BlackSquirrel, on 04 April 2013 - 10:46 AM, said:
Recently I keep reading this in various forums on mech building/load outs. However I've yet to see proof on if this is true or not...or even pseudo math.
So is this in fact true, and what is the math behind it? Seem dubious reasoning from the answers I've gotten so far.
Thanks.
Critical hits are randomly applied among all
occupied crit locations on a mech. (Note: the slots from Endosteel and Ferro Fibrious don't count for this, they are effectively empty and ignored by crits).
If you have a single Gauss Rifle in a hit location, the Gauss is the
only weapon that can get crits. There is no other crit slot occupied in that location. If you fill the other (up to 5, depending on location) crit slots, there is only a 7/12 chance that the Gauss gets hit.
Now, the Gauss is an easy case - the Gauss explodes, not the ammo. Other weapons have explosive ammo.
Why can it still be more useful to put in explosive ammo as "crit padding"?
Each item has by default 10 hit point. That means most items can actually survive multiple crit hits, as most weapons don't deal that much damage. Pretty much only PPCs, AC/10s, Gauss Rifle and AC/20s deal enough damage in one blow to take out all item hit points. All other weapons deal either less damage, or deal t heir damage in small damage ticks, and each damage tick has to determine crit and crit location seperately.
A single ton of ammo occupies only one crit slot. That it gets critted twice in sequence is only a 1/144 chance.
Padding an AC/20 with 2 tons of AC/20 ammo will not do all that much for the AC/20. The chance of it being it will still be 10/12, but this also means that the risk that your AC/20 ammo blows efore the AC/20 blows is much lower. But at least you're lowering the chance of an AC/20 hit a bit. Heavier weapons have better gains to have.
This is probably only a good idea for weapons that occupy more than 6 slots (otherwise the chance of a crit blowing up a randomly determined ammo location or the weapon itself are equal). But I'd say Gauss, AC/20, AC/10 and maybe LRM20s are okay.