Strength determines spell damage

[Fang]

Hello!
Just like the title said, Is it possible to do that (without using JASS)?

Oh yeah... And how can I make a spell folder? Thanks

 

[Warseeker]

• Storm bolt
  • Events
    • Unit - A unit Starts the effect of an ability
  • Conditions
    • (Ability being cast) Equal to Storm bolt
  • Actions
    • Unit - Cause (Triggering unit) to damage (Target unit of ability being cast), dealing (Real((Strength of (Triggering unit) (Include bonuses)))) damage of attack type Spells and damage type Normal

Something like this?

Oh yeah... And how can I make a spell folder? Thanks

Use a Spellbook?

 

[Rahko]

• Spell Damage Scaling
  • Events
    • Unit - A unit Starts the effect of an ability
  • Conditions
  • Actions
    • If (All Conditions are True) then do (Then Actions) else do (Else Actions)
      • If - Conditions
        • (Ability being cast) Equal to Storm Bolt
      • Then - Actions
        • Ability - Set Ability: (Unit: (Triggering unit)'s Ability with Ability Code: Storm Bolt)'s Real Level Field: Damage ('Htb1') of Level: 0 to ((Real((Strength of (Triggering unit) (Include bonuses)))) x 3.00)
      • Else - Actions

Do note, that the level is 0-indexed, meaning that level 0 in the trigger is level 1 in the ability and level 1 in the trigger is level 2 in the ability, etc.

And what do you mean by a "spell folder"?

 

[Fang]

• Storm bolt
  • Events
    • Unit - A unit Starts the effect of an ability
  • Conditions
    • (Ability being cast) Equal to Storm bolt
  • Actions
    • Unit - Cause (Triggering unit) to damage (Target unit of ability being cast), dealing (Real((Strength of (Triggering unit) (Include bonuses)))) damage of attack type Spells and damage type Normal

oops sorry, I meant the attack damage determines the spell damage.

@Sabe spellbook

 

[Pyrogasm]

Same concept, just use compute the unit's attack damage range by querying its base damage, number of dice, and sides per die. MAX = Base + (Num Sides)*(Num Dice). You can get all these values with triggers now, just look for it. There's no way to get 'green' damage numbers.

 

[Dr Super Good]

Since attack damage depends on randomness, one would need to simulate the dice rolls using a loop such as for integer a in range. This is to accurately simulate the random probability distribution without getting into complicated mathematics.

Using the obvious random integer in range function between the minimum and maximum damage is only correct for attacks with a single dice. This is due to the sum outcome of multiple dice rolls having a gaussian probability distribution.

 

[kentzeus]

oops sorry, I meant the attack damage determines the spell damage.

• Set real = (Real((Base Damage of unit for weapon index 1)))
• For each (Integer A) from 1 to (Dice Number of unit for weapon index 1), do (Actions)
  • Loop - Actions
    • Set real = (real + (Random real number between 1.00 and (Real((Dice Sides of unit for weapon index 1)))))
• Else - Actions

This gets you the correct attack damage

 

[Cespie]

...without getting into complicated mathematics.

This is due to the sum outcome of multiple dice rolls having a gaussian probability distribution.

YOU PROMISED NOT TO!

Jokes aside... in laymans terms (i.e. my half-assed understanding), what this means is that multiple dice rolls will end up closer to the average damage than fewer rolls as the distribution starts to even out. Kinda like if you were to roll a dice 6 million times vs 6 times, where certain outcomes would in all likelihood not even present themselves at 6 rolls, but where the distribution would start looking relatively even at 6 million.... or even 600k. Taking stuff like that into consideration may be a bit overkill for simulating a handful of dice rolls though.
You can also opt for the far lighter method of simply calculating the average damage, but that removes the randomness completely. Personally, I don't mind that, but to each their own. If the gap between min and max damage is relatively small, then I'd recommend just calculating the average.

 

[Dr Super Good]

YOU PROMISED NOT TO!

I did not get into the complicated maths. In theory one could correctly calculate a random damage number with linear complexity with respect to die count by getting a constant probability random number in a range, then transforming it based on the probability distribution and finally rounding. There may be precision issues when very large numbers are involved but the overall damage characteristics would be a lot more accurate than a constant probability in range approach or just taking the average.

This is similar to how one calculates a random point within a circle. The random radius value has to be transformed such that it is more likely to be a higher number than a low one.

 

[Cespie]

I did not get into the complicated maths.

It was meant as a joke.