Balancing Classes in Swords and Wizardry

DM: "The box is the size of a small trunk; it is latched but not locked." 
Dougal: "I'm looking for traps on the box." 
DM (rolling for Dougal's "find traps" ability. The roll indicates that Dougal has failed to find the poisoned needle in the latch.) "You don't find a trap." 
Morgan (the group's leader): "Black Dougal will open the box." 
DM: "Black Dougal, you find out that you missed a tiny discolored needle in the latch. Roll a saving throw vs. Poison, please!" 
Dougal (rolling): "Missed it!" 
DM: "Black Dougal gasps 'Poison!' and falls to the floor. He looks dead." 
Fredrik: "I'm grabbing his pack to carry treasure in."

Game balance. It's a tricky concept in RPGs. D&D 3rd and 4th editions made a very strong effort to balance the character classes with one another as well as to balance encounters. For my purposes, they went too far. As a Dungeon Master I found prepping for a 3rd and 3.5 edition game to be far too onerous a task. there were certain assumptions about balancing encounters, assumptions about what sort of magic items characters had, etc. Coming from a background of Advanced Dungeons & Dragons 1st and 2nd editions where I was used to making judgement calls it was a difficult adjustment for me.

On the other hand, you want to avoid creating a super-class or race. For those of you who remember the 2nd edition of AD&D there was the infamous Elven Bladesinger - a "kit" available to fighter/magic-users. It gave a large number of bonuses with, what most people felt, a small number of disadvantages.

In my opinion, Original Edition D&D and 1st Edition AD&D did the following to achieve balance:

  • Magic-Users were very weak at low levels but became very powerful as they reached intermediate and high levels
  • Fighters were extremely dominant at low levels but did not see as large a rise in power as Magic-Users. I have seen this described as the Linear Fighter, Quadratic Wizard problem.
  • Clerics are generally useful at all levels - they start off nearly as good in combat as fighters and gain magic which tends to assist the party.
  • Thieves always suck. (OK I exaggerate - thief abilities improve quite a bit but at low level thieves tend to be very poor at the abilities which define their class and failure at using these abilities can result in instant death - alas, Black Dougal.)
  • Non-humans are discouraged by having lots of abilities early on, the ability to advance in multiple classes at once, and limits on their advancement. This, I found, did not achieve what it set out to do. Most campaigns I was in maxed out in the mid levels. before level limits became issue.
Swords & Wizardry Complete is designed to model D&D as it was played just before AD&D came out. Overall, I think it does a good job of doing this. There's a few wrinkles here and there. I suspect most of there to avoid copyright violation - it recreates the feel of the Original Edition by using the System Reference Document of D&D 3.x and going backwards.

With regards to game balance it makes use of the same methods applied by Original Edition D&D. For me, that leads to some issues I had "back in the day". I should note that none of these problems were crippling - we still managed to have a ton of fun. But I suppose the engineer in me must tinker and I've been thinking of some house rules I might apply to a Swords & Wizardry game.

My general belief is all characters should be balanced for "fun". You should, in general, always be able to feel useful to the group. Not useful in all situations. If you are a magic-user you should be getting some enjoyment out of shepherding your resources for maximum effect. A fighter is not going to have much to contribute to examining dusty tomes of arcane lore. But in general you should not feel like a bump on a log.

My other guiding principle is that house rules should be as minimal as possible. Ideally they would mean no changes to a character sheet at all. A new player bringing a character to the table should be able to have the house rules explained to him very briefly and his character sheet should be useable as is if at all possible.

Going forward I will bring up the various issues I see and how I would address them.

Non-Human Characters

In addition to level limits, Swords & Wizardry Complete limits demi-human characters to the four core classes - fighter, magic-user, cleric, and thief. For the first three there are level limits, for the thief there are no so such limits. Therefore, aside from for roleplaying reasons there is absolutely no reason to play a human thief.

I would, in general, keep the level limits as they are. I would also give some tweak to all of the core classes for humans only - a little advantage to being a human member of that class. For each of the core classes I will mention what human bonus I would give.


At early level I find the fighter is in pretty decent shape. However, at high levels there is a tendency to find the fighter lagging behind the magic-user in combat utility with the mage wiping out rooms full of enemies. 

To be honest, I rather like the simplicity of fighters generally having one attack per round save vs. foes of minimal hit dice (where they can make a number of attacks equal to their level). Adventurer Conquerer King System (ACKS) has adapted the Cleave feat from the SRD as an ability of all characters, with fighters being best at it. With Cleave, if a character brings a foe to 0 hit points or lower, he or she can immediately attack another foe. And if that foe is dropped, there can potentially be more attacks. Fighters can do this a number of times equal to their level whereas other classes can do it less. For Swords & Wizardry I would give this to fighters and only to fighters. I would also not allow it to be combined with their regular multiple attacks ability. My inclination would be to let humans do this a number of times equal to their level and demi-humans equal to half of it. 


Generally speaking I think clerics are pretty well balanced. Even without starting with a spell at 1st level they are still good at combat, able to wear any armor, and able to turn undead. It's not a bad deal. As they level up they lag behind fighters a bit in combat but begin to get a nice combination of spells - healing and otherwise. 

There are two general modifications I would make. First, I would allow a cleric to prepare short-term scrolls, a concept I will discuss below with magic-users. Secondly (and this would be a slight reduction of ability), I would be strongly inclined to allow intelligent undead a saving throw to avoid destruction when being turned. If successful I would rule that the undead is still harmed, taking 1d6 of damage per level of experience of the cleric.

