Pick Your Own Rumour

David Petersen

I've never used rumour tables in my games, and I'm now realising that I probably should. They always seemed a bit dry and artificial to me, but I think they're important for fleshing out the world and giving players more options than just following the first plot hook that the GM introduces.

To make them feel less arbitrary, here's a micro-system that gives players a little bit of control over the rumours they get:

Each rumour has one word HIGHLIGHTED in its text. When PCs are in town, the GM reads out all the highlighted words and each player picks one. Their character hears that rumour. When the PCs return from an expedition, replace the known rumours with new ones and let the players pick again.

Example for my upcoming Mausritter game:

1. The frog village down the RIVER has not been heard from in several months.

2. Monks at the abbey of St. Ninian dug up a mysterious STATUE in their vegetable garden.

3. The ROSE Tombs contain strange treasures of mice who came before.

4. A vicious CORGI has been seen roaming the wilderland.

5. The dreaded cat, Tom-o'-the-Mount, has AWAKENED and left his underground lair.

6. The rat folk are mustering an army in the LAVENDER Hills.

Hopefully this will make the rumour process more interactive, give more hints about the world, and make players excited to return home and refresh the rumours. My gut instinct is that it's better to pick words that are distinctive but not necessarily the most informative. For the table above, FROG, ABBEY, TOMBS, CORGI, CAT and RAT might give the players more control, but it would make the process feel too much like just picking from a menu.

This system is not playtested, so I would be interested to hear feedback about it.

The Monsters' Saves in Knave are Too High! 3 Reasons Why

In Knave, the monster conversion/creation rules state:

Saves: Since OSR monsters usually don’t come with ability scores, assume that monsters have ability bonuses equal to their level, with the corresponding ability defenses.
Example: A typical 4 HD monster would have a bonus of +4 and a defense of 14 in all of its abilities by default, unless modified by the referee.

This makes the monsters' saves too high! Here are 3 reasons why:

1. It's not in line with classic D&D. The monster save tables in classic D&D (I'm look specifically at Old School Essentials here) give the monsters lower chances of saving. It's hard to compare one to one, because the old-school saves vary more and only change once every three levels. But in terms of average save value, Knave monsters are at least equal to OSE monsters, and sometimes have up to a 10% edge. (All this is assuming that the monster is saving against a static DC of 15. If the monster is making an opposed save against a PC's stat, then the variables are multiplied again - but in general the PC's stat is likely to be lower than 15.)

2. It's not in line with the way player characters progress. PCs only get to raise 3 of their 6 stats when they level up, but monsters gain +1 to all stats per level. Now, you might say that PCs who rolled well in character creation can start with higher than +1, which is true. But as you go up the level chart, most PCs will eventually be eclipsed by monsters.

3. It punishes rather than rewards creative play. This is the most important one. By the book, Knave monsters will usually have saves that are at least equal to their armour class, and sometimes even higher. Generally speaking, boring attack rolls will target AC whereas interesting stuff (spells, combat stunts, clever plans) will target the monster's saves. When the players use clever tactics they should have better chances of success, not worse.

The solution I propose is to assume that monsters have saving throw bonuses equal to half their level rounded up. This ensures that even against tough monsters with many hit dice, clever plans have a reasonable chance to succeed.

(P.S. - I didn't miss the phrase "unless modified by the referee" in the above quote. Of course you can and should modify saves where it fits the fiction. A lumbering armoured tortoise might have a crappy DEX save but a great CON save, for example. Nevertheless, I think it's important to have a suitable default that you can fall back on whenever you're not sure.)

Class Hack for Knave

Here are some simple but effective character classes to bolt onto Knave. These are not playtested so exercise your own judgment.

Magic-User
You can use panoply items to focus your magical power. Panoply items are wizardly accoutrements such as a staff, robes, amulet or pointy hat. Each panoply item must be different from the others.

Each panoply item can be used once per day for one of the following effects:
- Restore a spellbook that has been expended
- Force an enemy to reroll a saving throw against one of your spells
- Supercharge a spell as you cast it. Supercharging generally doubles the effect of the spell. The exact details depend on the spell and GM adjudication.

