let me first just say hi, new to the board here and just discovered it a few days ago and have enjoyed reading through it.
anyway, so a friend and i just started revamping an old house system we had from many years ago and we are having a problem working out the damage
from melee weapons.  it is a fairly generic system and firearms are all worked out, but for some reason we can't quite get hand weapons to do so.
so i'm wondering what other people's experience with doing so is and has anyone developed up a rough formula for doing so.
we want the system to be fairly realistic, meaning we shoot for realism but we err on the side of the player and fun.  so the system is sitting somewhere on a fine line of being just dangerous enough to discourage fighting your way through your problems but not enough so that you just run and hide or start prepping your next character when a fight does come up.

we have been playing with various combination of using the length, width, weight of weapons to get a base damage
we are looking at three types of damage: blunt, puncture, cutting
we don't want to just assign arbitrary numbers to weapons but rather to have a rough formula so that when someone picks up something like a stop sign and swings it the GM will have an idea of how to calculate the damage it could do quickly.  in other words, that there is some kind of rhyme  and reason as to where the damage for the listed weapons came from. 

so yeah just curious if anyone else has developed up any formulas based on the characteristics of weapons and if so what/how they went about it.

Hello, and welcome to The Forge, decline.

Rare is the game designer who has not put a fair amount of thought into the issue of how much damage should be done by a weapon.  The word "realism" is used to mean different things - by different people or even by the same people in different conversations.

Based on this selective reading of your post:

decline wrote:
(...) we err on the side of the player and fun 
(...) just dangerous enough to discourage fighting
(...) not enough so that you just run and hide

Decline,
  I personally, would like to give you some advice:
1) The wimpiest weapons can and do kill in real life.
2) No matter how deadly the combat is, it won't discourage players from resorting to violence

  So, in order to make the game more fun, I suggest a couple of tweaks:
1) Make the damage range between the strongest weapons an the weakest a small spread
2) Make other forms of conflict resolution more effective and more fun in play. Maybe talking someone down from stabbing you shouldn't be a single roll. And maybe it should be harder to hit someone who is in your head (so to speak). So you aren't bleeding to death by the time you convince them not to kill you.

  Does that make sense? Does that help? If not, let me know and I can approach my advice from a different angle.

I was starting to do something like what you describe for my survival horror game, under the assumption that characters would need to use their environment to survive. Basically, a weapon could be described by its length, weight, unwieldiness, sharpness, hardness etc. I would assign numeric statistics to the different adjectives, so a long and sharp/pointed weapon (fence post?) would have extra reach (depending on your system, mine would be +1 meter reach), and deal an amount of base damage (say maybe d6) and also some bonus damage for being sharp/pointed. I would also apply a penalty on the to-hit roll because a fence post is unwieldy and heavy.

If you're interested in this idea, let me know, I'd like to develop this system further but haven't had the chance.

Forget the notion of realism...how abstract or detailed do you want the system to be?...and in what contexts do you really want the detail to be manifest?

Damage is just one part of a weapon, and as John has mentioned in the above post, things like unwieldiness and sturdiness of the weapon can be used just as easily to differentiate between specific types of weapon.

I once went down the road of generating vast lists of tables for every real-world weapon then started pushing my tables into fictional settings...then decided that Palladium had produced a pretty extensive series of weaponry books in the 80's...why bother trying to reinvent the wheel? I moved on to the ideas of formulas to determine damage, weapon accuracy, unwieldiness and sturdiness, but found that most players just wanted a quick look at a chart to determine these things rather than having to recalculate them before equipping each new weapon.

My current theories in this regard have streamlined the notion a bit. Let a character buy a weapon for a point value and they can assign those points across fields:
Damage - Pretty obvious (more points spent here means more damage)
Range - How close you need to be to someone in order to hit them (more points here means you can strike from further away)
Accuracy - The chances of doing a critical or improved hit rate with the weapon (more points here means better chances of vital strikes)
Unwieldiness - The weapons speed (more points here means either an initiative bonus, a chance of a follow up strike or something similar)
Sturdiness - The weapons chance of breaking (more points here means a more resilient weapon).

