Random chance for status effects and debuffs?

[Misery]

Looking at some discussion about debuffs and the likeliness of inflicting status effects in the class ideas topic, I was reminded I've been hoping for a long time that someone would shed light on the matter. We know that luck affects it, but what about the actual formula? This could be a suggestion for the editor as well, it's likely to be something one would want to change in the game, as status effects are prone to becoming either broken or useless unless you balance them perfectly. The current debuffs wear off quickly and aren't even guaranteed to work, making them little more than a waste of a turn.

[Aile~♥]

Actually, part of the reason debuffs and ailments wear off so quickly is that their duration is ALSO based off the afflicted character's Luck stat.

[Misery]

Oh, I never realized...
Wait, have you verified this? If so, how?
In any case, that's just another unknown formula to me...
Stat debuffs would still be useful if they were just guaranteed to work though, even if they only lasted a single turn.

[Aile~♥]

I don't have any proof, except for the fact that against DD's bosses, random stat debuffs from unleashes only lasted for two turns, while whenever my party's stats get lowered, it lasts for something closer to four or five.

[leaf]

Easiest way to test this would be to make a boss in the editor that spams debuffs/ailments, and have a party with a range of luck values. Something like... 4, 25, 50, and 80 should do. Doesn't have to be exactly those numbers. Mia can hit up to 86 luck iirc.

[Rolina]

I thought it was 90-something?  Still, absurdly high.

[Misery]

I believe the effective cap is 99, I wouldn't restrict myself to what can be gotten from bonuses if I'm just testing...
And while testing would give a rough estimate of the success rates of current spells, it would be an awful lot of guesswork with possibly unknown factors, and wouldn't take me any closer to actually changing them.

[leaf]

Well, all you're testing for is how long the ailments last, not what percent chance they have of affecting the target. The latter would be overly difficult to figure out experimentally.

[Rolina]

To be honest, it probably would be better if we could actually pull the formula from the code.  This will work for guestimates right now, but hard data would be vastly more useful.

[Misery]

Leaf: The code for a spell determines how long it lasts, even if luck is a factor it may not be the only one, and this may differ between various statuses. Figuring out, and more importantly, changing status success rates is my main point of interest.

Role: Bingo. But I don't even know how to find the code, I may or may not be able to disassemble it, but if I even had time for testing right now I'd be doing character art for you guys instead ;P I just posted this since I was reminded of it, and can only hope someone else will look into it. This seems like something you plan to have some influence over in any case.

It may be worth to note that stat buffs are treated the same way as stat debuffs and ailments, and they definitely last for a preset amount of turns, so there's perfectly usable code for 100% success rate stat downs with a little tweaking. But tweaking the existing code for debuffs would be more logical, and is probably simpler as well.

[Aile~♥]

Luff and Rime's effect automatically inflicts a two-turn Psy Seal. The Bind Psynergy, if it actually triggers, gives a Psy Seal that can last anywhere from two turns to four or five.

Debuffs will last a seemingly random amount of time on player characters, but usually only two or three turns on bosses. Stun similarly lasts a random amount of time on players, but typically only two turns on bosses. Amusingly, Sleep seems to last longer than Stun, if the enemy doesn't randomly wake up when hit.

[Rolina]

Well, trends are easy to point out.  But unlike stuff like Psynergy cost to power ratios, this is something we can get hard data for by looking it up in the code.  We can guestimate until the cows come home, but what we really need is hard data.

[zman9000]

So then what we need is some one to get this data. Sure Atrius is who we would ask. but I'm starting to feel that we should have some one else also, at least be able to look at the code and find these little things that we need/want. It wouldn't be some one to work on the editor, but having some one, maby even an active member so they could find this stuff quick.

I my self am working on something that that I need this stuff for. and its a pain not knowing these stats...
I would try to take a look into it, but right now classes are a pain, and the home work is a lot...

[Misery]

Well, it's not like there's a desperate need for this at the moment, just putting it up on the list to make sure it doesn't get overlooked. Atrius isn't the only one, Teawater and Salanewt seem to be into this as well, and there may be someone else I have missed...? As for myself, I understand the general principles of ASM, but I'm not used to GBA assembly, and it seems I'll have a hard time deciphering 32-bit opcodes, at least until I get some experience with actually using said instruction set. Also, If I were told to find a given algorithm, I wouldn't have any idea where to start. But maybe someday I'll become curious enough to start looking into this again...