I am somewhat obsessed with medieval weapons and I have written extensively about them, trying to make sense of physics, history and, well, AD&D.

Unfortunately, I haven't been able to find or create a system that satisfies my needs (I want: simple, sensible, and detailed). Here is my latest attempt:

https://methodsetmadness.blogspot.com/2024/04/a-new-terminology-for-d-weapons-and.html

AD&D measures weapons in several ways I find interesting:

- Price, weight, and cost, as most RPGs do.
- Speed.
- Length.
- Effectiveness against armor.
- Effectiveness against large creatures.

3e also measures frequency and power of critical hits.

My main problem is that most systems feel inaccurate and too fiddly for me.

In AD&D, for example, a two-handed sword is heavier than the longest polearms and heaviest axes and maces, IIRC. And the weapon versus armor table is confusing and too extensive.

In 3e, all swords have good "crit ranges", making them good against plate armor - especially heavy and curved swords!

Now, my question is: does anyone else care about this stuff or is it just me? Do you want weapons to have any detail beside 1d8 damage? Do you know any systems that deal with this is a detailed, sensible and simple manner?

EDIT: I'm also considering writing a 20-40 pages PDF on these issues, maybe with optional rules for OSR games, but I don't know if anyone else is as interested in medieval weapons.

With weapons you can have simple or you can have detailed: not both.
For example, an european longsword and a katana (or better, a tachi) of the same period (let's say middle 15th century) are functionally identical against unarmored opponets.... but the materials they're made of are different enough that the defensive techniques for them are wildly different.
How do you track that (important) difference in game while staying "simple", especially in a D&D like framework ?

Well, there's always GURPS and Rolemaster.

GURPS Low-Tech is probably the best I've said for dealing with different weapons vs. armor types. Fantasy Hero offers a fair amount of customization without quite that level of detail.

The longsword/katana distinction might be more detail than I want... but maybe longsword is slightly better against armor (or parrying), and katana does slightly more damage to large creatures or unarmored foes.

GURPS was good at this, but too complex to the point of becoming unrealistic: a 10-second fight between two people has 20 or more sword blows, most being parried.

You needed additional optional rules ("lulls", not sure which product) to make it more realistic.

The 1-second rounds also ruined axes and maces IIRC, because they attacked once in every two rounds (not sure about that).

Rolemaster was pretty fun, but relied too much on tables.

The thing I'm looking for is similar to AD&D, but a bit simpler. For example, reducing AC bonus to a single number:

You shouldn't be using categories and then collapsing them into real names.  A long sword is a real thing, it shouldn't be derived from your categories.  You should have a category that describes a long sword and then apply it to that.  In your system, for instance, you would have the idea of a large bladed slashing weapon, and that's the rules that your system would use for a katana or a longsword or a khanda or anything that works that way.

It's unusual to see the classification "great" (which is used in some real world names) over "large", but that's ok.  Did you require two hands on that classification somewhere?

Anyway, you can't serve two masters here.
You can go for realism, which is a great goal, and you'll definitely end up with your system favoring some weapons over others.
You can go for game balance with realism fulfilled as much as possible, which will give you a huge table like in multiple AD&D books.  That's fine too, and you can do the game balance better than those tables had too.  Note that in certain systems, weapon length is really important, and in others, it is not. Similar for speed factor.  So if you have either or both of those, make sure it's clear what you are balancing it for.
Finally, you can go for simplicity.  This gives you the game balance mostly for free, and your table is small, and weapons fit into your categories easily enough.

But you CAN NOT do simplicity and realism.  This is because the real world weapons are not simple in their differences. You'll have to file off enough details to consider reasonably different weapons as the same in your system, or you'll have to proliferate stuff.  The idea that "curved weapons crit more" and "pointy weapons crit harder" isn't really realistic, for instance, even though the idea is simple.

I don't want to choke on weapon details but something I've been homebrewing for years is a damage plus effect type of thing.  Here's a sample:

A spear can keep an enemy at bay for a turn if the attacker gets 2 or more above the target.
An ax will reduce the armor of a target by -1 if the attacker gets 2 or more above the target and -2 on a crit.
A mace will concuss a target with a crit.

