Attack/defense formula for game

[zzo38]

This is basically how D&D works..

Roll for Attack, I got a 12
Roll for Defense, they got a 8

I did 4 damage.

I don't know what version that is.

OHRRPGCE has attack minus defense, and then multiplied by elemental damage modifiers.

GURPS has first you roll attack and defender rolls their defense (there are some choices to make on both sides); if the attack is a hit and defense is a miss, then you roll for damage, and then deduct the target's armour and then apply the wounding multiplier, and then a few other stuff if applicable.

[gauauu]

This FAQ explains in-depth how Dragon Warrior 1 worked.

The simple answer is:

There are two formulas for attack damage for enemies. The standard range is
from:

(EnemyStrength - HeroDefense / 2) / 4,

to:

(EnemyStrength - HeroDefense / 2) / 2

[yaros]

I don't know what version that is.

OHRRPGCE has attack minus defense, and then multiplied by elemental damage modifiers.

GURPS has first you roll attack and defender rolls their defense (there are some choices to make on both sides); if the attack is a hit and defense is a miss, then you roll for damage, and then deduct the target's armour and then apply the wounding multiplier, and then a few other stuff if applicable.

My example was loosely based on how Fallout 1-2 calculates the damage. I was trying to present the idea, not the actual system. There are many of them exist.

[nocash]

For example, this calculation:
damage = attack / (100 / (100 + defense));
where all variables are global zeropage variables of type unsigned char, requires 11 scanlines worth of CPU time.

Unsigned char means that your attack & defense values are in range 0..255?
Or is it only smaller range like 0..15? Then you could easily pre-calculate all possible results and store them in a table.

Ehm, how or why do you calculate "100 / (100 + defense)"?
The result would be always zero?
Unless defense were zero, or the unsigned chars were negative?
Or unless your unsigned chars were converted to floating point values? Floating point won't be very fast for sure.
Or did you use fixed point fractions not shown in the formula?

In general, "damage=max(0,attack-defense)" doesn't sound so bad. It's simple, and more or less logical.
If it works depends on what you are doing...
For example, a paper hat might help against rain drops, but not so much against bulldozers.
In that case the formula would be fine. But it might fail if you had other situations in mind.

PS. Just wondering "giving him more initial energy" means that the opponents are all males?
I've recently found a similar gendered sentence in old source code that I had written a few decades ago : )

[tepples]

PS. Just wondering "giving him more initial energy" means that the opponents are all males?
I've recently found a similar gendered sentence in old source code that I had written a few decades ago : )

Use of "he" to mean "he or she" is older usage that may have been more common in previous decades.

[DRW]

Thanks for your tips.

I guess I'll do the simple max(1, (attack - defense)) then since it really seems to be used in a lot of places.
(I use 1 as the default value because enemies who can't be attacked and temporary hero invincibility are a conscious decision, so this shouldn't happen for regular calculations.)

The attack value gets calculated by hero attack value + weapon attack value + item attack value.

Unsigned char means that your attack & defense values are in range 0..255?
Or is it only smaller range like 0..15?

Probably in the higher range, i.e. not just 1-15.

Ehm, how or why do you calculate "100 / (100 + defense)"?

It was a calculation that I found on the internet. I didn't actually use it myself. But yeah, for int values, that's indeed not a good approach.

PS. Just wondering "giving him more initial energy" means that the opponents are all males?

Seriously?