Secret of Evermore (SNES)

TheAngryPanda

Edit history:

TheAngryPanda: 2014-01-04 10:00:17 pm

2014-01-04 06:47:20 pm

Speed > Safety

I don't think underflowing repeatedly would work, but then again, anything seems to be possible with this game and how absolutely (or horrifically) broken it is right now. The issue becomes when you underflow to a massive number, enemies can no longer hit you because they're chances are literally above 100% to miss due to you having an evade% that is way above that as well. So that leaves you with options like the desert in order to die, so you'd have to time the poison/plague/whatever with when you'd die to desert at the proper interval for the stat amount you'd want to achieve... I don't have a formula for this because it's not useful, but I'm sure you (or someone else) could derive one fairly easily.

A small note about Thraxx, I haven't derived a complete formula or description of the ACTUAL damage calculations, and never intend to do so. This is simply because it's faster and more efficient to take what is known about the fight and brute force my way to a good RNG. The damage formula(s) provided to me by various locations all are not compatible with the fight, even with various methods of inclusive and exclusive (then additive/multiplicative) properties thrown in. So that knowledge in hand, I just said "don't give a damn!" and moved right along since I KNOW that I can just use positioning and AI aspects to provide damage needed as opposed to literal frame count and damage outcomes. Currently the max my TAS route can hit Thraxx for is 677, luckily, it's 9x HB due to Magmar not being possible/faster otherwise, however, should something prove otherwise, 8x will take over the 9x slot and save me a few frames. Currently though, if you go 8x, Magmar needs 1 more HB at the end because HB didn't level up during a critical moment (just before heat wave). By it not leveling up, I lose frames due to more text (also lose inputs during text frames, 4 of them) AND Magmar needs to be hit 2 times on top of the rest of the HBs (possible for 1 hit, but still takes longer, surprisingly) which takes longer in total frame contributions than simply casting an extra HB at Thraxx to make up for the lack of HB level pre-Heat Wave. I might be remembering some of this slightly wrong, but I do know that 8x Thraxx is possible but was NOT faster for something along the lines of above... short term memory is tearing me up lately.

If there's anything else regarding Thraxx's damage stuff you wanna know about, lemme know. More people playing with it/looking at it the better.

edit: For those who don't know or are curious, this is basically what's involved with the Thraxx fight:

1) Manipulate good cutscene skip and eventual AI positioning before fight begins
2) Manipulate AI to path correctly
3) Manipulate PC (Player Controlled) character to proper manipulation area (there are 4 locations for this, 2 are preferable... upper left and upper right)
4) Manipulate damage from hardball as AI boy moves into position from previous AI manipulation (this is where it gets horrible)
5) Manipulate damage amplification, while technically a part of the previous manipulation, it is independent and based on AI positioning, simply pressing B or R can change the entire damage outcome
6) Manipulate AI to go where needed shortly after damage is dealt
7) Manipulate PC attack (dog attack) to cage
-7.1) Make sure any manipulation keeps AI in proper location
-7.2) Make sure outcome of damage to cage is max damage on cage
--7.2.1) Make sure heart takes max damage from cage damage, this is actually a thing...
-7.3) Make sure PC is within hitbox of cage
-7.4) Make sure AI is on screen
8) Manipulate 8x Damage, while technically all individualized, you can still affect all of them via other means post-casting since they haven't made contact yet so getting frame perfect casts = better
9) Manipulate PC attack to deal max damage
-9.1) Make sure 8x damage is still high
-9.2) Make sure heart amp damage is decent to high, this is separate from "deal max damage"
10) Manipulate quick cutscene by adjust character slightly over the course of 8x casts travel time, dog movement may not be an option here
-10.1) Make sure Dog is active character post-scene

Anyone running the game regularly don't worry about this, some people asked to see what kind of RNG Thraxx had in Meta's chat and I figured I'd post it here.

Other good examples of RNG in this game are Carltron's fight at the end of the game and how he operates. You can prevent jumping by simply pressing R on the magical frame of happiness. The route test TAS doesn't do this because I stopped caring at that point since I knew the route worked.

Other news:
-BBMSkip needs to be learned by ANYONE running this game. The time it saves is comparable to skipping Defend. So you get to keep your crutch... and save the same amount of time (roughly)? Yeah, that's a damn good deal. It's also easy to do with only one real annoyance, the maggot. Oh no! A reset in the first 5 minutes? Pretty sure that's normal, so no problem there.

I don't think underflowing repeatedly would work, but then again, anything seems to be possible with this game and how absolutely (or horrifically) broken it is right now. The issue becomes when you underflow to a massive number, enemies can no longer hit you because they're chances are literally above 100% to miss due to you having an evade% that is way above that as well. So that leaves you with options like the desert in order to die, so you'd have to time the poison/plague/whatever with when you'd die to desert at the proper interval for the stat amount you'd want to achieve... I don't have a formula for this because it's not useful, but I'm sure you (or someone else) could derive one fairly easily.