You can carry a number of panoply items equal to your level.

Fighter
You can exert yourself for extra power in battle. Exerting yourself means filling one of your empty inventory slots with "Exerted". The slot must already be empty - you cannot throw away an item and exert that slot on the same combat round. Exerted slots are cleared after a night's rest.

By exerting yourself you can use one of the following effects:
- Gain advantage on an attack or defense roll
- Add 1d6 damage to a successful attack
- Add a stunt to a successful attack

The maximum number of slots you can exert is equal to your level.

Rogue
When purchasing items in town, you can instead opt to fill one slot with a "Useful Item" and note the value of the item. At any time, you can declare what the Useful Item is. It must be a mundane item you could conceivably have purchased for that value or less. It can also be a specific magic item (not a spellbook) that you already own but left in storage.

The maximum number of slots you can fill with Useful Items is equal to your level.

Notes

- Credit: the panoply items are from Benjamin Baugh and Courtney Campbell, exertion is based on the stamina rules from Grave, and the Useful Items are found in Kidnap the Archpriest, Dungeon World, and probably others.

- What I like about these classes is they are simple, but distinctive and relatively potent. I've seen a few class hacks for Knave that add very light classes with just a few minor bonuses. This might seem in the spirit of the base game, but I feel that it's better to have noticeable classes or none at all.

- The Fighter's ability encourages them to carry minimal items "Conan style". I restricted them from tossing items in combat because I thought it would get silly if they were shedding random stuff during every battle (a problem I also have with the "shields will be splintered" rule).

- One issue with this system is that PCs' item slots might get chocker-block at high levels. A 10th level Fighter will have 10 Exert slots, but how many will they actually be able to use if their Constitution is at 15 or 16?

- A lot of Knave hacks seem to interact with the encumbrance system, unsurprisingly since it's at the core of the game. But the more things you load up into encumbrance slots, the more likely it is that Constitution becomes a must-have stat.

- The Rogue's ability might not scale up very well compared to the other two classes. The idea is that at low levels the Rogue is whipping out useful mundane items, but at high levels they are more focused on magic items. The limitation on not bringing spellbooks is rather inelegant, but I think if it weren't there then the Rogue would end up being a bit of a "spellbook valet" for the Magic-User.

- One oddity about the Rogue's ability is that you can use it to "protect" valuable items. Say you have a McGuffin that mustn't fall into the villain's hands. You can leave it in town and bring a "Useful Item" which might be the McGuffin if you need it, but won't be the McGuffin if the villain captures you and searches your pack. I think this is a feature, not a bug, since it lets the Rogue be Grant Morrison-era Batman, always two steps ahead of the enemy. But some people might find it a bit weird.

Spells for Carcosa, or, The Fungi From Yuggoth

The Last of Us

Spells are alien lifeforms native to the jale moon Yuggoth. They exist alternately as memetic information and fungal growths. Spells are not intelligent, but they have evolved a symbiotic relationship with intelligent beings.

In its memetic form, a spell can be encoded as writing, numbers, images, musical tones or many other formats. When an intelligent creature memorises the spell, it grows as a fungus inside their brain.
At the host's command, the spell is ejected from their brain through the nasal cavity and becomes pure magical energy. The actual effects of the spell are, in biological terms, its spores. By proving itself useful to its host, the spell ensures that it will be copied again and again.

Dedicated magic-users undergo training to increase their spellcasting capacity - literally creating cavities in their brains for the fungus to take root.

A O Spare

Rules
- The first time you try to memorise a particular spell, make a Will/Wisdom save. On a success, your brain is incompatible with the spell and you will never be able to learn it.
- You cannot memorise multiple copies of the same spell. Either you have it or you don't. After you cast a spell, you can't memorise it again until the next day.
- Once you have memorised a spell, you can only get rid of it by casting it. 
- When you eat the brain of another magic-user, their memorised spells will be transferred to you. Any spells they recently cast (within the last 10 minutes or so) have a 50% chance of being transferred as well.
- Spells cannot be written in cipher or otherwise concealed by translation. A spell is always a spell no matter what form it has been transcribed into.

