost entirely a subset of the things needed for TAS for this game. I'll post that when I get it semi-finalized. I was going to time out many of the spirits eventually, too.
Side Note: Guilty is extremely powerful.
Here is my working document complete with all spirits, and I think all the chests. I also wrote directions to follow for non-TAS. The spirit counts before the early bosses assume no encounters. If I missed any chests please let me know. (The FAQs aren't terribly reliable.)
http://pastebin.com/gwgXYGrh
In general, only grab items that are needed. Giant's Shoes are probably only usable at Solvering, Silent Flutes are worthless, Silver Amulets and Celine's Bells are worthless. Use Mint Leaves before Larapool and grab a free one in the inn. Replicas on the path may be useful for running from the enemy 'Rocky' in Boil Hole non-TAS. Any spirit costing ~45 seconds or more after Fargo is definitely not worth it.
I think I found out why encounters are so strange.
00977FFA (unsigned short)
This value increments each time you walk 50 units. It also increments when the encounter values (7FF4 and 7FF6) rollover. It seems when this number gets larger, your probability of getting an encounter on the next rollover gets higher. So I think this number sort of represents "Probability of encounter on next rollover." This number refreshes any time you enter a battle, or enter a new loading zone.
Donamer's attempts to dodge enemies, as it would have rolled to:
Plains: 1200, 650, 800, 1050, 1050, 550, 200 (Intentional?)
Connor Forest: 1100, 1950, (1450 no encounters), (Intentional at 250)
Plains: (1400 no encounters)
Connor Forest (1200 no encounters)
Dondoran Flats: 1200, 1150, (1700 no encounters)
Glencoe Forest: 1250, 1550, 1600, 1550 (1550 no encounters)
Dondoran Flats: 1150, 800, 1400, (1100 no encounters)
It seems getting over 2000 is very difficult. I'm trying to figure out how to see if certain corners of the map have no encounters or not. I tried searching for a byte that is "in a no encounter zone vs in an encounter zone", it doesn't seem to exist.
Thank you Kirkq, this was a great find! Now I understand why there are some places that ALWAYS start a battle. Remember that last Blood Jell encounter in Cull Hazard or the part in Baragoon Tunnel right after the long bridge? Thats because the counter still increases on those places. You can't trigger any battles on places like bridges or stairs, but they are still on the same map.
Oh and I agree with you about Guilty. I always found him being the hardest boss!
Current project: Gex 3 any%
Paused: Gex 64 any%
There are no N64 emulators. Just SM64 emulators with hacky support for all the other games.
Finally. I maxed Brian. I finally got all spirits to 50. Yes, it's possible if anyone was wondering....there aren't 2 or 1 spirits missing, nor is the battling system flawed in that you can't max the spirits if you haven't fought enough. That is, you may never get a spirit from a battle and decide you want to start training right before Guilty or even Mammon; well you can get all the battling spirits at that time. Few I'm just so glad I finally managed to do this after years of having this as something of a goal lol.
Wow what a sigh of relief, I just had to share this with someone and this place was the first thing that came to my mind. Oh well, just glad I said that :P
If anyone's willing to tas this game good luck, it certainly isn't a great choice to of game to tas (playing this game has been very fun though), but definitely worth watching at least once when it's done, just to see how shocking of an outcome a tas can bring.
And I'm certainly willing to help with criticism and other personal information only, too bored to tas :P
I wish a 100% run of this game was possible, but it would take forever.
Probably looking at about 10 hours or grinding just to get Spirits to Max Level.
I wrote a decent faq that has a good amount of info on spells, stats, and enemy list.
http://www.gamefaqs.com/n64/198386-quest-64/faqs/66316
I have been doing some testing to figure out more stuff. I recently found out some interesting info while debugging the game.
Is anyone interested in helping me understand the RNG? So far I found the subroutine for the damage, and the address that determines the randomness of the damage.
I still question your methodology.
Solvaring's "Earth Spikes" can definitely Miss and you give it a 100% accuracy, and then you give other attacks a 90% accuracy, which just seems like numbers you are pulling out of your ass because I'm fairly confident the Miss chance is agility dependent.
You have 90% accuracy on Magic Barrier and 90% accuracy on Confusion. From my experiences Confusion barely ever misses, which is less often than magic barrier. Also from a coding standpoint 15/16 chance = 93.75% makes way more sense than 90% = 231/256. (EDIT: Maybe this is wrong?)
From my experiences, Escape has around a 1/3 = 33% chance of working. You give it 55%. From experience this seems to go up with agility level? (EDIT: ?)