A small note about Thraxx, I haven't derived a complete formula or description of the ACTUAL damage calculations, and never intend to do so. This is simply because it's faster and more efficient to take what is known about the fight and brute force my way to a good RNG. The damage formula(s) provided to me by various locations all are not compatible with the fight, even with various methods of inclusive and exclusive (then additive/multiplicative) properties thrown in. So that knowledge in hand, I just said "don't give a damn!" and moved right along since I KNOW that I can just use positioning and AI aspects to provide damage needed as opposed to literal frame count and damage outcomes. Currently the max my TAS route can hit Thraxx for is 677, luckily, it's 9x HB due to Magmar not being possible/faster otherwise, however, should something prove otherwise, 8x will take over the 9x slot and save me a few frames. Currently though, if you go 8x, Magmar needs 1 more HB at the end because HB didn't level up during a critical moment (just before heat wave). By it not leveling up, I lose frames due to more text (also lose inputs during text frames, 4 of them) AND Magmar needs to be hit 2 times on top of the rest of the HBs (possible for 1 hit, but still takes longer, surprisingly) which takes longer in total frame contributions than simply casting an extra HB at Thraxx to make up for the lack of HB level pre-Heat Wave. I might be remembering some of this slightly wrong, but I do know that 8x Thraxx is possible but was NOT faster for something along the lines of above... short term memory is tearing me up lately.

If there's anything else regarding Thraxx's damage stuff you wanna know about, lemme know. More people playing with it/looking at it the better.

edit: For those who don't know or are curious, this is basically what's involved with the Thraxx fight:

1) Manipulate good cutscene skip and eventual AI positioning before fight begins
2) Manipulate AI to path correctly
3) Manipulate PC (Player Controlled) character to proper manipulation area (there are 4 locations for this, 2 are preferable... upper left and upper right)
4) Manipulate damage from hardball as AI boy moves into position from previous AI manipulation (this is where it gets horrible)
5) Manipulate damage amplification, while technically a part of the previous manipulation, it is independent and based on AI positioning, simply pressing B or R can change the entire damage outcome
6) Manipulate AI to go where needed shortly after damage is dealt
7) Manipulate PC attack (dog attack) to cage
-7.1) Make sure any manipulation keeps AI in proper location
-7.2) Make sure outcome of damage to cage is max damage on cage
--7.2.1) Make sure heart takes max damage from cage damage, this is actually a thing...
-7.3) Make sure PC is within hitbox of cage
-7.4) Make sure AI is on screen
8) Manipulate 8x Damage, while technically all individualized, you can still affect all of them via other means post-casting since they haven't made contact yet so getting frame perfect casts = better
9) Manipulate PC attack to deal max damage
-9.1) Make sure 8x damage is still high
-9.2) Make sure heart amp damage is decent to high, this is separate from "deal max damage"
10) Manipulate quick cutscene by adjust character slightly over the course of 8x casts travel time, dog movement may not be an option here
-10.1) Make sure Dog is active character post-scene

Anyone running the game regularly don't worry about this, some people asked to see what kind of RNG Thraxx had in Meta's chat and I figured I'd post it here.

Other good examples of RNG in this game are Carltron's fight at the end of the game and how he operates. You can prevent jumping by simply pressing R on the magical frame of happiness. The route test TAS doesn't do this because I stopped caring at that point since I knew the route worked.

Other news:
-BBMSkip needs to be learned by ANYONE running this game. The time it saves is comparable to skipping Defend. So you get to keep your crutch... and save the same amount of time (roughly)? Yeah, that's a damn good deal. It's also easy to do with only one real annoyance, the maggot. Oh no! A reset in the first 5 minutes? Pretty sure that's normal, so no problem there.

assassin17

2014-01-04 10:05:53 pm

I love YaBB 1G - SP1!

from my "Unsaved stat boosts fix" Readme:

"Thus, using the bug to game your Evade % or Hit Rate % stat isn't feasible unless you're
willing to study Function 8F/BA06, do some math to figure out where the pointer winds up,
and then if the pointer is 8000h to FFFFh, study the relevant ROM addresses in Bank 8F
(if it's 0 to 7FFFh, don't even bother). And then be prepared to do it all over again
whenever your character levels up. :P"

so it's a matter of knowing whether or not the combination of your character level, Speed's level, and the enemy attacker's Hit Rate will produce a pointer that references data that's advantageous. a pain in the arse, because hit rates vary from enemy to enemy.