Quote from: Venka on April 22, 2024, 12:02:37 PMIt's unusual to see the classification "great" (which is used in some real world names) over "large", but that's ok.  Did you require two hands on that classification somewhere?

I used "great" over large because greatsword, greataxe, great mace, and greatclub are all easy to understand. Yes, these would require 2H.

Quote from: Venka on April 22, 2024, 12:08:43 PMBut you CAN NOT do simplicity and realism.

I see what you're saying, but there are shades of gray. B/X is too simplist for me, AD&D too complicated. There must be a middle ground somewhere (Hyperborea, which I haven't finished reading, seems to aim at that).

I struggle with this too and the pinnacle of realism AFIAC is a matrix of weapon vs armor with each cell describing initiative, to-hit, and damage. AC, weapon speed, etc are just abstractions that simplify the matrix at the cost of realism. But even that matrix does not account for the defensive capabilities of weapons. So then we need a matrix of kit vs kit. 

I've created small matrices like this but then I flip the other way and think it should all just be role played with ad hoc bonuses or penalties. Such is life. I'm defintely interested to see what you come up with.

You could look at pendragon, weapons there do differing effects by type such as damaging shields or armor or tripping or what have you.

There are compromises between detail and realism.  As an example, for me the D&D weapon vs armor table was too much.  But for my previous project I decided to go with a 3x3 matrix, with slicing, piecing, and blunt weapons on one side, and no armor, flexible armor, and stiff armor on the other side.  It works alright if you're willing to take them time.  In my current project, I went another way entirely.  There are about 30 combat maneuvers, and each weapon is better or worse at a few of them.  So axe gets a plus on furious attack and a minus on defense and so on.

I'm also interested in this kind of thing.  I don't think it is quite so stark as what Venka expressed--though what he said is still mostly on target.

I would say instead that you can get some mix of realism/simplicity/other if you limit yourself to "good enough".  Or in the case or realism, you instead think of it as "nods to realism".  That is, a game is a model, which is already not entirely accurate even if you toss out all concerns for simplicity, ease of play, game options, etc. 

For "nods to realism" to be of any consequence, they must be reflected in the model.  It can't be entirely shallow, either.  For example, from a realistic modeling stand point, just looking at medieval melee weapons, you'd like to distinguish cut, slash, bludgeon, pierce, and then also some of the combination effects (say, bludgeon/cut from an axe).  However, if your system has no reasonable mechanic to model the different effects of cut versus slash, then the distinction is useless.  If you roll everything up into a damage against hit points, then even the slash/bludgeon/pierce may be overkill.

That's where the "nod" comes in--when you have a mechanic like occasional damage reduction against certain types of damage, so that slash/bludgeon/pierce matters a little.  It's only a nod, because it doesn't get to modelling any types of attacks or specific armor piece but it does give a reason in game tenuously related to reality why someone would use a different weapon.

So my answer is that first you determine what mechanics your system is wiling to support that will meet your simplicity goals.  Only after that is done can you then decide what nods to realism will work. 

That's also why this kind of thing never really works well when reverse-engineered on top of D&D while trying to be backwards compatible.  The D&D mechanics don't work that way.  If you change them so that they do work that way, you are moving away from compatibility.  The hoops that one jumps through trying to retain that compatibility and simplicity while tacking on the realism have an even worse complicating effect than Venka outlined, compared to building a system from the ground up.

Eric,

i like where you're headed, and it somewhat echoes a project I'm working on. But I ditch realism:

1. Four weapon categories: d4, d6, d8, d10
2. Five weapon types: axes, blades, clubs, flails, polearms
3. Imagine a matrix (cat x type) but not all of the cells are filled.
4. Two handed = d10 weapons and polearms

Example:

Blades
d4: Dagger
d6: Short sword
d8: Broadsword
d10: Great sword

Clubs
d4: Sap
d6: Club
d8: Mace
d10: Great club

Some design principles:

1. Rules light
2. Playability > realism
3. No historical accuracy
4. A sword is a sword. "Broadsword" covers a saber, katana, scimitar, tulwar, falchion, etc.
5. Each type has a feature (polearms can keep enemies at reach, etc.)