I think you might have some useful information in your FAQ, but your methodology is unexplained and perhaps inaccurate in some spots.
You are right about my methodology being unexplained, mostly due to the fact I submitted the faq kinda early and I'm not done with my testing. I'll add more to my faq once I understand the mechanics behind it.
For accuracy, there are multiple factors. 100 accuracy in a spell does not mean it will always hit because your agility and your enemy's agility also factor in whether an attack hits or misses. RNG seems to affect it too. The numbers I give are the actual values given to the spells. I didn't make anything up. The only spells that seem to go purely by the accuracy of the spell are spells that effect yourself, like healing which has 100 accuracy by default.
If you want to take a look at the spell variables, I can give you the N64 addresses to them. Keep in mind I'm using NTSC, if you're PAL, i can go search it for you.
The addresses of the spells go in order from Fire to Earth to Water to Wind. Within the elements, the order goes by the level requirement. So the beggining of the addresses for spells is 800C06A0 . Which is Fire Ball Level 1. Since I use cheat engine to keep a list of addresses, I will be using the Little Endian format, so if you are using Big Endian, you will have the swap the last byte of the address. For example 8007BAA4 in little endian = 8007BAA7 in big endian.
Also I don't fully know what each variable does so I will just skip them.
800C06A0 is the address that determines Area of Effect for the spell as well as when you are able to cast the spell (inside of battle, outside of battle, or both). I'm not sure what 800C06A1 does, but 800C06A2 determines the Level Requirement of the spell. 800C06A4 - 800C06A7 seems to effect menu. 800C06A8 - 800C06A9 effect the number of frames the attack lasts. This is the main variable that affects range aside from whether or not it's a projectile. 800C06AC - 800C06AD is Accuracy. 800C06AE - 800C06AF is damage.
If you want to write down or look at the addresses for the spell variables, here's the order in Little Endian.
Byte 1 determines AOE and when you can cast spell
Byte 2 Something to do with hit stun
Byte 3-4 Determines Level Requirement
Byte 5-8 Menu stuff
Byte 9-10 Range (seems to just be the number of frames it lasts)
Byte 11 Unknown
Byte 12 Something to do with menu
Byte 13-14 Accuracy
Byte 15-16 Damage
Byte 17-18 Projectile (whether moves or not)
Byte 19-20 Determines whether you do damage, MP drain, healing, Invalidity, etc
Byte 21-22 Target (whether auto hits or not)
Byte 23-24 Unknown
Byte 25-26 Unknown
Byte 27-28 Unknown
Byte 29-30 Unknown
Byte 31-32 Size
Byte 33-34 Number of (frames delayed / 2) between hits
Byte 35-36 Number of Hits
Byte 37-38 Starting point and direction of the spell
Byte 39-40 Number of (frames delayed / 2) after animation
Byte 41-42 Unknown
Byte 43-44 Homing
Byte 45-46 Something to do with the velocity of the spell
Byte 47-48 Unknown
Byte 49 Vamps Touch (crashes game when used with dmg spells)
Byte 50 Compression
Byte 51-52 Direction of attack??
Byte 53 Agility (1 = Evade lv.1, 2 = Evade lv.2, 3 = Slow Enemy)
Byte 54 Movement (1 = Lower movement, 2 = Higher Movement)
Byte 55 Restriction
Byte 56 Attack (1 = power staff lv 1 , 2 = pow staff lv. 2)
Byte 57 Confusion
Byte 58 Magic Barrier
Byte 59 Soul Search
Byte 60 Silence
Byte 61 Unknown
Byte 62 Gives 2048 to status effect
Byte 63-64 Unknown
Byte 65-66 Unknown
Byte 67-68 Unknown
I'll give you the starting address of the spells you mentioned.
Confusion starts at 800C0D00 so accuracy will be at 800C0D0C - 800C0D0D
Escape starts at 800C10B8 . Accuracy is at 800C10C4 - 800C10C5
Magic Barrier starts at 800C0DCC . Accuracy is at 800C0DD8 - 800C0DD9
Earth Spikes starts at 800C16D4 . Accuracy is at 800C16E0 - 800C16E1
Maybe soon, I can take a look at the accuracy formula, if I'm fortunate enough to find it.
When you take a look at the mechanics of the game, it will look funny to you because it just seems so arbitrary, the way they decided things.
Edit: I did some testing on accuracy, item drops, movement, RNG, damage, etc, so I updated my faq. It now fully explains the accuracy formula as well as the movement range formula.