then as mentioned, there's the matter of things changing when your character(s) level. i'm not sure what you'd do there if the dying underflow bug is already "activated", and the level-up makes it work against you:
- recast the alchemy to get the stat into positive territory temporarily?
- level up the alchemy so it alters the stat more?
- save and reload so the "Unsaved stat boosts" bug will zero out the alchemy-related stat boost?
- try to work around it with manual dexterity (i.e. dodge everything), and hope it resolves when your character levels again?

the second and third options burn valuable time, particularly the second. whatever path you take, it's still a bitch to make Speed underflow work in your favor, due to all the variance.

i'd tend to agree with you that it's not worth it.

from my "Unsaved stat boosts fix" Readme:

"Thus, using the bug to game your Evade % or Hit Rate % stat isn't feasible unless you're
   willing to study Function 8F/BA06, do some math to figure out where the pointer winds up,
   and then if the pointer is 8000h to FFFFh, study the relevant ROM addresses in Bank 8F
   (if it's 0 to 7FFFh, don't even bother).  And then be prepared to do it all over again
   whenever your character levels up. :P"

so it's a matter of knowing whether or not the combination of your character level, Speed's level, and the enemy attacker's Hit Rate will produce a pointer that references data that's advantageous.  a pain in the arse, because hit rates vary from enemy to enemy.

then as mentioned, there's the matter of things changing when your character(s) level.  i'm not sure what you'd do there if the dying underflow bug is already "activated", and the level-up makes it work against you:
- recast the alchemy to get the stat into positive territory temporarily?
- level up the alchemy so it alters the stat more?
- save and reload so the "Unsaved stat boosts" bug will zero out the alchemy-related stat boost?
- try to work around it with manual dexterity (i.e. dodge everything), and hope it resolves when your character levels again?

the second and third options burn valuable time, particularly the second.  whatever path you take, it's still a bitch to make Speed underflow work in your favor, due to all the variance.

i'd tend to agree with you that it's not worth it.

Nitrodon

2014-01-04 11:23:06 pm

Even if you manage to control your underflowed evasion to give good results, you still can't hit anything with underflowed accuracy.

Code:

$8F/BA7C 68          PLA           (pull accuracy from the stack)
$8F/BA7D 10 01       BPL $BA80
$8F/BA7F 7B          TDC           (bound below by 0)
$8F/BA80 C9 64 00    CMP #$0064
$8F/BA83 30 03       BMI $BA88
$8F/BA85 A9 64 00    LDA #$0064    (bound above by 100)

(P.S.: The function technically starts at 8F/B9F8, not 8F/BA06.)

assassin17

Edit history:

assassin17: 2014-01-05 05:29:18 am

2014-01-05 05:25:03 am

I love YaBB 1G - SP1!

Ooh, good catch. I had that flooring noted in my local disassembly, but somehow failed to account for it in the patch Readme.

I almost wonder whether Square added that check later as a result of playtesting, to compensate for the BMI they use there instead of the more sensible BCC. (As well as those employed when first capping the stats in the Boy's and Dog's functions.)

I don't think these instances of inferior branch instructions are ever a problem outside of underflows.

That said, the BCC-->BMI botch is directly consequential (for a little while) when comparing the Dog's current experience to his experience needed for the next level. A mistake that Square curiously did not make for the Boy. (See the "Erratic levelling fix" for that one.)

P.S. Also thanks for the function start info. Is the earlier stuff some invincibility flag?

MetaSigma

2014-01-05 07:52:04 am

Evermore Extrodinare ᕦ(° Д°)ᕤ

Yes...BBM skip should be in instantly with everyone. I had trouble with it at first, but I'm doing it so quickly now it's hilarious. The maggot is the maggot...he's a hero, but it saves roughly 7 to 10 seconds (Defend saves a lil over 15 I've noticed) depending on the skill of the player. It's a risk reward everyone. Trust me it's rough, but it is something we all should be doing. Yes you lose 3 Clay but if you need em that bad grab em at the hut with the 3 clay jar. it's not a tough trick and you'll be cursing more over Hero Maggot than you will the fact you failed the trick.

TheAngryPanda

2014-01-05 08:00:55 am

Speed > Safety

You also save time by it forcing 12 HB on you. Most people run 14 HB initially, thus you gain even more time by switching to the BBMSkip route(s). Just to tag on to what Meta said.

GreenAmbler

2014-01-05 01:15:47 pm

Even if you the Hero Maggot is spawned, you can usually still wedge yourself between it and the adjacent corner, take the hit, then go around it for only a 2-3 second loss. So even with the maggot there you will still likely be gaining time.

Nitrodon

Edit history:

Nitrodon: 2014-01-05 10:40:27 pm

2014-01-05 10:39:57 pm

Assassin, that check is after "lower Hit Rate by 30 vs. non-player-controlled character" is handled (as well as some -10 hit rate effect I never looked into). While your document shows that every enemy with that flag has 50 or 100 hit rate, so hit rate shouldn't go negative anyway, I would still guess this is the reason for that check.