If a player spends a certain amount of XP (or in-game currency), they might gain a bonus item in the form of a weapon. Once they've acquired this weapon, they can spend a bunch of points to determine the weapon they've acquired [some pre-generated weapon templates could give some ideas on how to spend these points to reflect existing weapons in the setting].

If a player wants to use a found item as a weapon, they could gain a number of points to spend based on the success of their search attempt. The more successful you are when looking for a suitable hitting tool, the more points you get to spend on it. Such makeshift weapons would never be as good as the weapons permanently acquired by characters though. [Again, a series of pre-generated templates can give some ideas on how points could be spent to reflect such things as a broken bottle, a plank with a nail in it, a lead pipe, or other things appropriate to the setting].

That's where I'm heading with my ideas, I just thought I'd share.

V

Vulpinoid wrote:

My current theories in this regard have streamlined the notion a bit. Let a character buy a weapon for a point value and they can assign those points across fields:
Damage - Pretty obvious (more points spent here means more damage)
Range - How close you need to be to someone in order to hit them (more points here means you can strike from further away)
Accuracy - The chances of doing a critical or improved hit rate with the weapon (more points here means better chances of vital strikes)
Unwieldiness - The weapons speed (more points here means either an initiative bonus, a chance of a follow up strike or something similar)
Sturdiness - The weapons chance of breaking (more points here means a more resilient weapon).
V

In my experiences, you produce a good mix of realistic and gameable damage output, by using a system with a fairly fixed range of outcomes, where the lethality of a weapon makes higher outcomes more or less likely. I mean, I'd rather be stabbed than shot with a .223 rifle, and I'd rather be shot with a .223 rifle than with a 40mm autocannon. But I'd rather not get stabbed at all, because, you know, a knife can still kill you.
Dice pools (like in World of Darkness or Burning Wheel, where you roll a number of dice based on skill, damage etc, and then get different levels of success based on how many dice come up with certain numbers) work with this principle, at least if the dice explodes in one way or another (highest result on a die grants you an additional die, for example), as a poor weapon wielded by an unskilled opponent can still kill you, it's just not as likely as surviving a shot from a rifle fired by a skilled marksman.
Another system which follows the same principle, is a system where weapon damage isn't a single number which you add to your rolls or dice pool, but rather a system where each weapon has a number of damage levels. For example, if we use a d10 to roll for damage, a knife might cause a superficial wound on 1-2, a severe wound on 3-8 and a lethal wound on 9-10, while an assault rifle perhaps would cause a superficial wound on 1, a severe wound on 2-5 and a lethal wound on 6-10.
A less clunky way of getting the same result as in the last example, would be to use a die or a playing card or something else with a fixed range of possible outcomes, and having the range of hit points (or similar) be within that same range. You then roll a number of dice (or draw a number of cards) equal to the damage value, and choose the highest (doing that many points of damage). For extra versatility, some weapons or sources of damage could have a negative value, where you calculate damage in the same way, but choose the lowest.

In both of the latter systems, you could have the attacker's level of success affect the damage roll in some way, but then you'd either have to use a bunch of tables (which wouldn't be that much of an issue if you include relevant tables on the character sheets), or find that the damage system only works with the intended level of lethality under a certain range of skill levels.

(the above may or may not make any sense due to lack of sleep and caffeine overdose)

John wrote:
I agree with this system, I would say maybe you could have a certain amount points for a weapon based on the quality. So a poor weapon might have 5 points,  an average weapon might have 10 and a superior weapon might have as many as 20 points.

John wrote:
I was starting to do something like what you describe for my survival horror game, under the assumption that characters would need to use their environment to survive. Basically, a weapon could be described by its length, weight, unwieldiness, sharpness, hardness etc. I would assign numeric statistics to the different adjectives, so a long and sharp/pointed weapon (fence post?) would have extra reach (depending on your system, mine would be +1 meter reach), and deal an amount of base damage (say maybe d6) and also some bonus damage for being sharp/pointed. I would also apply a penalty on the to-hit roll because a fence post is unwieldy and heavy.

If you're interested in this idea, let me know, I'd like to develop this system further but haven't had the chance.