Spell Slots
- All PCs begin with spell slots = 0. Gain spell slots through whatever system of character advancement you are using.
- Anyone can memorise any number of spells, regardless of their spell slots.
- When you cast a spell, if the number of spells in your brain exceeds your spell slots, you must roll on the Spell Misfire Table. Roll one die for each spell that exceeds your spell slots, and take the highest die.

Spell Misfire (d10)
1. Spell is cast successfully.
2. Spell works only for a moment, or is at half strength.
3. Spell refuses to come out. Choke on spores for 1 round, but the spell stays in your brain.
4. Spell fuses with another spell in your brain, creating a new hybrid spell which is cast immediately.
5. Spell is not cast, but instead copied into the brains of everyone else in the area.
6. Spell fails and is lost.
7. You give birth through your forehead to a child, half-spell and half-human. Has the same hit dice as you and abilities based on the spell's effects. Roll reaction to see how it regards you.
8. Spell succeeds, but comes out through your face. Take 1d10 damage and permanent facial scarring.
9. Spell possesses your body permanently, while your spirit becomes a spell.
10. Your head explodes as the spell is cast with incredible potency.

Social Mechanics for OSR D&D

I'm looking for feedback on this system, which I haven't playtested yet. The purpose is to give more structure to tense social encounters, such as with monsters in a dungeon. The Patience stat puts a time limit on the conversation, and the 'End of conversation' boxes make the stakes clear to the players. I hope this will make the social encounters feel more 'fair' and thus allow the DM to dole out serious consequences if the social encounter goes poorly.

 This system combines a few ideas from OSR bloggers (I don't think I'm the first to come up with a 'Patience' stat) with dice math based on Apocalypse World moves. Design notes are in italics.

The Conversation Game

This subsystem is for use when encountering unfamiliar or potentially hostile NPCs, particularly in the wilderness or dungeons. It is not necessarily applicable to simpler social encounters such as with a quest-giver or a shopkeeper. The purpose is to create tension and interesting choices for the PCs.

These rules refer to NPC (singular) but can also be applied to a whole group.

Basic Procedure

-       When the NPC is encountered, make a Reaction roll to determine their Attitude and Patience.

-       PCs converse with the NPC. Each significant action in the conversation takes one Conversation Turn.

-       After each Conversation Turn, reduce Patience by 1. When it reaches zero, the NPC ends the conversation in a manner appropriate to their current Attitude.

The Reaction Roll

When the NPC is first encountered, roll 2d6 and add the highest Charisma modifier of the PCs’ group. The total amount is the NPC’s Affinity. The number shown on the highest die is the NPC’s Patience. (This is the only place where Charisma affects the encounter. I did it this way so that a player with high CHA gets to feel rewarded, but everyone can still participate in the conversation without being suboptimal.)

 Affinity represents the NPC’s overall disposition toward the party. Patience represents how long they will converse before taking an action that ends the conversation.

Affinity Track	2	3	4	5	6	7	8	9	10	11	12
Affinity Mod	-3	-2	-1	-1	0	0	0	+1	+1	+2	+3
General demeanour	Hostile	Unfriendly			Neutral			Friendly			Helpful
End of conversation	Attacks; flees		Leaves; pursues own goals even if in conflict with PCs					Leaves; pursues own goals if not in conflict with PCs			Offers alliance or gift

(I envision this track on a sheet of paper in the middle of the table, with a token to show where on the Affinity Track the NPC is sitting.)

 

Conversation Turn

A Conversation Turn is an abstract measure of how the conversation has progressed. One Conversation Turn is something that moves the conversation forward and has a measurable effect. One Conversation Turn might cover:

-       A few sentences of back-and-forth dialogue

-       A single important question asked by the PCs

-       One of the specified Conversation Actions below

Conversation Actions

Improve Affinity: Covers all attempts at flattery, boasting, charming, gift-giving, and anything else where the main purpose is to make the NPC like the PCs more.