TheAngryPanda

2014-01-06 05:49:33 am

Speed > Safety

Since we're sort of on the same topic as earlier... Maybe you guys can help me with something.

Here's the results of a 6 frame sequence attack post 7x cast from the TAS used to get 500+ damage (it's worked on literally every previous TAS):

-nothing over 450 combined damage (highest was a 37 and 410)
-no maggots ever spawn due to cutscene skip (I think this might be key in the problem)
-assuming movement wasn't a factor, there's essentially a binary situation, where-in you have 0 being no input, and 1 being R or L (both achieve same effect).
-introducing movement only alters things if the AI is altered on the non-player controlled character OR on key frames between certain events (it's best to not manipulate movement here if you don't have to anyways due to causing lag)

So my problem is... somehow there is much less damage being done even if I go and increase damage output from various other means possible through the cutscene skip. Here's some theories:

1) The maggots cause some minor lag in the game from sprite limitations, which allows the "combination" of amp damage to take place and when the maggots are not present, it simply doesn't occur and I cannot achieve a "perfect" combination.

2) I'm just really damn unlucky? This makes no sense given previous attempts at the same goal and the fact that they were met within a 3 binary permutation of the above with only minor adjustments to movement (because it was 1 frame faster). That being said, I'm not one to argue how statistics work, so this is definitely a possibility.

3) Again referring to maggots, perhaps they are key in forcing the AI controlled boy to move in ways that are more beneficial to my goal(s) during the fight. This is the one I'm HOPING is the problem because it means that everything is still possible to be done, where as the other maggot theory implies that the cutscene skip simply cannot be utilized without backtracking and acquiring more clay somewhere, which won't be happening since I'd need 2 or 3 instead of just 1. Even with IFG, that's a no-go.

So anyone have any ideas?

Since we're sort of on the same topic as earlier... Maybe you guys can help me with something.

Here's the results of a 6 frame sequence attack post 7x cast from the TAS used to get 500+ damage (it's worked on literally every previous TAS):

-nothing over 450 combined damage (highest was a 37 and 410)
-no maggots ever spawn due to cutscene skip (I think this might be key in the problem)
-assuming movement wasn't a factor, there's essentially a binary situation, where-in you have 0 being no input, and 1 being R or L (both achieve same effect).
-introducing movement only alters things if the AI is altered on the non-player controlled character OR on key frames between certain events (it's best to not manipulate movement here if you don't have to anyways due to causing lag)

So my problem is... somehow there is much less damage being done even if I go and increase damage output from various other means possible through the cutscene skip. Here's some theories:

1) The maggots cause some minor lag in the game from sprite limitations, which allows the "combination" of amp damage to take place and when the maggots are not present, it simply doesn't occur and I cannot achieve a "perfect" combination.

2) I'm just really damn unlucky? This makes no sense given previous attempts at the same goal and the fact that they were met within a 3 binary permutation of the above with only minor adjustments to movement (because it was 1 frame faster). That being said, I'm not one to argue how statistics work, so this is definitely a possibility.

3) Again referring to maggots, perhaps they are key in forcing the AI controlled boy to move in ways that are more beneficial to my goal(s) during the fight. This is the one I'm HOPING is the problem because it means that everything is still possible to be done, where as the other maggot theory implies that the cutscene skip simply cannot be utilized without backtracking and acquiring more clay somewhere, which won't be happening since I'd need 2 or 3 instead of just 1. Even with IFG, that's a no-go.

So anyone have any ideas?

assassin17

2014-01-06 07:08:52 am

I love YaBB 1G - SP1!

Nitrodon: ah, that does make more sense. the -10 effect is for Ruby Heart.

assassin17

2014-01-06 09:09:44 am

I love YaBB 1G - SP1!

ok, that 8F/B9F8 stuff is the "Do projectiles just pass over target?" check. i would've known that if i'd bothered to check my own Monster Statistics Guide yesterday. so that entry point is probably just used for projectile attacks, while 8F/BA06 is used for normal melee ones.

TheAngryPanda

Edit history:

TheAngryPanda: 2014-01-08 10:10:15 am

TheAngryPanda: 2014-01-06 02:58:23 pm

2014-01-06 02:58:03 pm

Speed > Safety

So if a character is knocked all the way back after skipping the Thraxx cutscene, they trigger the cutscene and become locked in the lower section of the room. This could very easily remove the trick from the TAS, but I'm hoping to find a way around it. GreenAmbler, if you got any ideas, I'd really like to hear them, lol... cuz I'm at a loss here...

