Too easy Psi Attack

[redv]

The problem.
I'm sure, Psi Attack is too easy in OpenXcom.

Formula Psi Attacks in vanilla Xcom:

Attack Strength (AS) = INT(Psi Strength * Psi Skill / 50)        Attacker stats
Defense Strength (DS) = INT(Psi Strength + (Psi Skill / 5))     Defender stats

Attack Success (A%) = 100/56 * (Constant + AS – DS – Distance)
where:   Constant = 25 for Mind Control;   45 for Panic

https://www.ufopaedia.org/index.php?title=Psionics

In one string, formula for Mind Control:

A% = 100/56 * (25 + INT(APSt * APSk / 50) – INT(DPSt + DPSk / 5) – Distance)

If A% < 0 then attack always fail. If A% > 100 then attack always successful.
For example:
My unit: APSt = 90; APSk = 30.
Victim – Sectoid: DPSt = 40; DPSk = 50.
Distance: 20.
Attack Success: A% = 16,07%

Formula Psi Attacks in OpenXcom:

Code: [Select]

double attackStrength = action->actor->getStats()->psiStrength * action->actor->getStats()->psiSkill / 50;
double defenseStrength = victim->getStats()->psiStrength + 30 + (victim->getStats()->psiSkill / 5);
int d = distance(action->actor->getPosition(), action->target);
attackStrength -= d/2;
attackStrength += RNG::generate(0,55);
…
else if (action->type == BA_MINDCONTROL)
{
if (attackStrength > defenseStrength)
{
https://github.com/SupSuper/OpenXcom/blob/master/src/Battlescape/TileEngine.cpp#L1780

APSt * APSk / 50 – Distance / 2 + RNG(0,55)   >   DPSt + 30 + DPSk / 5

Or, after transformation:

A% = 100/55 * (25 + APSt * APSk / 50 – (DPSt + DPSk / 5) – Distance / 2)
No difference between "Mind control" and "Panic" attacks!

Example:
My unit: APSt = 90; APSk = 30.
Victim – Sectoid: DPSt = 40; DPSk = 50.
Distance: 20.
Attack Success: A% = 34,55%

OpenXcom easier to play versus vanilla Xcom.

The suggestion.
1. Suggest to use formula such as in original Xcom. Should be a difference between "Mind control" and "Panic" attacks.
2. Suggest to give less experience points for "Panic" attack, and only if attack is successful. This will reduce the possibility of training exploit.

For example:

Code: [Select]

if (action->type == BA_PANIC)
{
if (attackStrength > defenseStrength)
{
[b]action->actor->addPsiExp();[/b]
int moraleLoss = (110-_save->getTile(action->target)->getUnit()->getStats()->bravery);
if (moraleLoss > 0)
_save->getTile(action->target)->getUnit()->moraleChange(-moraleLoss);
return true;
}
else
{
return false;
}
}
3. Can be added to the formula unit Morale.
For example:

A% = 100/56 * (Const - (Morale / 10) + (APSt * APSk / 50) – (DPSt + DPSk / 5) – Distance)
where: Const = 33 for Mind control;    53 for Panic attack.

My unit: APSt = 90; APSk = 30.
Victim – Sectoid: DPSt = 40; DPSk = 50; Morale = 80.
Distance: 20.
Attack Success: A% = 16,07%

[Volutar]

ufopaedia not always contains correct data.

actual vanilla xcom code is:

Code: [Select]

AttackForce = PsiSkill*PsiStrength/50 - distance/2 +random(0..55);
DefenceForce = TargetPsiStrength+panic?10:0+MC?30:0];
if (TargetPsiSkill>=5) DefenceForce  += Skill/5;

if (AttackForce>DefenceForce)
{
  xcom_psi_exp_count+=2;
  if (panic)
  {
    if (10*Bravery<=Morale) https://bravery0==110
    Morale-=10*Bravery; else Morale=0;
  }
  if (mc)
    MCTargetUnit();
}

In openxcom code there was only 1 mistake - it used +30 for both panic and MC. So actually it is harder to panic than in vanilla xcom.

[redv]

May be you right.
I read the code, because i sense, vanilla Xcom was more difficult.