Roll 2d6+Affinity modifier:

On a 10+, move one step up the Affinity track. 

On a 7-9, the NPC may be unmoved, or may demand a gift or promise; if the demand is met, move one step up the Affinity track. 

On a 6-, the NPC is irritated and one extra point of Patience is deducted.

Bonuses from +1 to +2 can be applied by giving a gift the NPC likes, making a concession (e.g. throwing down weapons), or using techniques relevant to the NPC (e.g. flattery gets a bonus on a prideful NPC).

Make an Offer: Covers all attempts to trade, swap information, make an alliance, etc.

Roll 2d6+Affinity modifier:

On a 10+, the offer is accepted. 

On a 7-9, the NPC makes a counter-offer. 

On a 6-, the offer is rejected. Move one step down the Affinity track.

Bonuses and penalties from -2 to +2 are applied if the offer is unusually favourable or unusually unfavourable.

Make a Threat: Covers all attempts at extortion, scaring away, forcing submission, etc.

Roll 2d6 (no Affinity mod)

On a 10+, the NPC submits to the PCs' will, but moves 1d4 steps down the Affinity track.

On a 7-9, they retreat or make a counter-offer, and move 1 step down the Affinity track.

On a 6-, they attack or flee.

Bonuses and penalties from -2 to +2 are applied based on the relative strength of the PCs versus the NPCs.

Tell a Lie: Covers all attempts at deception.

Depending on the content of the lie, it can instantly set the NPC’s Affinity track to a higher number (e.g. if you convince them you are an agent of their king, their Affinity is set to 11 or 12).

When you tell a lie, roll for its starting HP based on how plausible it is:

Implausible: 1d6-3

Plausible: 1d6

Airtight: 2d6

Add the NPC’s Affinity modifier to this roll.

(I envision a space on the sheet, below the Affinity track, where you can write down lies and put dice to show how much HP they have left. These lies could potentially continue across multiple sessions and multiple encounters with the same NPC.)

Lies

Each lie told has its own pool of HP (Hoodwink Points). If this pool goes to 0, then the NPC no longer believes the lie. A lie’s HP can never go higher than 12.

-       When the NPC sees/hears something that seems to disprove the lie, then the lie takes damage (1d4 for mildly suspicious, 1d6 for quite suspicious, 2d6 or more for blatant contradiction).

-       When the NPC sees/hears something that seems to be evidence for the lie, the lie is healed (1d4 for mild evidence, 1d6 or more for strong evidence).

-       Whenever a lie takes damage, the PCs have a chance to negate the damage if they make up another lie that explains the inconsistency. Roll HP for this lie separately.

-       When a lie reaches 0 HP, it ‘dies’ and the NPC stops believing it. When this happens, all other lies told to the NPC take 1d6 damage (possibly causing a chain reaction) and the NPC’s Affinity track moves down 1d6 steps.

Status Effects as Choices

While riding my bike today it occurred to me that maybe combat status effects would be more interesting if they were always designed to offer players a choice. For example, instead of "Stunned: miss one round", you could have "Shaken: -2 to all rolls until you take one round to clear your head." That leaves the player with something to think about and do on each combat round.

Below is a table that cross-combines different negative effects to produce various choices.

Effect	Damage	Penalty to rolls	Drop an item	Slowed
Miss a turn	Acid on you: Lose 1hp per round until you take a turn to brush it off.	Shaken: -2 to all rolls until you take a round to clear your head.	Pack torn: Drop a random item each round until you take a turn to tie it back up.	Weighed down: Weird heavy jelly stuck to you. Slowed by 30’ until you take a turn to wipe it off.
Damage		Sucking wound: Take action carefully (-2 to rolls) or forcefully (lose 1d4 hp). Effect continues until you take 10 minutes to bandage yourself.	Heat Metal: All metal items you carry grow red-hot. Drop them or take 1d4 damage per round.	Iron Brambles: You got iron brambles stuck all over you. Move slowly (30’ max) or take 1d6 damage. Take 10 minutes to carefully unwrap yourself.
Penalty to rolls			Wood Growth: All wooden items you carry begin growing green shots. Drop them or take -2 cumulative penalty to all rolls.	Inner Ear Disruption: Sonic attack has wrecked your sense of balance. You can move or you can take an action, but not both in the same round. Ends after 10 minutes of rest.
Drop an item				Heavy Metal: All metal items you carry become progressively heavier. Drop them or be slowed by 60’. (LotFP encumbrance: any regular-sized metal item becomes an oversized item; any oversized metal items add double to encumbrance.)

