View previous topic :: View next topic |
Author |
Message |
Newbie_Power

Joined: 04 Sep 2006 Posts: 1762
|
Posted: Wed Sep 06, 2006 6:18 pm Post subject: Some questions about Attack accuracy. |
|
|
I have been wondering what the tilde symbol does mathematically when looking at the formula "Aim * 4 ~ Evade".
It may sound unimportant, but I would like to know so I can find it easier to adjust accuracy and dodging accordingly to my tastes.
On Extra Damage... Does the extra damage percent factor before "Attack - Defense *.5" or after that formula is calculated? I was hoping to be able to multiplay the magic attack stat itself by some percentage before calculating in defense. |
|
Back to top |
|
 |
JSH357

Joined: 02 Feb 2003 Posts: 1705
|
Posted: Wed Sep 06, 2006 6:37 pm Post subject: |
|
|
I'm not TOTALLY sure about the damage formula, but:
In accuracy, the tilde works as such:
1) Take the aim value * 4 and compare it to evade.
2) A random number is taken between 0/1 and each value.
3) The higher number wins. If there's a tie, I'm not sure what happens.
I THINK attack - defense happens before multipliers.
Last edited by JSH357 on Wed Sep 06, 2006 7:05 pm; edited 1 time in total |
|
Back to top |
|
 |
Newbie_Power

Joined: 04 Sep 2006 Posts: 1762
|
Posted: Wed Sep 06, 2006 6:58 pm Post subject: |
|
|
I am not quite sure I grasp that concept completely.
Do I need to compare (Aim * 4) to (Evade * 4), or the actual Aim value with Evade * 4?
If the latter, wouldn't comparing Evade to Aim * 4 make more since since Evade would be the lower stat in this case?
And I think I can understand the rest. Thank you. |
|
Back to top |
|
 |
JSH357

Joined: 02 Feb 2003 Posts: 1705
|
Posted: Wed Sep 06, 2006 7:04 pm Post subject: |
|
|
It's (Aim*4) vs. (Evade)
And I don't understand why it was designed that way either. Makes it really hard to balance. |
|
Back to top |
|
 |
Newbie_Power

Joined: 04 Sep 2006 Posts: 1762
|
Posted: Wed Sep 06, 2006 7:13 pm Post subject: |
|
|
I see the edit in your post...
And yes, I am already tasting the hard to balance. Say someone's accuracy is 10, and the enemy's evade is 5. That would be a random number between 40 and 5, and then compare that random number to 0-1, in which the higher number wins between those two?
I think I am missing something. |
|
Back to top |
|
 |
JSH357

Joined: 02 Feb 2003 Posts: 1705
|
Posted: Wed Sep 06, 2006 7:16 pm Post subject: |
|
|
Ohhh, wait, I think I confused you. When I said 0/1 i meant 0 or 1. (I'm not sure which) The comparism is something like this:
randomnumber(0,Acc*4) ~ randomnumber(0,Evd)
But you are right about balancing woes. |
|
Back to top |
|
 |
Valigarmander Bye-Bye

Joined: 04 Mar 2006 Posts: 750 Location: Nowhere
|
Posted: Wed Sep 06, 2006 7:20 pm Post subject: |
|
|
In mathematics, the tilde, often pronounced "twiddle," is often used to denote an equivalence relation between two objects. Thus x ~ y means x "is equivalent to" y (Note that this is quite different from stating that x equals y). Especially, it can be used to denote the asymptotical equality of two functions. For example, f(x) ~ g(x), means that limx→∞ f(x)/g(x) = 1. In English it is often used to represent approximation, for example ~10 would mean "approximately 10". Similar symbols are used in mathematics, such as in π ≈ 3.14, "π is about equal to 3.14". Since the double-tilde (≈) is not available from the keyboard except on the Macintosh (where it is Option-x), the tilde (~) became a substitute for use in typed entry. There is also a triple-tilde (≋), which is used to show congruence. In mathematical logic, ~ is used to represent negation. ~p means "not p". See the logic section above. In statistics and probability theory, ~ means "is distributed as". |
|
Back to top |
|
 |
Newbie_Power