I am a little uncertain if I'd give additional abilities to human clerics. The only other race that can become a cleric in Swords & Wizardry Complete is the half-elf, a race that gets somewhat modest abilities. At most I would grant human clerics a single spell at 1st level, but I am uncertain that is needed.


The poor 1st level magic-user. 1d4 hit points. One 1st level spell. A dagger or staff. No armor. This is not a character who should engage in melee combat.

I believe the Holmes Basic D&D rules has a very minor enhancement to the magic-user which I would be inclined to make in a modified form. Swords & Wizardry does not provide hard rules on how much a spell scroll costs to make or when it can be made. Original D&D suggests that a spell scroll should take 1 week per level of the spell. While this is fine at low levels it quickly becomes unbalancing, with a high level wizard able to make a nice collection of 8th and 9th level spell scrolls for a fairly small amount of cash.

With this in mind I began thinking of the Dresden Files series of books. In them, the protagonist, the detective and wizard Harry Dresden, often prepares a series of potions that he believes may be of utility on the case or mission he is on. He doesn't seem to have a stock of potions on reserve with which to draw from. That makes me think they don't last long. So what I propose is that magic-users be able to scribe Lesser Spell Scrolls (for lack of a better term). They cost 100 gold pieces per spell level and take either a week or a day per spell level to scribe. However, they only last a number of weeks equal to the spell level of the scroll. The magical energy can only be contained for a limited amount of time. This allows a magic-user to prep for what he thinks he will face in his adventure but prevents him from amassing a gargantuan collection of scrolls with the cash he will build up at higher levels. My inclination would be to allow clerics to do this as well. It is quite possible a 1st level magic-user would have enough cash to scribe a single one of these scrolls, giving him more utility in his or her first adventure.

I would keep these Lesser Spell Scrolls separate form more traditional spell scrolls. I would apply more thorough requirements for scribing traditional spell scrolls - costly and rare components, more cash to be spent on rare materials, etc. The advantage to these greater scrolls would be that they last as long as the scroll they are written upon does. 

Note that there is no reason these Lesser Spell Scrolls need be actual scrolls. They could, for example, be imprinted upon an otherwise non-magical wand, staff, or holy symbol with the approval of the Judge.

I've spent a lot of time talking about these Lesser Spell Scrolls. However, I believe this is also a case where human magic-users likely need an edge over their elven counterparts. I'd considered an additional spell per level but that seems too much an advantage. When reading the Complete Rules the following gave me an idea:

In Swords & Wizardry, a beginning Magic-User’s spellbook contains as many of the eight basic first level spells as the neophyte character can know. Check each spell to see if the Magic-User can learn and know it, using the “Learn Spells” column on the table in the description of the Intelligence attribute.
[Note that there are actually ten first level spells in the Complete Game.]

I'd propose the following modification. All characters start with Read Magic, followed by the check described above. Then, human characters can swap out one known spell and make it unknown and in its place know a spell they failed to learn (and include it in his or her spellbook). This is a modest improvement but could prove quite useful.


Poor thieves. At least the magic-users 1st spell will always work. At 1st level they have a 15% chance to pick pockets, find traps, and remove traps. One can see why Black Dougal met such a terrible end. And that 15% chance to pick pockets seems to suggest there's a lot of thieves who lose their hands after a day in the market. I would suggest the following modifications to thief skills:
  • If a thief triggers a trap, he or she gets a saving throw at +2 to avoid the trap. If they fail this save but the trap allows a normal save (for example, a poison needle allows a save vs. poison) they may still attempt that as well. (This is in keeping with the statement "Thieves gain a +2 bonus on saving throws against devices, including traps.")
  • The extraordinary nature of hiding in the shadows and moving silently should be emphasized. I would rule that failure does not automatically result in detection but rather allows for a normal surprise roll.
  • If a thief fails at picking pockets by more than 20% he or she will be detected unless a saving throw is made. (I'd suggest no bonus to this roll.)
  • The rules suggest applying a penalty when trying to pick a difficult lock. I'd also allow a bonus when picking an easier lock.
Given demi-humans can advance without limit in the thief class as well as gain bonuses to certain skills I'd strongly suggest giving some bonus to human thieves. While I'd be tempted to give humans variable bonuses to these skills as well, that adds a little more complexity than I'd prefer. Rather I'd like to have something that reflects some inherent "luck" of humans or make them a little tougher in combat like the Grey Mouser. My inclination would be to give them 1d4+1 hit points per level. It makes their average hit points on par with clerics, with a higher minimum but lower maximum.

Other Classes

The other classes would need less changes in my opinion. I'd not share cleave with other fighting classes. The Lesser Spell Scroll is probably not needed for druids. The interpretations for thief skills should be shared with other classes who share those skills.

Swords and Wizardry Appreciation Day

I wrote this post for Swords & Wizardry Appreciation Day Frog God Games is offering a 25% discount on purchases at their store with the coupon code SWApprDay for today (April 17, 2013 only). I've no affiliation with Frog God Games but would like to share this. The blogroll at 2000 Coppers lists other blogs for this day.

Popular posts from this blog

Jules Verne Translations That Don't Stink

RPG Review: Lamentations of the Flame Princess Weird Fantasy Role-Playing

First Thoughts on Call of Cthulhu 7th Edition

Go Support Golden Age Champions