Race-as-Class vs. Race-as-Race, And a Proposed Synthesis

In D&D-like games, is it better to have Race-as-Class (the idea that non-human races each have their own character class that defines their abilities) or Race-as-Race (where Race is a separate axis to Class)?
Here I will lay out some arguments in favour of either approach and then propose a hybrid.

THE QUESTION OF RACE

Ok, so first of all - the phrase "Race as Class" just makes me cringe every time I write it. It sounds like something you would read about on Stormfront. Most of my IRL players are not familiar with the history of D&D like I am, and I'm pretty sure that if I started talking to them about Race-as-Class then their first reaction would be: "Umm. What?"

Their second reaction probably wouldn't be that much better. The whole concept of Race-as-Class is basically a mechanically codified way of stereotyping the Other: Humans are a diverse group who can choose all kinds of different professions. Non-humans, on the other hand, are all fundamentally homogeneous. The fact that we're talking about wholly fictional 'races' softens the blow a bit, but it still feels wrong to me. And that's assuming your non-human races are wholly fictional and not just thin paint-overs of real world cultures.

Now, I think I can present a logically sound argument why Race-as-Class isn't (or doesn't have to be) stereotyping an entire race. It's like this: character generation is an inherently metafictional mechanic, and its primary concern is not realism but game balance. Race-as-Class is saying "All playable Dwarves are fundamentally the same", not "All Dwarves in this game world are fundamentally the same". If you think it's saying the latter, then by the same logic you would have to conclude that all humans in the game world are either Fighters, Clerics, Thieves or Magic-Users, which obviously isn't true.

So I can make a case for why it isn't racially insensitive, but that won't get rid of the initial gut reaction people are going to have.

THEME: THE SELF AND OTHER

However, there are a few arguments in favour of Race-as-Class which can explain why it has stuck around for so long.

Some people argue that it's necessary to ensure that most of the adventuring party is made up of humans, which is desirable for thematic reasons. I totally get this. D&D to me is fundamentally a game about journeying from the Normal World to a mythic Otherworld. That journey has less weight if all of your player characters are already natives of the Otherworld. But, people still want to play as Dwarves or Elves. Trying to limit the number of non-humans in the party is an attempt at compromise, but ultimately it's a pretty clunky one. A better solution is to make sure your playable non-humans are coded as native to the Normal World. Give them more human-like cultures and psychology. Limit their knowledge of, and power over, the Otherworld: playable Dwarves should have only a small understanding of the Underdark, playable Elves should have little connection to the Fae, and so on.

Also, it's totally fine to say a hard No to a character concept that is definitely going to impinge upon your game's Otherworld. "Can I play as a dragon?" "No - for the same reason you can't play as Yog-Sothoth in a Call of Cthulhu game."

CHARACTER BALANCE

Another reason to use Race-as-Class is for the sake of player character balance. Here's why:
When someone chooses to be a Elf, there's a good chance that they want 'Elf Stuff' to be a major part of their character. So you, as the GM/game designer, want to give them a mechanical package of Elf Stuff to enable that.

If Race is separate from Class in your game, then the Elf Stuff package will have to be balanced against the Human Stuff package. And Human Stuff tends to be fairly weak because humans are the baseline.

If you're using Race-as-Class, though, the Elf Stuff is balanced against Fighter Stuff and Magic-User Stuff. Now you can put a lot of mechanical juice into the role of Elf.

