øAslickproductions.org/forum/index.php?PHPSESSID=5f0fck550j2m4m2fpbtkj2vkm1&topic=817.msg9276e:/My Web Sites/Slick Productions - FFIV Message Board/slickproductions.org/forum/indexdadd.htmldelayedslickproductions.org/forum/index.php?PHPSESSID=5f0fck550j2m4m2fpbtkj2vkm1&topic=817.60e:/My Web Sites/Slick Productions - FFIV Message Board/slickproductions.org/forum/indexdadd.html.zx7Rg^ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÈ…¬$OKtext/htmlISO-8859-1gzip8:Ö$ÿÿÿÿÿÿÿÿÑHTue, 10 Mar 2020 08:55:42 GMT0ó°° ®0®P®€§²ð®6Rg^ÿÿÿÿÿÿÿÿß+$ FF4 Algo - Input Requested

Author Topic: FF4 Algo - Input Requested  (Read 30808 times)

Deathlike2

  • Moderator
  • *
  • Posts: 3,538
  • I'm looking at you, bitch!
    • View Profile
Re: FF4 Algo - Input Requested
« Reply #60 on: January 06, 2009, 06:43:57 PM »
Consistency question - do  you have any issue with me renaming / reordering things for consistency?

It depends, I'm sure I get off tangent (as evidenced by all the random-ish posting on thoughts).

Quote
For example, when referring to bows and arrows, in one place you refer to the setup as "arrow in primary hand" whereas later it's "bow in recessive hand."

The check for Bow+Arrow attack power is dependent on where the Bow is placed, but that was late in the findings (past the point when I started writing said document).

Quote
Also, the order in which you present the formulae for standard/unarmed/edge/yang/bow&arrow/monsters differs by section.

Yea, I'll have to look at that... when I started writing the sections, I didn't actually put an order together. Unarmed should go first, then standard, then Bow+Arrow, then character specific (Yang first, and then Edge). Monsters should always be last.

The last part is addressed... because there should be a logical order of sorts... although I wonder if I should Bow+Arrow ahead of the standard...
Working on the next Yet To Be Named FF4 "Hardtype" Hack Download Latest: v1.48

Video Demos: #1 #2 #3

odditude

  • Vargas
  • *
  • Posts: 91
    • View Profile
Re: FF4 Algo - Input Requested
« Reply #61 on: January 06, 2009, 07:03:40 PM »
i'd say standard before special cases, personally.

Dragonsbrethren

  • Forum Overlord
  • *
  • Posts: 1,820
    • View Profile
    • Dragonsbrethren Industries
Re: FF4 Algo - Input Requested
« Reply #62 on: January 07, 2009, 06:27:24 AM »
i'd say standard before special cases, personally.

Yeah, me too.

odditude

  • Vargas
  • *
  • Posts: 91
    • View Profile
Re: FF4 Algo - Input Requested
« Reply #63 on: January 07, 2009, 03:23:50 PM »
The following special cases are missing from the Critical Hit modifier section:
- bow OR arrow equipped
- Edge w/ one weapon, other hand empty

Also, according to the algorithm, the spell multiplier does not affect spell damage at all, only hit rate.  Is this correct, or an oversight in the formula?

"There is a bug where monsters count the dead in the algorithm, found in one rare instance. This only applies to the SNES version of the game."
What is the rare instance?

"Cure4/Curaja - When an enemy target takes more than 16383 damage, this spell will heal the MP of that target instead due to overflow."
What happens if the MP healing overflows (say, if over 34000 damage has already been dealt)?
« Last Edit: January 07, 2009, 03:57:28 PM by odditude »

Deathlike2

  • Moderator
  • *
  • Posts: 3,538
  • I'm looking at you, bitch!
    • View Profile
Re: FF4 Algo - Input Requested
« Reply #64 on: January 07, 2009, 06:50:55 PM »
The following special cases are missing from the Critical Hit modifier section:
- bow OR arrow equipped

I believe it follows the fists critical hit formula.

Quote
- Edge w/ one weapon, other hand empty

It follows the regular critical hit formula like all characters... half the weapon's attack power.

Quote
Also, according to the algorithm, the spell multiplier does not affect spell damage at all, only hit rate.  Is this correct, or an oversight in the formula?

The spell multiplier certainly does affect the spell damage (it assumes the spell is deals damage of course)... however if the spell power of a spell is 0 (Venom/Poison and the Cockatrice summon's attack), it's going to deal 1 damage regardless.


Quote
"There is a bug where monsters count the dead in the algorithm, found in one rare instance. This only applies to the SNES version of the game."
What is the rare instance?

If I hadn't mentioned it in the bugs section, the tallest Magus Sister (from the Tower of Zot) casts Cure2/Cura to every living sister (the middle sister is dead to perform this). It happens that factors in the dead... which reduces the overall healing dealt.

