I think one of the issues I am seeing is that people here are talking in avg rolls which in the end result does not represent the whole picture but only in part.

If the system only worked on avg rolls it would be much easier to handle but it doesn't. At times there are big jumps and huge jumps that can have up to above avg and massive hits.

Such as if the attacker/defender rolls avg and the opposing party rolls high and massive effect if one rolls severely poor and the other rolls exstreamly well which I have been seeing and its not rare.

I think one of the issues you have going is that its pretty much a all or nothing system with very little mitigation going on to help even out to a degree the hits that are taken.

This is all and well if you want a nitty gritty game and don't mind rolling up characters. But for Players and GMs alike this can be a issue when the game is more about the story and fun then it is rolling up characters.

This is not to say a game should not be dangerous. There should always be chance of death from to much firepower incoming or just plain stupid actions on behalf of the players.

But its about risk vs reward. To a degree I think a games deadliness should be some what balanced on how much time it takes to build a char vs realism.

The tweaks we have done makes armor some what less reliable as in the all or nothing damage but add a mitigation factor in that helps mitigate the incoming damage depending on the armors lvl stated in the books.

Before I got further, I will give you our combat formula so you will understand what I am talking about. At the moment we are happy with it for it was designed around our play style but it may not suit yours but it still open to tweaking as we go.

® = Rolled Dice, (AR) = Armor Rating

Step 1) Agility ® + Skill® vs Reaction ® = Net staging
(same as before)
Step 2) Net staging + DV (Gun/Spell) vs Half AR + (Half AR ®) - AP from Weapon/ammo = Penetration
(if penetration is sucessful)
Step 3) Net staging (from step 1) + DV Weapon + DV ammo/burst vs half base AR + body = Damage taken.

It looks like more then it is since I wrote it out to be simply understood, but after you get use to it, it flows smoothly.

Now its easier to take a hit but the hit is now not has harsh as it once could have been and reaches more of a happy medium.

This was made to take into account that with decent armor avg joe with avg hits could not make a dent in the armor and those people who min max armor where close to juggernought. This evens the playing field out more but still alows it to stage up with better grades of protection.

That was our Combat formula which we are also applying to magic where its physical.

Physical magic still works off of impact but impact now has its full value + any value that is purchased in the forms of Fire, Thermal, Electrical protection when needed.

Also value rating for added protection is a flat 6 to be divided up as the player wishes. Could have a + 6 to fire added on to the armor or + 2 in 3 separate categories that add up to 6.

That is just for Physical Magic, the none Physical magic is much harder to deal with but we came up with a idea that may help some based off of the concept behind the none physical magic.

mundane Stun: Body + Will
Magical Stun: essence Rating + Will
Mana Attacks: essence Rating + Will

Some body brought up fetishes for the mundane and I think that is a very interesting idea there and fits in well with history which brings even more realism to the game.

(This will need more tweaking for a obvious reason that it adds to mages too and could possibly make them to hard to hurt with magic, further testing will be needed)

See the issue is, that the concept behind Magical Mana/Stun attacks was to be the balancing factor between Mages and Sams. The more borged you were the more vulnerable you were to a Mage.

Now there is a second issue. In the older versions mages rarely got off more then one action and went last most of the time.

In the new system a mage can now get off more then one action and it doesn't always means he goes last. So you now have a compounded system where one was to make up its weakness, and now its weakness can be taken out if properly done which has caused some issues.

One idea which we have not applied yet is that only one magic spell can be cast per round and the rest he will have to spend shooting. Concept behind that is the body can only handle so much magic going through it per round, not per action phase. But its still in debate. But it would make mages more tatical in the use of their magic.

But again it all comes down to risk vs reward, if magic is too powerful then its over balanced and if you handicap them to much, why play one?

So when you balance it, it should be Power of Attack vs How often you can attack vs Drain vs Defense.

I don't think Drain is a issue, I think how it is defended against and how often you can cast it is.

The problem is not only easy of hitting a mundane but also the damage done.
With a magic of 5 and spellcasting of 4 and spell cast at force 10 against a will of 3 and counter spell of 4 if any will most like end up being mage 13 mundane 2. And I know people say the sammy has two shoots to equal that but the mage gets a chance react to it and armour to protect him

yes that is true but there are many way to raise reaction such increase reflex. I
f you do not want to have it on a fous all the time and play a PC for the fun you can spare a point of essencs and get bio or cyber.
Where as the mundane does not have that option for will at this time.

FYI, the increased reflexes gives a straight boost to initiative, not reaction. You could use Combat Sense in a focus for that purpose, though.

oops I guess that increase attribute. I guess you can choose which attribute spell to learn.

@Tinkergnome:
I´d use attribute as threshold and add counterspelling successes.

How would you introduce a chance for the mage to occasionally fail? Without spell defense?

Overcasting isn´t the only problem, as illusion spells for example will rarely ever be overcasted - why should they?

I think that attribute value + counterspelling successes as THRESHOLD (no need for a net success) will work just fine. A specialised mage would have a noticeable advantage for once.