I throw a lot of complexity out the window, both from AD&D (weapon speed, weapon length, weapons vs. armor, etc.) and 3.5 (simple vs. martial, double weapons, weapon sizes, etc.)

The D&D combat system was designed to be simple and abstract. Tacking detail on it generally leads to unsatisfactory results. The basic D&D combat chassis of rolling to hit, and rolling simple damage vs a pile of mounting hit points starts to have issues the more you try to jazz it up.

I have my own issues with it, the primary one being variable weapon damage. It is a system that creates winner and loser weapons simple due to the assigned damages. My current project that I am working on for a planned OSE campaign is removing all weapon damage completely and starting over. All of these medieval weapons are very deadly and can inflict brutal wounds. You will almost never see a classic D&D hoplite fighter using spear & shield when using var weapon damage because choosing a d6 over a d8 weapon is sub-optimal. I am moving away from specific damage from particular weapons. I have all weapons classed as small, medium, large, and great. The actual damage inflicted depends on the expertise of the wielder. Martial classes have the best damage, semi-martial classes have medium damage, and non-martial classes have the lowest damage.

I am also doing away with multiple attack rolls. Instead, base weapon damages go up with level gain. The more martially focused the class, the quicker weapon damage rises. I am also giving martial and semi-martial classes a base AC boost with armor adding on to that. Still tweaking the numbers on the scaling damage amounts but they are significant. A barely trained militiaman might only do 1d3 damage fighting with a dagger but a grizzled 7th level fighter will be doing 1d10 with that same dagger.

Oooh... We're touching on a subject which is near and dear to my heart.

For my money, if you want a fairly realistic approach to the way ancient/medieval/early modern weapons and armor interact, the optimum way is to address three factors for weapons:

--Ability to circumvent armor: i.e., the ability to strike at gaps or weak points in armor.
--Concussive force: i.e., blunt force trauma,
--Penetrative capability: slashing and piercing.

The ability to circumvent armor by striking at weak points is critical, because it is the entire purpose of many weapons, such as the rondel dagger and most types of late medieval sword. Some historians assert that it is also the purpose of massed bow fire. As far as concussion and penetration goes, the balancing there is fairly obvious: A piercing sword like a rapier has good penetration, but almost no concussion, a cutting sword a bit more concussion and less penetration, an axe a fairly even mix of the two, and a mace pure concussion.

Armor would then optimally reflect the same three factors, with a rating for thoroughness of coverage, and concussion and penetration resistance stats. Mail for example, would have good penetration resistance, but minimal concussion resistance. Textile armor would have little of either, but perhaps better against concussion than penetration, and renaissance style full plate good levels for all.

This is the only system I can think of which represents the choices of weapon vs. armor as they were in history. E.g., if you expect opponents to be in head-to-toe mail, as you might if you were fighting 12th or 13th century knights, you really want a lance (which should have an extremely high penetration rate when couched from horseback) mace or hammer, but if those aren't an option, a battleaxe is a good second choice, which is exactly the trend in battlefield weapons during that time.

The additional wrinkle you could add in is a bonus to combat checks (attack if you're using D&D rules, but optimally both attack and defense), to represent the combination of reach and wieldiness. That further ups the realism of the decision making, since while an axe might be a more damaging weapon, a sword is much more nimble.

The point of doing all this for a game IMO, is to give your players a more interesting choice to make when choosing their loadout, and to allow you as DM to mix up the dynamic by changing the weapons and armor in play. Being forced by social circumstance to carry "civilian" weapons such as swords and knives makes much more difference when a sword and an axe no longer have the exact same utility.

Admittedly, the level of complexity this requires is possibly too much to be practical. I haven't tried to implement it in my own games, but where I can, I do aim for half-measures that move towards the same effect.