Quote
"Cure4/Curaja - When an enemy target takes more than 16383 damage, this spell will heal the MP of that target instead due to overflow."
What happens if the MP healing overflows (say, if over 34000 damage has already been dealt)?

I'd rather not find out, I'm sure someone will come along to answer that (or I perform a test of that)...
Working on the next Yet To Be Named FF4 "Hardtype" Hack Download Latest: v1.48

Video Demos: #1 #2 #3

Dragonsbrethren

  • Forum Overlord
  • *
  • Posts: 1,820
    • View Profile
    • Dragonsbrethren Industries
Re: FF4 Algo - Input Requested
« Reply #65 on: September 30, 2012, 12:14:07 AM »
You interested in covering the PlayStation version? I've been playing through it for Tomato and decided to check out some of the bugs.

Back row bit removal: Not fixed
Spell timers + Stop: Only tested with Sleep so far. No crash on monsters; still extends status indefinitely.
Item duplication: Back after being fixed for Easy Type. (I'd hazard a guess that it was intentionally left unfixed, but who knows)

I didn't take notice to whether arrows are still infinite or not but I'm pretty sure they are, from playing this version in the past. I'll confirm the next time I play. I'll also test critical hits.


New bugs:

Valvalis's spin form doesn't seem to get its defense/evasion boost. This confirms what Ian Kelly claimed in his differences guide, shockingly, given how innaccurate it is about other things. (It still claims you can't hit her at all in the SNES version, which simply isn't true). I dumped her attack sequence in that version and it is identical to the others, so it's not that that was changed. I didn't take notice to Cagnazzo's shell, unfortunately, and I think I've already overwritten my saves from that far back.
« Last Edit: September 30, 2012, 12:22:03 AM by Dragonsbrethren »

Deathlike2

  • Moderator
  • *
  • Posts: 3,538
  • I'm looking at you, bitch!
    • View Profile
Re: FF4 Algo - Input Requested
« Reply #66 on: September 30, 2012, 10:59:47 PM »
The only time you can't hit Valvalis/Barbariccia is when your hits accuracy is far below the evasion data for the boss. I had a fond memory of trying to attack her with Jump and the Blood/Drain Spear. Despite the hit being a miss more often than not, the Jump aspect still causes her to counter/react. The only other thing probably involves an underleveled Cecil... but meh.

I've always felt the FF Chronicles versions were primarily the ROM+some emulation-ish overlays

PS is not my forte.. but it's good to know nonetheless.
Working on the next Yet To Be Named FF4 "Hardtype" Hack Download Latest: v1.48

Video Demos: #1 #2 #3

Dragonsbrethren

  • Forum Overlord
  • *
  • Posts: 1,820
    • View Profile
    • Dragonsbrethren Industries
Re: FF4 Algo - Input Requested
« Reply #67 on: October 01, 2012, 02:21:29 AM »
Yeah, the Blood Spear is the only thing that missed when I was fighting her, and I knew why that was the case. I forgot to test if her magic evasion gets applied, but luckily I did keep a save from then, so I'll get around to that the next time I play.

In general, I'd say you'd be completely safe saying your guide covers the PSX version. It's a straight port of the SNES version, and now it's just a matter of figuring out if any of the other non-critical bugs were fixed (doubtful) or if they broke anything else.

Grimoire LD

  • FF4 Hacker
  • *
  • Posts: 1,684
    • View Profile
Re: FF4 Algo - Input Requested
« Reply #68 on: October 01, 2012, 07:27:05 AM »
Does the Stoneblade still freeze the game?

Deathlike2

  • Moderator
  • *
  • Posts: 3,538
  • I'm looking at you, bitch!
    • View Profile
Re: FF4 Algo - Input Requested
« Reply #69 on: October 01, 2012, 09:13:17 AM »
Yeah, the Blood Spear is the only thing that missed when I was fighting her, and I knew why that was the case. I forgot to test if her magic evasion gets applied, but luckily I did keep a save from then, so I'll get around to that the next time I play.

It's either the fact that 255 really means invulnerability or it's the evasion/magic evasion multiplier (99x 99%) that causes it. I haven't actually thoroughly tested that.

I know while in tornado form, she is definitely immune to Slow since her magic defense (along with defense). You can cast Slow on her when she's not in tornado form.
Working on the next Yet To Be Named FF4 "Hardtype" Hack Download Latest: v1.48

Video Demos: #1 #2 #3

Dragonsbrethren

  • Forum Overlord
  • *
  • Posts: 1,820
    • View Profile
    • Dragonsbrethren Industries
Re: FF4 Algo - Input Requested
« Reply #70 on: October 06, 2012, 04:55:32 PM »
I tested the Valvalis battle again, and she definitely got the defense bonus this time. So that means either there's a bug that only strikes at random, or I'm an idiot. I'd go with the latter, personally, although I swear I was doing full damage against her in tornado form the first time I played this battle. I checked the screenshots I grabbed, but none of those show the damage I was doing. The magic defense is handled the same way as the original, too.