Alright, the cutscene skip will not be in the TAS, it simply cannot work with the current route and altering the route in any way ends up costing more time than the trick saves. In some cases this frame loss is 3-5 frames, in others it's well over 2500. I'm going to comb over the rest of Act 1 to see if there's anything I feel like changing, if not, we'll be picking up Act 2 and discussing 100% marathon routes on my stream post-AGDQ.

Sorry I couldn't get it working, kind of disappointed tbh. Sad

GreenAmbler

2014-01-08 03:12:12 pm

Do characters really get knocked all the way back with your setup? I figured you could just let a non controlled player get knocked back or put one far enough to the left or right so they won't hit the trigger, even if it were a bit out of the way it should still save time no? Is this why it doesn't work for the TAS or is it related to damage manipulation?

TheAngryPanda

Edit history:

TheAngryPanda: 2014-01-08 05:09:40 pm

2014-01-08 05:00:46 pm

Speed > Safety

It's related to both. Something about skipping the scene doesn't allow for massive damage to occur when I attack with both the dog and the 7x cast. There is no way to deal enough damage to make the trick worth it without dealing damage with the dog on top of the 2x or 7x casts either, so the dog will fly all the way back regardless of what I do. I cannot switch to boy because then boy dies instantly to Thraxx roar because he's at 0 hp due to zombie boy.

I've spent well over 100 hours on this alone with multiple ideas/suggestions from others and cannot make it work. The most damage I've gotten with the v1.7 method was 440, nowhere close to what's needed. Adjust tactics, I could kill Thraxx, but then the cutscene is played due to the dog hitting the trigger (again, nothing I can do about this) which starts the cutscene up and essentially wastes any time saved. Adjust amount of HB costs anywhere from 3 to 2500 frames, unfortunately the 3 option isn't enough, so the next closest thing is a frame loss of around 450, and that's still not enough. Removing Zombie Boy isn't an option because it's not worth the tons of frame loss I'd get from the inn scene not being skipped and the travel around that area. It really is super disappointing that I can't get this damn thing to work, it's an awesome skip and I've spent tons of time on it now. Sad

Just to clarify, the real-time route could easily handle this with a minor adjustment to the route if it's even consistent enough to be implemented. The TAS, however, absolutely cannot include it under current routing/strategy.

It's related to both. Something about skipping the scene doesn't allow for massive damage to occur when I attack with both the dog and the 7x cast. There is no way to deal enough damage to make the trick worth it without dealing damage with the dog on top of the 2x or 7x casts either, so the dog will fly all the way back regardless of what I do. I cannot switch to boy because then boy dies instantly to Thraxx roar because he's at 0 hp due to zombie boy.

I've spent well over 100 hours on this alone with multiple ideas/suggestions from others and cannot make it work. The most damage I've gotten with the v1.7 method was 440, nowhere close to what's needed. Adjust tactics, I could kill Thraxx, but then the cutscene is played due to the dog hitting the trigger (again, nothing I can do about this) which starts the cutscene up and essentially wastes any time saved. Adjust amount of HB costs anywhere from 3 to 2500 frames, unfortunately the 3 option isn't enough, so the next closest thing is a frame loss of around 450, and that's still not enough. Removing Zombie Boy isn't an option because it's not worth the tons of frame loss I'd get from the inn scene not being skipped and the travel around that area. It really is super disappointing that I can't get this damn thing to work, it's an awesome skip and I've spent tons of time on it now. :(

Just to clarify, the real-time route could easily handle this with a minor adjustment to the route if it's even consistent enough to be implemented. The TAS, however, absolutely cannot include it under current routing/strategy.

TheAngryPanda

2014-01-09 02:54:20 pm

Speed > Safety

Sorry for the double posting, but it's time to talk marathons:

So I need to make a 100% marathon route for/with Metasigma in order for C4L2's SoE run to not be a disastrous outcome due to how much I optimized the route previously. So I need to know what kind of incentives we'd like to have routed into the game for future marathons AND what some people consider to be exhibition glitches or required glitches. This way, we can show off some stuff post-run as well as include what's needed/entertaining in the run.

