Due to the relitave compelxity of shadowruns firearms resolution mechanics, I decided to try to write a program to analyze specific outcomes so that I would have a better understanding of how the mechanics work. If you have a copy of Matlab or similar program you can just copy and paste the code and run it. Things I have learned:
Damage value is a Huge deal, as is called shots/long bursts. If possible you should always go for a higher effective damage value.
-Wide bursts are useless in almost every situation.
-Armor is not very important except for a threshold where several points can make the difference between mostly stun damage and mostly lethal damage. An example would be with the having 20 armor instead of 17 armor gives you 40% less chance to take lethal damage, while only lowering the actual damage you will take by ~5%
-It is almost impossible to disable a moderately optimized combat character in one simple action, and almost easy to do it in two.

I'm Posting this here so that some other people can play around with it and fix bugs/learn more about how the mechanics of Shadowrun firefights ACTUALLY work.

clear all
close all
format short g
format compact
screen_size=get(0,'ScreenSize');
clc
fig=figure;
set(fig,'units','normalized','outerposition',[0 0 1 1]);
tic
for a=6:2:22
for b=4:2:10
for j=2:2:26
for n=[0,1]
for l=4:2:20
for e=[0,1]
clearvars -except a b j n l e fig
%a=12; %attack Pool
%b=6; %Gun Damage
%c=20; %Recoil Controll
d=6; %Burst Count(semi=1, short=3, long=6)
%e=1; %Narrow burst(1=yes, 0=no)
f=0; %Off hand(1=yes, 0=no)
%g=0; %called shot(1=yes, 0=no)
h=0; %range penalty
%j=12; %Defense pool
%l=7; %Armor
m=1; %Health
%n=2; %armor piercing value
o=ceil((l-3)/3); %body value
c=d-1;
g=e;
info=sprintf('Attack Pool %1.0f, Gun Damage %1.0f, Shot Count %1.0f, Narrow Burst %1.0f, Called Shot %1.0f, Defense Pool %1.0f, Armor %1.0f, Body %1.0f, Armor Piercing %1.0f',a,b,d,e,g,j,l,o,n);
i=a-4*g-2*f-(d-1-c);%total attack pool
k=j-(1-e)*(d-1)-n;%total soak pool
m=zeros(50,1);
m(1,1)=2/3;
m(2,1)=1/3;
if i>1
for x=[2:i]
m(1,x)=m(1,x-1)*2/3;
for p=[2:50]
m(p,x)=(m(p,x-1)*2/3)+(m(p-1,x-1)/3);
end
end
m=m(1:50,i);
end
if k>0
for x=[2:k+1]
m(1,x)=m(1,x-1)+(m(1+1,x-1)/3);
for p=[2:49]
m(p,x)=m(p,x-1)*2/3+(m(p+1,x-1)/3);
end
end
m=m(1:50,k+1);
end
for x=[1:50]
m1(b+e*(d-1)+g*4+x,1)=m(x,1);
end
m=m1;
%shock=stun
%lethal=physical
shock=zeros(50,1);
lethal=zeros(50,1);
oneHkill=zeros(50,1);
twoHkill=zeros(50,1);
for x=[1:50]
if x>l-n
lethal(x,1)=m(x,1);
else
shock(x,1)=m(x,1);
end
end
if l+o-n>0
for x=[2:l+o-n+1]
shock(1,x)=shock(1,x-1)+(shock(1+1,x-1)/3);
for p=[2:49]
shock(p,x)=shock(p,x-1)*2/3+(shock(p+1,x-1)/3);
end
end
shock=shock(1:50,l+o-n+1);
end
if l+o-n>0
for x=[2:l+o-n+1]
lethal(1,x)=lethal(1,x-1)+(lethal(1+1,x-1)/3);
for p=[2:49]
lethal(p,x)=lethal(p,x-1)*2/3+(lethal(p+1,x-1)/3);
end
end
lethal=lethal(1:50,l+o-n+1);
end
total=lethal+shock;
for y=[1:50]
lethal1(y,1)=sum(lethal(y:50,1));
shock1(y,1)=sum(shock(y:50,1));
end
total1=lethal1+shock1;
x=[0:49]';
plot(x(1:20),shock1(1:20).*100,'- .b',x(1:20),lethal1(1:20).*100,'- .r',x(1:20),total1(1:20).*100,'.k',x(1:20),shock(1:20).*100,'-. .b',x(1:20),lethal(1:20).*100,'-. .r',x(1:20),total(1:20).*100,'ok');
axis([0 20 0 100])
grid minor
xlabel('Damge')
ylabel('% Chance, Solid: At least this much damage, Dashed/Hollow: Exactly this much damage')
legend('Shock Damage','Lethal Damage','Total')
title(info)
print(fig,info,'-dpng')
end
end
end
end
end
end
toc