Joined: 04 Sep 2006 Posts: 1762
|
Posted: Wed Sep 06, 2006 7:23 pm Post subject: |
|
|
That makes MUCH more sense now. (EDIT: I typed that before I saw the history of using Tilde in mathematical formulas, obviously).
I am actually deciding to use Aim * 2 compared to Evade for most attacks, with Aim * 4 being for very accurate attacks.
Last edited by Newbie_Power on Wed Sep 06, 2006 7:24 pm; edited 1 time in total |
|
Back to top |
|
 |
Valigarmander Bye-Bye

Joined: 04 Mar 2006 Posts: 750 Location: Nowhere
|
Posted: Wed Sep 06, 2006 7:24 pm Post subject: |
|
|
I got all that crap from Wikipedia. I have no idea what it means. |
|
Back to top |
|
 |
Newbie_Power

Joined: 04 Sep 2006 Posts: 1762
|
Posted: Wed Sep 06, 2006 7:25 pm Post subject: |
|
|
Heh. I thought you were out to truly pwn me for a second there. |
|
Back to top |
|
 |
Valigarmander Bye-Bye

Joined: 04 Mar 2006 Posts: 750 Location: Nowhere
|
Posted: Wed Sep 06, 2006 7:30 pm Post subject: |
|
|
Nah. To do that I'd need full access to your belly button. But enough about me. How do for you, Kemo-Sabi? |
|
Back to top |
|
 |
Newbie_Power

Joined: 04 Sep 2006 Posts: 1762
|
Posted: Wed Sep 06, 2006 7:49 pm Post subject: |
|
|
Err...
Anyway, I have an easier time thinking about accuracy now, thanks to JSH. |
|
Back to top |
|
 |
Moogle1 Scourge of the Seas Halloween 2006 Creativity Winner


Joined: 15 Jul 2004 Posts: 3377 Location: Seattle, WA
|
Posted: Wed Sep 06, 2006 8:29 pm Post subject: |
|
|
Val, that is all very fascinating and has nothing to do with the tilde in the aim formula. _________________
|
|
Back to top |
|
 |
msw188
Joined: 02 Jul 2003 Posts: 1041
|
Posted: Wed Sep 06, 2006 9:01 pm Post subject: |
|
|
I am nearly certain that extra percentage is calculated AFTER the defense is subtracted. However, it is fairly easy to simulate the behavior you want. Let's say that you would like your ATK stat to be increased by x percent (if decrease, then x is a negative number). You will have to make a chain of attacks. That is to say, you will have multiple attacks that connect to each other.
The first attack will most likely be an attack that does no damage, but stores the target (thus the player can choose who that attack will end up damaging).
The second attack will then target the 'caster', and will be set to damage the caster's ATK stat as "ATK=(100+x)% of current". To do this, you choose the % of current option for damage calculation, and then use the extra percentage damage for the x.
The third attack will then target the stored target. It can be set to regular damage with no extra percentage. This should also have the bitset on "clear stored target", so that this target is not chosen next time this attack is used.
The fourth attack should target the caster. This should be the attack to restore the ATK stat to normal. This is more difficult. You will want to use the % of current option again, but this time the extra damage should be calculated by the following formula:
extra damage= (-100*x)/(100+x)
Thus, if x was positive (gained ATK), this will be negative, and vice versa. If you would like an explanation of where this formula came from, feel free to ask.
Note that due to rounding errors, this may gradually change your character's ATK stat as the battle continues (all stats but HP and MP are reset after a battle is concluded). If this attack is the ONLY thing that has a chance of affecting the given ATK stat, then you can replace the fourth attack by an attack that is set for damage as "ATK=100% of maximum".
All four of those attacks should chain to the next one in line with a 100% chance of chaining. I hope that this was understandable. Any questions please feel free to ask. |
|
Back to top |
|
 |
Newbie_Power

Joined: 04 Sep 2006 Posts: 1762
|
Posted: Thu Sep 07, 2006 10:00 am Post subject: |
|
|
It's not that big of a deal, really. I would rather balance an enemy's balance defense based on something I can do much easier and in less time. |
|
Back to top |
|
 |
|