So, here's what I've gathered thus far:

Included in route:
-All bosses incentive

Exhibition Stuff:
-IFG shenanigans (I'll have to teach people this)
-Messing with AI loops during raptors or maggots @ Thraxx

Blah blah blah blah blah blah, let me know so I can do a general route and have variations as we go (probably in separate colors for easy reading).

assassin17

Edit history:

assassin17: 2014-01-09 03:52:45 pm

2014-01-09 03:41:28 pm

I love YaBB 1G - SP1!

the Boy's and Dog's base Hit Rates are always less than their base Evades (and even with Jade Disk, the former won't exceed the latter). so it's theoretically possible to repeat the stat underflow via dying bug tons of times (like over 1000), and wind up with a Hit Rate that's ~32866 or less, and gets capped at 99 by the character's stat calculation function. alongside an Evade that is a little over 32866, thus left uncapped by said function, and can possibly point to advantageous data when used by the Chance to Hit function.

obviously, this is pointless in a time attack, and in civilized society. Tongue

MetaSigma

2014-01-09 04:30:54 pm

Evermore Extrodinare ᕦ(° Д°)ᕤ

I'm going to try to recreate the glitch in Gomi's Tower where I got the boy stuck in mid air flailing while the dog could run around. That was actually pretty funny. Also yeah all Bosses is pretty doable, if all bosses gets met then that damn snake is going to be the biggest dick of the run since Magmar is going to be a cakewalk due to the upgraded dmg of Hard Ball by that point. Still I personally hope for an all bosses run as it would be nice to in a way show what's actually in the game and at the same time explain what is actually in and out of a true 100% run as well as how it differs from any%. This is something else I've debated...I honestly wanna do a no defend run and take a more...difficult attempt at a run. Like...NOT using perma-aura or Wings glitch on the final boss rush. I still can just aura spam or barrier, but I'd like it to be a little more on edge as it might make a more exciting run overall. What are your thoughts on that Panda?

TheAngryPanda

2014-01-09 04:51:08 pm

Speed > Safety

I think no invincibility would be a more interesting run personally and a great donation incentive, but it does need a semi-high price considering how vastly the danger of losing the run increases. That being said, there is the consideration of potentially losing the run to player fault as opposed to RNG. I don't think all bosses + no invicibility is a good idea though, because optional boss > your face. Running the gambit of hoping barrier/aura doesn't wear off at an appropriate time is a great way to waste what is probably going to be a 2:15 run.

As for All Bosses, we're grinding HB 3 out, there's not really any way out of it that's logical. Then we also need to decide which market route you intend on using, the whirlpool of boredom or actually trying. Considering it's not a real run, staying with the 14 minutes isn't exactly priority #1, thus our shift is towards showcasing knowledge/execution/entertainment. Probably player preference, but personally I believe that not doing the whirlpool would be best. This is especially true if you're under the no invincibility glitch incentive. This being said, I absolutely hate All Bosses as a category/route because of how incredibly and intentionally inefficient it is. Seems like a cop-out to make the game last longer. No Invincibility is a much more interesting incentive, but would require explanation.

I guess my thoughts on 100% no invincibility are that it should be fine if you're not reckless. There's something to be said though about how the WR run of 100% WILL have invincibility because it's faster due to picking up Defend, and intentionally making something slow to be entertaining isn't exactly my idea of a speedrun. But this is a marathon route/run, so I'd imagine that's not nearly as important.

As for literal details of the run, we can polish all that out on skype or something. Practice will iron out anything else that sneaks up on us probably.

I think no invincibility would be a more interesting run personally and a great donation incentive, but it does need a semi-high price considering how vastly the danger of losing the run increases. That being said, there is the consideration of potentially losing the run to player fault as opposed to RNG. I don't think all bosses + no invicibility is a good idea though, because optional boss > your face. Running the gambit of hoping barrier/aura doesn't wear off at an appropriate time is a great way to waste what is probably going to be a 2:15 run.

As for All Bosses, we're grinding HB 3 out, there's not really any way out of it that's logical. Then we also need to decide which market route you intend on using, the whirlpool of boredom or actually trying. Considering it's not a real run, staying with the 14 minutes isn't exactly priority #1, thus our shift is towards showcasing knowledge/execution/entertainment. Probably player preference, but personally I believe that not doing the whirlpool would be best. This is especially true if you're under the no invincibility glitch incentive. This being said, I absolutely hate All Bosses as a category/route because of how incredibly and intentionally inefficient it is. Seems like a cop-out to make the game last longer. No Invincibility is a much more interesting incentive, but would require explanation.

I guess my thoughts on 100% no invincibility are that it should be fine if you're not reckless. There's something to be said though about how the WR run of 100% WILL have invincibility because it's faster due to picking up Defend, and intentionally making something slow to be entertaining isn't exactly my idea of a speedrun. But this is a marathon route/run, so I'd imagine that's not nearly as important.

As for literal details of the run, we can polish all that out on skype or something. Practice will iron out anything else that sneaks up on us probably.

GreenAmbler

Edit history:

GreenAmbler: 2014-01-13 06:41:04 am

2014-01-12 06:41:30 am

If you're looking for glitches to show off, perhaps you could do the Inn Skip so people actually see that you can play the game as Zombie Boy. With a bit of practice it's nearly as fast as the regular route anyway. Other glitches that could be fun to showcase are the Post Aegis shenanigans like having the boy run off screen, and also maybe Forcing Characters OoB. A few others come to mind but they're all frame perfect. I messed around with the Floating Boy/Dog on Gomi's Tower awhile ago but I think it can softlock the game if I remember correctly, although there's probably a way to avoid it.