Stand back, he's using MATH!

You might want to put the code in spoiler tags. It looks like you were examining automatic capable weapons. Did you perform any analyses of pistols?

1: how do?
2: The weapon is defined by: damage, recoil compensation, burst type, and armor piercing value. I think that those values cover basically every posible firearm with the exception of special ammo like S&S.

One point about armor - the thing with it is that it's relatively "cheap" compared to other things. For example, for 30 bp, you can get +3 agility (for +3 dice to hit, which if you were already going to hit is +1 DV on average) or +3 Body and thus +6 armor for 9 extra soak, and thus about -3 DV.

Well, that's the reason that everyone uses a combination of defense (diminishing returns and differing costs). It's less complex in Diablo 3, but you see the same thing: armor, dodge, and resistance all reduce incoming damage (and increasing HP lets you take more), but the mechanic mostly pushes you to raise them all.

I was just pointing out that the reason armor seems lackluster is when you look at how valuable raising soak by a few points is, without accounting for its relative cheapness.

Electing a Wide burst is sensible if an agile opponent has a dodge pool close to your attack pool, but not the armor to just take a hit. As long as you can assume a hit, -2 dice on the opponents pool is dominated by +2 damage. Where do you figure in the possibility of a successful dodge? I am challenged reading that code.

The value of armor depends on the campaign, much. We play close to street level currently, and high armor is much easier to get away with than a large firearm. If you are in the "2 hits = gone" bracket, go for initiative. Hard.

While it may seem lackluster, it ends up being very important for deciding if you are unconscious or dead



Ignore the legend. the blue lines are defense values from 15 to 35 in increments of 5. the cluster above is a narrow full burst, the five on bottom are wide bursts. Acording to the math, wide bursts are completley inferior except in extreme edge cases.

[IMG]http://i.imgur.com/K4BTS.png[/IMG]

Worse: Algebra.

Except for one of the most common edge cases, spirits. If you can't count the increased damage from narrow burst for overcoming ItNW, then wide burst may be the better option. Still not as good as called shot.

I agree.

Who has defense pools of 15 to 35?

I take it we aren't talking about reaction here.
Is this Reaction + dodge + Body + Armor?

Never mind - I also don't see Characters with Attack pools of 20. 12, sure, sometimes, but not 20.

Ditto. 20 is pretty much the upper-end maximum possible by the rules. 15 Reaction is also strongly weighted towards the absolute max (hell, most people don't even have 15 armor).

20 is not uncommon for an Adept specialist

How common is that? (IMG:style_emoticons/default/smile.gif)

His analysis is seriously flawed... you only do more damage if you actually hit. If you look purely at soak then yeah +1 damage for -1 dice is always the best deal.

On the other hand... if your'e having trouble hitting (going from 50/50 odds of hitting to say 66/33 is quite significant and doesn't take a lot) and this can skew your results quite a bit since you've gone from 1 for 1 to 2 for 1.

