Thread:Randomouscrap98/@comment-5106053-20140429031049/@comment-9531161-20140429033006

OK, first, the random amount looks fixed. So even if your attack is 1, it could vary by -5 to 5 points. Alternatively, if your attack is 1000, the variance is still only +-5. I don't know if that's what you were going for, but a percentage variation might be better (but it's up to you). Remember, the higher the variance, the more likely it is that the battle's outcome is left to chance. That's why I made the variance so high for the chat RPG: I wanted there to be a higher level of interest in the battle since there's no input. If you're making an RPG like pokemon, I would leave the variance at something low like 5%.

Second, your attack calculation will always be negative as long as the log of defense+1 is 1/9 or greater. Even with a defense of 1, that still makes the attack negative. Is that what you were going for? Like negative amounts to health?

Finally, you should plot graphs for increasing values of attack vs increasing values of defense. Then you should compare that plot to the rate at which health normally increases. If the plot is much higher than the rate of health increase on the attack side, then attack is unbalanced and will overpower high defense players, and vice-versa for a plot lower than the HP increase rate.

Don't worry about how complicated the damage formula is; as long as it works the way you want it to and you understand it, then it's fine. Just make sure one thing doesn't overpower something else (unless you want it like that).