Edit: The game can softlock when switching during the "Falling" action depending on when you switch. If you fall as the Boy and switch to the Dog right as it starts falling (basically while the sprite is still on the trigger initiating the fall), the action is never carried out and the game softlocks. You can wait until the Dog has fallen past where the trigger is and then switch, locking the dog in place while the boy falls again from the Dog's position. From there, if you want, you have time to switch to the Boy and then his position is locked and the dog falls to the ground. But, if the "locked" character is still in its Falling animation, once you go and trigger the door for the middle tower, the "Falling" character can not finish its action to walk over to the door and the game once again softlocks. If the Falling character is still doing the Falling animation, you'd need to Wings out and do the Tower again before trying to enter the door.

The "Locked" character does not always stay in the Falling animation, I think it has to do with when you first fall as the Boy. If you trigger the floor collapse on the bottom section of the floor, the Boy has to walk up ~2 steps and then fall, and when the Dog starts falling, it will not be in the Falling animation but rather the normal Sitting sprite. You can get a character locked in the air like this and still have them walk directly between the towers if the other character triggers the door. So, you could do it the first way, which is funny because a character is stuck permanently in the Falling animation (but you have to Wings out of tower); or do it the second way without the falling animation but at least you get to see them walk an invisible bridge between the towers. If you do try for this, bring an extra Wings for safety, and make SURE the dog has fallen a bit before switching to him (you could hit X to bring the menu onto the Dog and verify he is not still on the Falling trigger).

If you want Horace to talk to the Dog, you just need to find a audio/visual/time cue for when the Aegis death animation/flashing/shaking is just about finished so the boy is still being knocked backwards when he gets "warped" to the center where he does the Victory pose. If you bring up a menu during the white flash, the flash disappears, just if you want to see what you're doing for reference

If you're looking for glitches to show off, perhaps you could do the Inn Skip so people actually see that you can play the game as Zombie Boy. With a bit of practice it's nearly as fast as the regular route anyway. Other glitches that could be fun to showcase are the Post Aegis shenanigans like having the boy run off screen, and also maybe Forcing Characters OoB. A few others come to mind but they're all frame perfect. I messed around with the Floating Boy/Dog on Gomi's Tower awhile ago but I think it can softlock the game if I remember correctly, although there's probably a way to avoid it.

Edit: The game can softlock when switching during the "Falling" action depending on when you switch. If you fall as the Boy and switch to the Dog right as it starts falling (basically while the sprite is still on the trigger initiating the fall), the action is never carried out and the game softlocks. You can wait until the Dog has fallen past where the trigger is and then switch, locking the dog in place while the boy falls again from the Dog's position. From there, if you want, you have time to switch to the Boy and then his position is locked and the dog falls to the ground. But, if the "locked" character is still in its Falling animation, once you go and trigger the door for the middle tower, the "Falling" character can not finish its action to walk over to the door and the game once again softlocks. If the Falling character is still doing the Falling animation, you'd need to Wings out and do the Tower again before trying to enter the door.

The "Locked" character does not always stay in the Falling animation, I think it has to do with when you first fall as the Boy. If you trigger the floor collapse on the bottom section of the floor, the Boy has to walk up ~2 steps and then fall, and when the Dog starts falling, it will not be in the Falling animation but rather the normal Sitting sprite. You can get a character locked in the air like this and still have them walk directly between the towers if the other character triggers the door. So, you could do it the first way, which is funny because a character is stuck permanently in the Falling animation (but you have to Wings out of tower); or do it the second way without the falling animation but at least you get to see them walk an invisible bridge between the towers. If you do try for this, bring an extra Wings for safety, and make SURE the dog has fallen a bit before switching to him (you could hit X to bring the menu onto the Dog and verify he is not still on the Falling trigger).

If you want Horace to talk to the Dog, you just need to find a audio/visual/time cue for when the Aegis death animation/flashing/shaking is just about finished so the boy is still being knocked backwards when he gets "warped" to the center where he does the Victory pose. If you bring up a menu during the white flash, the flash disappears, just if you want to see what you're doing for reference

Grumly

2014-01-12 04:09:00 pm

Hi everybody,

Using Action Replay, it is possible to fix the amount of damage of any alchemy. For example, if you put the value 64 for 7E358E, the basic strength of your spells will be always 100.

If you cast HB with 7E358E64, HB will always do 100 damage to Tar Skull, 75 to Maggots and 50 to Carniflowers for example, ever.