Also as brought up spirits and vehicles... both commonly have reaction pools of 5-9 dice. Every dice out of this pool you cancel is one less dice reducing the damage below their ignore the damage hardened level. A tank doesn't care how many spitballs you fire at it on full auto... it only cares how many end up in vulnerable spots.

In a low level ganger campaign... I'm only tossing ~8 dice on attack. Against opposition in cover also with ~3 dice (3 reaction + 4 cover). That shifting of the pool to to 8 vs 5 from 8 vs 7 is huge for actually hitting most of the time and doing reasonable damage. Simply hitting and doing damage is a big help as well.. wounded opposition has more trouble with penalties on them. Missing and still not doing enough damage to drop them in one shot doesn't help much. If I'm needing two bursts to down a target... two wides do the job far more often than two narrow simply because they hit far more often.

as has been pointed out, wide bursts can help improve your base DV for the purposes of getting through hardened armor, damaging vehicles, and dealing with immunity to normal weapons. as well as making damage more likely to stun, and of course dealing with opponents that would otherwise just completely avoid getting shot at all, or using ammunition where hitting once is the most important thing, such as capsule rounds and such.

that said... those situations (wanting to damage vehicles, wanting to damage spirits, dealing with hardened armor, needing to deal physical damage, needing to be certain to hit) are not really all that uncommon at all. they may not be the majority, but i wouldn't be surprised to see some need for these things come up at least once a session.

wide bursts are also more likely to be valuable to NPCs than the runners. odds are good the runners are throwing enough dice to regularly hit for good damage anyways. NPCs, especially when dealing with the runner team, will likely need every advantage they can get just to hit in the first place. runners may want to use wide bursts once per session. NPCs may very well be trained to use wide bursts for almost any situation.

it definitely adds some benefit.

Your general conclusion is based on a high attack pool, where misses are rare. You are also looking at a weapon that will usually not cause damage without several net hits, and has little gain from hitting just so. You need to model misses if you go into lower pools.

Look at a guard (Agi 5 Skill 5 Smart) shooting a runner on full dodge (Rea 7 Dodge 3 Ranged+2) or in cover. Those are not fringe cases the way we play.

agi 5 skill 5 is not exactly a standard guard. for an ordinary guard, you're probably looking at around agi 3 skill 3, for a total dice pool of 8 after smartgun. maybe 10 if they specialize, but even that would be pretty unlikely for an ordinary run-of-the-mill guard.

The initial values posted were for my mostly optimized gun bunny shooting his clone.



I would love help in finding any math errors, but if there are none there, than im not sure how the analysis can be "flawed" unless the assumptions are flawed. The assumptions I made were that a mostly normal metahuman shooting at another mostly normal metahuman with a "bullet" firearm. Situational modifiers such as cover bonuses should be factored into the relative dice pool(ex: -2 to attackers dice pool due to concealment). Appart from that, saying " analysis is seriously flawed" is basically saying "math doesn't work"...and math does work.



You are right, however, vechicles and hardened armor are outside the scope of this program. The program DOES account for dealing 0 damage if the attacker fails to get >0 hits on the attack-defense roll.




I am modifying the program so that it will make image files for each permutation that a metahuman vs. metahuman firefight would reasonably have. I would aprechiate suggestions on the lower and upper bounds for values such as attack pool, defense, damage, AP, armor, etc. I also need a web based service for sharing over 5000 image files in a searchable-by-filename format. I am thinking a .torrent, but I don't have much experience in this area so suggestions are appreciated.

Well, at a minimum, you have at least one flawed assumption. Cover bonuses are ADDED to the Defender's Pool, not subtracted from the Attacker's pool. (IMG:style_emoticons/default/smile.gif)

depends on if your using standard 4th ed or anniversary edition.

In standard it was subtracted from attacker's pool. Anniversary edition changed it to add to defenders pool instead.

True... Forgot some people could still be using vanilla SR4. (IMG:style_emoticons/default/smile.gif)
The change in 4A makes a LOT more sense, though.

They both 'make sense'. It's just a question of where you want the glitches.