I think this is the clearest reason why Race-as-Class remains popular, and in particular why OSR bloggers have written approximately 10,000 homebrew Racial Classes but barely any homebrew Races-Separate-From-Class. It's a framework that gives the designer more room to play around.

THE GRID

But - there's also a class design reason to prefer Race-as-Race. Having two axes of character choice is a really smooth and easy way to create space for creativity. Four classes and three races creates a grid with twelve boxes to fill, some of them obvious (everyone can picture a Dwarf Fighter) some more mysterious (what exactly does an Elf Cleric look like?)

One of my earliest roleplaying memories is of fiddling around with the character builder app for D&D 4th edition, slotting together combinations of races and classes and trying to come up with a backstory that made sense. (The grid was a lot bigger in that game - probably 20x20 or larger.) There is something inherently appealing about the formula of "X + Y", and it's particularly desirable for OSR play because it makes the characters more distinct without adding very much time to the chargen process.

SYNTHESIS

So those are the major arguments for and against Race-as-Class. With them in mind, I propose a synthesis that will hopefully provide the best of both worlds:
- At chargen, players choose their Race and then their Class.
- Certain Races, if they have powerful innate abilities, can be chosen as both a Race and a Class. Their Class features will then expand upon their racial abilities. In the fiction, this represents a character who has chosen to train their natural powers at the expense of learning any particular adventuring trade.
- If multiclassing is available in your game, then characters can potentially mix and match between a Class-Class and a Race-Class.
This means that PCs are not wholly defined by their race, but they can specialise into it if they want.

Here's a basic example:

1. Choose a race.
Human: You get +10% XP
Dwarf: You have darkvision.
Halfling: You can reroll one die per session.
Elf: You have one spell (from Wonder & Wickedness) that you can cast once per day.

2. Choose a class.
Fighter, Cleric, Specialist, Magic-User: Refer to LotFP corebook (or whatever text you're using)
Elf-as-Class: You may only choose this if your race is Elf. In addition to the spell you get for being an Elf, you gain one extra spell per level and one spell point per level, which you can spend to cast any of your known spells.

The whole 'casting with spell points' thing is untested and would need to be refined, but my aim is just to show how race and class interact mechanically. You can be an Elf Fighter, in which case your main thing is fighting but you also get one unique Elf spell. Or, if you want to lean harder into the Elf part of your character, you can level up as an Elf and get more Elf magic.

Lastly... I've typed the word 'Race' about fifty times in this blog post and it's made me realise that it rubs me the wrong way. In future I might try using a word like 'Kin' or 'Lineage' instead, even though they both sound somewhat clunky.

Is There Logic Behind Old-School Saving Throw Tables?

In my last post I gave some reasons why I like the original D&D saving throws as a general concept. Wanting to make up my own setting-specific saves took me down the rabbit hole of looking over the old books and trying to figure out the logic behind the actual numbers that are given for different saves. I looked over my copies of the Rules Cyclopedia, the AD&D Dungeon Master's Guide, and Lamentations of the Flame Princess.

The short answer is that I think it's all basically arbitrary, though I'd be interested to hear if anyone wants to correct me.

The long answer is this:

How do the classes compare?

- In RC and LotFP, which use race-as-class, the demihuman classes all have superior saves to any of the human classes. I guess this makes sense in RC where the demihumans are sort of like rare Pokemon due to their stat requirements. It doesn't make that much sense in LotFP, but I imagine that Raggi just copied it over for the sake of tradition.
- Halflings have the best saves in the game in both RC and LotFP. This makes sense in the same way as Dungeon Crawl Classics giving special Luck bonuses to Halflings - the logic being, if Halflings aren't supernaturally lucky, then why aren't they all dead already?
- AD&D separates race from class, so it only has four saving throw tables for the four 'core' classes.

- In all three books, there is a common theme of Thieves/Specialists having the worst saves, Clerics being middling, and Fighters starting out poorly but gaining saves at a faster rate until they eventually overtake the other core classes.