I agree with the fact that there are two types of magic attack failures for spells like weakness. Sometimes it fails and gives an MP exp and sometimes it fails and does not give MP exp. I still wonder if escape is dependent on enemy agility levels, as my cast success doesn't feel even close to 55% :(
Does the chance of getting an item go up if you fight 1 hell hound vs 4 hell hounds?
Have you looked into the formula for how spell damage is affected by other elements? It's something like rock damage = N*earth + (N/2)*fire + (N/2)*water + (N/4)*wind. It's probably a little more complicated than that though.
With escape, i gave the enemy 65535 agility, and I was still able to escape. I will have to download mupen to see if the RNG is broken on it, like it is with certain video plugins. With the Jabo's Direct 3D6, the seed advances a little more than once per second, but with RiceVideo, it only advances when you do certain things like opening doors, doing attacks, getting attacked, etc.
I'll admit, theres a small chance I could be wrong about item drops, but when I look at the RNG routine, it always appears to be 1/4 chance from what I've seen. I will need to fight more types of enemies to confirm. The confusing part is that it goes through the RNG process twice. I'm guessing the 2nd time decides which item, if you fight multiple monsters that drop different items. I'll go ahead and do some testing right now, since you asked a good question that I forgot to look into. I guess I'll add the rest of the formulas in my next update.
Edit:
I would have posted my damage formula already, but the formula is ridiculously long and I still don't fully understand the rounding. I understand everything about damage, except for the slight difference in rounding, which I'd like to say is negligible, but in some random case where you hack the game to do extra damage, there's a very small chance you could be off by one.
What I'll do is just make a formula's section I guess.
I wasn't aware of RNG emulation issues. Most of my testing has been on mupen, but most of my experience is on console. In Mupen while in battle I think the RNG cycles if you walk a certain distance or a certain amount of time passes (around a second). I don't trust my memory fully on this one though.
Present the information however you want, I'm just asking questions I'm curious about.
I can post the element bonus here, the other stuff I'm writing in my faq right now, so it should be updated by tomorrow at least.
The way the element bonus works is, it checks the element levels of your other spells. So say you're using a fire spell. Your element bonus would be water /16 rounded down + earth / 8 RD + wind / 8 RD . So basically the opposite element is /16 while the neutral elements are 8. The reason it seems confusing is because the higher your total element count is, the more you hit through defense. That's the reason the damage is buggy when you change your elements during a battle, because that total element variable is only written at the start of a battle, so it won't update since the game assumes your elements never change in battle.
RNG is interesting, it determines the exact spot your avalanche will fall. When i froze the RNG seed, the rocks stacked in 1 spot. It's very hard to manipulate avalanche since it's a multi hitting attack though, without cheating of course. Fortunately for single hitting attacks, it is very easy to manipulate with emulator. I have the formula and know the sequence of the seeding.
What seems to be the furthest progress anyone has made on the game before restarting, seem as though new tidbits are being discovered all the time, and I'm actually quite interested on what a TAS can pull off on this game.
Also I find the Avalanche manipulation, if able to harness, will would make some boss fights a breeze to get through.
I suppose you could position yourself and use math to get the most out of an attack. RNG determines where the rocks will land. Maybe I'll look into that.
The hard part is the fact that you can't really control all 10 attacks because the seed changes at least 3 times after every hit, so the best you can do is make as many rocks hit as possible.
I was wrong about PJ64 effecting the RNG. It's the video plugin I was using that broke it, not the emulator itself.
Joined: 11/13/2006
Posts: 2822
Location: Northern California
Something must have snapped in my brain, because I'm honestly considering taking this on as a future project. Specifically a future project. I've been on a pretty solid streak of TASing one "underappreciated" game that I unironically enjoy a year, and I think this is the perfect candidate for this year.
This is merely an interest post to hopefully start the discussion back up, despite the discussion being 3 years old and outdated. I definitely won't be getting around to this for a while, with a few other low and high profile projects on my plate at the moment, but eventually it will happen. Perhaps I'll begin this Quest in 64 days, who knows.
TASvideos Admin and acting Senior Judge 💙 Currently unable to dedicate a lot of time to the site, taking care of family.
Now infrequently posting on Bluesky
There have been numerous glitches found recently, particularly one that allows you to get like 50 levels (TAS mashing) while dying at the same time as you cast, and battle arena clips and various other methods that can get out of bounds in pretty much any area with the right seam. If you plan to take this on, please get in touch with us: http://www.speedrun.com/quest64