FUN WITH FORMULAS

The regular attack damage algorithm has a Level^2 in it:

Damage = Battle Power + ((Level * Level * Attack) / 256) * 3 / 2

(Attack has Vigor * 2 that caps at 255 and a bit more)


Compared to the magic damage algorithm:

Damage = Spell Power * 4 + (Level * Magic Power * Spell Power / 32)


So stats matter more with magic attacks than physical ones

I wonder if there is a way to tweak them so that magic and physical attacks are more on par while still keeping the FFVI feel.

A big part of the problem is magic is used in the formula for pretty much anything that isn't the fight command. Even things like Blitzes, Swordtechs, things you'd think would rely on physical stats are still magic driven, so characters who suck at magic get outclassed pretty quickly in raw damage.

The end of the game is still basically a race to get Ultima so you can Gem Box+Quick+Ultima everything into the ground.

Why would you even need to do this

author=Feldschlacht IV
EDIT: And this also applies in the GBA Dragon's Den, which the game actually does have actual challenge, and very interesting enemies and strategies that you have to work around and figure out, while trying to resist the urge to press the 'fuck it nuke everything' button.

So, how does the 'nuke it' option function? Is it like Dragon Quarter where it has plot-related impact or limited use? Or is it just an option they give you that makes things too easy if exploited?

author=hedge1
First, if you're going to have multiple attacks in a round, ensure they're fast and don't waste time. I'd say even the sword slashes from FFT would be too long. Look for animations of 1-2 seconds or less.

You made a very useful comment. I'll take it into account when working on animation. Currently, the game has no animation: the delay between attacks depends on how fast you press the attack button again.

author=hedge1
Second, make sure that the optimal strategy isn't always to move as little as possible to maximize damage. Since 3 attacks is literally 300% damage versus one, it's easy to just sit back and wait for the enemy to move towards you and do one damage, while you counter for three damage next turn.

This is something I noticed while building the maps. Defending your positions is too often the best strategy; except when your cavalry needs to rush the enemy archers. I'm going to introduce an impact bonus when a unit moves to an enemy and deliver its first attack: moving two hexagons towards the enemy then attack would cause 2 damages and remove some Action Point to the target (stunt effect).

author=hedge1
Third, a good rule to follow is the simpler your game, the shorter it should be. So you're probably not looking at a long game. That's probably for the best though, since it gives you a greater chance to finish it and, if it's popular, gives you a reason to make a sequel.

Totally agree! A minimum viable product of my game is already available. And some members made useful comments on it. So I know what improvements to focus on (even before thinking of a sequel).

author=kentona
okay, I found this formula in my notes txt file, with no explanation or source:

(5+slv)*sqrt(a.str/b.vit*(a.atk*x*skill_power)*a.lv/16) - b.def


thoughts? anyone know what this is from?

I have no idea what this is from, but the a and b makes it look like a VX Ace formula, so I'm not sure if it's from a commercial game in the first place.

Anyway, a few analysis. slv is likely skill-level. If this is true, then the skill increases in power with 1/5 for every skill level up if the formula reads skill level 1 as 0 and with 1/6 if it reads skill level 1 as 1.

On the offense side, the game has both str and atk and on the defensive side, it has vit and def. Presumable str and vit is character governed and atk and def is equipment governed and the game did want to keep them separated. str and atk are multiplied with each other, but since both are put into a square root, it works out so that if you have double str and double atk, you deal double damage prior to defense subtraction instead of quadruple.

If slv really is skill level, then a.lv is probably attacker's level. I have no idea what x is, but assuming it usually stays constant, there are within the square root three variables that raises the damage as the characters progress trough the game and one which decreases the damage. This works out as the square root function increasing linearly (except maybe from the very first few levels where the a.lv part may force it to increase much faster early on). Another plausible case is that str and vit stays static, but it would still work out as the square root function increasing linearly.

However, the "5+slv" part makes it look like the damage goes up exponentially. Then the target's def is subtracted from that exponential damage. Typically, if you have attack and defense in your game, you don't design it so that defense has to quadruple or something like that every time attack doubles, but if my guesses are correct, it kind of has to in order to remain relevant.

There are some other seemingly mind boggling details in the formula. Putting the skill_power (which is presumable replaced by a number in the actual formula) inside the square root instead of outside it just causes extra work. Assume you want a single target skill to be one and half a time as strong as a multi target one, you have to make skill power 1,5 * 1,5 = 2,25 times as high instead of just 1,5 times as high. Not that hard, but it still means extra work, especially if you want a number you cannot square in the head.

The middle parenthesis inside the square root has no purpose since there's only * and /. I can only guess that it's there because someone altered a formula where the parenthesis had a purpose in a way so that it no longer has a purpose, but didn't think to remove it. This makes it increasingly hard to find the source.

Anyway, to summarize a bit; It's a skill who's damage doubles after you level it up five or six times. It's from a game which separate equipment governed attack and defense from character governed strength and vitality. The attacker's level also by itself raises damage dealt as it goes up. It's all guesses though.

Technically Zeromus didn't either!

My two guesses is that they didn't feel it was necessary due to the multistage nature of the fight (Neo Exdeath was similar, same HP cap but it wasn't necessary since there were four of him), or it fucked up the scripting and nobody cared too much for a game made for kids in the hellscape that is game development.