However, HB will do variable damage to Thraxx even when 7E358E is fixed. Most of the time, I do between 100 and 105 damage per HB and sometimes, up to 140. Using this, it is possible to get rid of one of the RNG, maybe it could be useful in order to discover the way the second one works.

Fun fact, the change in HB damage works only for 1 HB at once. If I cast 2 HB at the same time, I do 120 damage.

Now, I've got a 6 HB strat ;-)

TheAngryPanda

2014-01-12 04:12:54 pm

Speed > Safety

Interesting. I will definitely play around with this. Thank you for posting! You MAY have just saved the TAS.

Grumly

2014-01-12 04:35:16 pm

Hey AngryPanda,

Of course, here is the link :

If I got some time, I will try to find how it works.

TheAngryPanda

Edit history:

TheAngryPanda: 2014-01-12 05:15:24 pm

TheAngryPanda: 2014-01-12 05:01:41 pm

TheAngryPanda: 2014-01-12 04:57:07 pm

2014-01-12 04:54:17 pm

Speed > Safety

~~Do you know if the damage for the 64 value is set at cast or on contact with the target?~~ I'd presume on cast, which means I'd need to identify patterns in which I can obtain said damage amount for each and every cast possible.

Testing shows the actual value is set as you select the target.

Building a results table with various things. Is the game actually capable of setting the value to 64 on it's own?

Interesting result here: Value 11 gives me 20 damage. Value 17 gives 19. Both occur on the exact same frame, the only difference is pressing R. Positional data remained the same as well. This leads me to believe that the Value is not independent of RNG and thus not capable of being manipulated in real-time via setups of any sort. This also means conversely that I can definitely obtain whatever damage amount I want so long as the value range is large enough to reach said max damage. The problem is not having an equation to obtain said max damage value per RNG session. Finding this equation could easily break this game in two TAS-wise, and potentially lead to better routing for real-time.

I think Thraxx has a special formula for damage on his own, so I'm not sure if I'm getting screwed by the Thraxx amp formula on the heart or the actual value itself. Possibly both? Stars will align eventually and it looks like you've prevented me from getting any sleep tonight. <3

[s]Do you know if the damage for the 64 value is set at cast or on contact with the target?[/s] I'd presume on cast, which means I'd need to identify patterns in which I can obtain said damage amount for each and every cast possible.

Testing shows the actual value is set as you select the target.

Building a results table with various things. Is the game actually capable of setting the value to 64 on it's own?

Interesting result here: Value 11 gives me 20 damage. Value 17 gives 19. Both occur on the exact same frame, the only difference is pressing R. Positional data remained the same as well. This leads me to believe that the Value is not independent of RNG and thus not capable of being manipulated in real-time via setups of any sort. This also means conversely that I can definitely obtain whatever damage amount I want so long as the value range is large enough to reach said max damage. The problem is not having an equation to obtain said max damage value per RNG session. Finding this equation could easily break this game in two TAS-wise, and potentially lead to better routing for real-time.

I think Thraxx has a special formula for damage on his own, so I'm not sure if I'm getting screwed by the Thraxx amp formula on the heart or the actual value itself. Possibly both? Stars will align eventually and it looks like you've prevented me from getting any sleep tonight. <3

Grumly

Edit history:

Grumly: 2014-01-12 05:19:34 pm

2014-01-12 05:16:33 pm

The value of 7E358E is set at cast. When the spell hit the enemy, it take the value of the damage until a new spell is launch.

i never noticed before that if you launch a spell on multiple target of the same baddies, damage will be the same on all enemies.

Furthermore, since I observed 75 damage on maggots and 100 damage on tar skull when the value of 7E358E is fixed, whatever alchemy you cast against those enemies, damage on Tar Skull will always be 133% of maggot's (for example 11 and 8 because of rounding up). The same thing for skelesnail and winpy flower for example (21 vs. 10 HP).

Just see your post Panda, sorry about your sleep Wink

TheAngryPanda

Edit history:

TheAngryPanda: 2014-01-12 05:27:57 pm

TheAngryPanda: 2014-01-12 05:20:40 pm

TheAngryPanda: 2014-01-12 05:20:30 pm

2014-01-12 05:19:57 pm

Speed > Safety

So it's essentially a "critical" hit, ie; max range damage amount for the spell. I still need to confirm that the higher values can be set by the game itself instead of via ram poking or whatever. I will figure this out and you may have found a holy grail for TAS'ing this game.

also, I have a nasty habit of editing my posts constantly; lol.

edit: If that's the case, as it's calculated subsequently per each new cast, perhaps that means we could re-roll our value during multi-casts until we encounter a promising number. I know earlier you said the value was permanently set on 64, but if it reads 64 each time, perhaps it met with some complication? Especially during Thraxx damage formulas.