Critical Hits

When a unit in combat is hit, it runs a risk of suffering a devastating critical hit, also known as an armor-defeating hit in the combat log. In short, the risk to take a critical hit is normally 1/36, or 3/36 when above 50 fatigue, or 6/36 when immobilised. A critical hit reduces protection by 25%.

Note that critical hits are a separate mechanic from head hits.

On a hit, first armor-piercing effects (AP weapon, AN weapon, piercing weapon) are applied to the targeted unit's protection. AP and piercing are done separately, so they are multiplicative (60% armor reduced when stacked).

• A protection DRN (exploding 2d6) is rolled.
• If the target has cursed luck, it has a 50% chance of taking -1d6 (open roll) to its protection DRN.
• If the target has fatigue >= 50, protection DRN is reduced by 1.
• If the DRN is negative, it is set to 0.
  • If the unit is immobilized, e.g. entangled or fatigue >= 100, but was not already immobile (e.g. not a Monolith), then it is a critical hit if the protection DRN is 4 or less.
  • Otherwise, it is a critical hit if the protection DRN is 2 or less.

Note that this differs substantially from the manual's description. In particular, Fatigue does not apply a "-1 per 15" scaling reduction, only a small bump at 50 and a larger one at 100.

When a critical hit occurs, the unit's current protection (after AP stuff above) is reduced by 25% (rounded down, all divisions are btw). Afterward, the protection DRN is added to that protection to get the protection used against damage.

The manual says critical hits reduce prot by 25%, but the code below appears to show a 50% reduction to base prot. Since the manual is wrong on this topic in several ways, can we confirm the code is more accurate?

Note that it is the same DRN roll that is used for both reducing damage and deciding whether a critical hit happened, which leads to critical hits being even more devastating.

This disassembly snippet was posted by iRFNA's in 2015:

 v128 = opendice(6, 2);
 protdrn = v128;

 if ( hascurseluck > 0 && randnum(100) < 50 )  // curse luck is internal effect 255
   protdrn -= openroll(6);
 if ( *(__int16 *)((char *)fatigue + hitunitoffset3) >= 50 )
   --protdrn;
 posprotdrn = ((protdrn <= 0) - 1) & protdrn;  // set negative prot drn to 0
 if ( posprotdrn > 2 && (immobileunit(hitunitid2) || !immobilized(hitunitid2) || posprotdrn > 4) )// mobile unit has 1/36 chance to crit (protdrn <= 2)
                                               // immobilized unit has 1/6 chance to crit (protdrn <= 4)
                                               // this is ignoring the two protdrn alterations above
   protaftercrit = prot;
 else
   protaftercrit = prot / 2;
 isprotneg = posprotdrn + protaftercrit < 0;
 adjprot = posprotdrn + protaftercrit;
 prot = adjprot & (isprotneg - 1); //set negative prot to 0