I guess I'll try it a few more times just so I'm 100% certain.

I forgot to mention it, but I confirmed that arrows aren't fixed. No idea on the Stone/Gorgon sword, I'd probably have to cheat that into the inventory if I want to test it.

Deathlike2

  • Moderator
  • *
  • Posts: 3,538
  • I'm looking at you, bitch!
    • View Profile
Re: FF4 Algo - Input Requested
« Reply #71 on: October 11, 2012, 01:48:04 PM »
The only thing I have to say in regards to the algo, is that there's no special algo for Double Meteor...

Apparently there is a range cap...

When Spell Power > 510, the range is capped @ 255 (which is always 1/2 of 510). This is also why Meteor (and subsequently Double Meteor) has a very weak range of damage. Where it was originally thought that Meteor would have a range of 400 (800 is the Spell Power), the math doesn't seem to agree with that.

That is something to look at/into...
Working on the next Yet To Be Named FF4 "Hardtype" Hack Download Latest: v1.48

Video Demos: #1 #2 #3

Dragonsbrethren

  • Forum Overlord
  • *
  • Posts: 1,820
    • View Profile
    • Dragonsbrethren Industries
Re: FF4 Algo - Input Requested
« Reply #72 on: October 20, 2012, 10:10:09 PM »
The bug where Kain repeats item/magic use with the Avenger in the Zeromus battle is fixed in the PSX version, like Easy Type. The other Avenger bugs remain (you'd really think they would've caught the bug with berserk Cecil first).

I haven't decided yet, but I may take a look at the WSC version. PSX version just seems to be Easy Type's fixes with FF4's data and the item duplication bug restored for some reason.

Dragonsbrethren

  • Forum Overlord
  • *
  • Posts: 1,820
    • View Profile
    • Dragonsbrethren Industries
Re: FF4 Algo - Input Requested
« Reply #73 on: October 21, 2012, 10:41:42 AM »
Decided to give the WSC version a go. Not sure if I'll play through the whole thing or not, but I've already made a pretty interesting discovery: Enemy rows are still functional in this version. Seems they broke during the GBA port, not this remake. The targeting works the same way as the GBA version, so sometimes the enemy it starts you on is actually in the back row (three Dive Eagle formation near Baron, for example).

Thinking about throwing together a little site dedicated to documenting this version. "The Forgotten IV" or something. It's closer to the SNES version in terms of presentation than I remembered.

 :edit:

Tellah missing doesn't progress the Bard battle, same as the first GBA version.


Confirmed that dark knight Cecil auto covers critical allies.

Gilbert/Edward doesn't seem to auto hide. (Nevermind, just saw him do it.)

Gradual petrify actually works in this version, and seems to go much faster than in the SNES version. In addition to the light gray squiggles characters get in the GBA version, the character's sprite gets progressively grayer as they petrify. This isn't like the SNES version where they turn to stone from bottom up, they start with a dark gray tint and it gets more pronounced at the next level of petrification. Only certain colors of the sprite change, seems like they were going for skin and hair on Rydia, but Cecil's armor changes.

When I healed Cecil of stone status, the shades of yellow on his boots never updated and stayed gray. Stone bug affects Edward too, probably everyone. (Maybe this is an emulation issue? No real way of telling, but seems almost too obvious to overlook.)

Double turns seem to be present. Wasn't sure at first, there's no ATB bar to make it apparent in this version and it's not uncommon for Cecil to get more turns than anyone else in the beginning, but I've had the command window pop up immediately after attacking with Cecil and Edward, same as the bugged GBA versions.

There's an annoying feature in the field equip menu where the item menu stays open even after you've equipped something. Additionally, trying to equip more of the same arrows will just swap the quantities instead of adding them.

I forgot to mention it, but elemental rods are single target, like the GBA version. Already knew about that, so I completely neglected to post it.

Bows and arrows seem to work correctly from both rows.

It doesn't seem possible to make a character permanently long range.

Yang's Power command seems to "charge" a lot faster in this.
« Last Edit: October 21, 2012, 12:38:30 PM by Dragonsbrethren »

Grimoire LD

  • FF4 Hacker
  • *
  • Posts: 1,684
    • View Profile
Re: FF4 Algo - Input Requested
« Reply #74 on: October 21, 2012, 02:19:59 PM »
Those are some really weird changes. Seriously, Dark Knight Cecil using Auto-Cover? That's just such a strange function to include to begin with. Does that mean that Paladin Cecil no longer Auto-covers or is this a shift in data so they both can do so?