How do the individual saves compare?
Mostly, the numbers for each save seem to have been picked more or less at random. There are a few points of intentionality that stick out:
- Breath Weapon is consistently the hardest saving throw to make, across all classes. I suppose this was to make dragons especially threatening.
- Magic-Users often have the best saving throw against spells, implying they have a chance to counter the magic with their own. Clerics have the best saves against Poison/Death Magic, which also seems thematic. I can't see any other places where one class is notably good at a particular save. And even in the case of the Cleric, it's probably going to make a difference of 20% at most, on a roll that is not going to come up all that often over the character's lifespan.

How do the different books compare to each other?
- LotFP has roughly the same numbers as the RC, but they're usually better by one or two points. Maybe that was Raggi's idea of giving his players a break?
- AD&D's numbers are totally different to the RC's. For example, a first level AD&D Fighter has a Poison save of 16, whereas the same save in RC is 12. What's more, AD&D Fighters improve their saves every two levels instead of every three.

It's at this point that I start to doubt that there was any real logic or maths happening behind the scenes when Gygax and other designers were choosing these numbers. I know that some people like to read into AD&D and find hidden nuggets of wisdom, but it seems just as likely to me that Gygax just made up a bunch of new numbers so he could say his game was different from Basic D&D.

Frankly, trying to wrap my head around all these tables feels like more cognitive effort than it's worth for the amount of impact it will have on gameplay. The complexity of the tables feels like a sham - an effort to keep the game's design mysterious by obscuring how banal and arbitrary it really is.

If I were going to write a new, setting-specific table of saving throws, I'd be inclined to just write one table that applies to all characters. However, I would take two lessons from my analysis of the old-school tables:
- You can emphasise a particular danger by making it a harder save (for example, dragon breath.)
- You can emphasise the 'luck' of a particular class by giving them a bonus to saving throws or a separate saving throw table. The latter is a bit more clunky, but I think it's worth it because it will make the player feel like their character is "breaking the game" in a cool way.

Why I Like The Original Saving Throws

Whenever someone sets about updating the classic rules of D&D, one of the game elements most commonly seen on the chopping block is the original list of saving throws. In the Rules Cyclopedia the saves are listed as “Poison or Death Ray”, “Magic Wands”, “Paralysis or Turn to Stone”, “Dragon Breath” and “Rod, Staff or Spell”. This rather haphazard list has been replaced in many modern iterations of the game by the simpler and more realistic “Reflex, Fortitude, Will”.

This change seems to make sense at first glance—but I can’t help feeling that something valuable is lost when the original saves are discarded. Firstly, they have the advantage of being traditional. Some people might see this as irrelevant or even as a negative point. It’s true that there are many cases (in gaming and in society in general) where tradition has to be swept aside to get rid of things that just aren’t working anymore, or never worked in the first place. But when a tradition is still serving its purpose and isn’t causing any serious problems, it’s worth trying to preserve it.

The clumsiness of the original saving throws is also what gives them their charm. If, as is commonly believed, the rules of the game were invented gradually through play, then these five phrases might very well be a list of the first five bad things that ever happened to player characters in a game of D&D. To my mind there is something rather beautiful about the idea that the game still contains a trace of the earliest days of the hobby, like tree-rings encoding the history of the seasons the tree has lived through.

It is not only the past that is encoded in the original saving throws, however, but also the future: they contain an implicit preview of the kinds of dangers that players can expect to face in the game. What better way to present the call to adventure than to say: “here is the number you will need to roll when a dragon breathes fire at you”?

Because the saving throw list can be seen as a kind of primer for the action that will happen in the game, it’s interesting to see how other games with D&D’s DNA have created their own unique saving throw lists. Flying Swordsmen, while hewing fairly close to the original five categories, has renamed them after the five Chinese elements (Wood, Water, Metal, Fire and Earth). And James Young’s extensive LotFP houserules have replaced the “Spells” and “Magic Wands” categories with “Chaos” and “Law” to emphasise the Moorcockian cosmology of his game world. I’d be curious if anyone knows of other examples like these.

I love the idea of different games or campaign settings having their own saving throw lists based on the most common (or most iconic) dangers present in the game world. A campaign based on Greek mythology might have a saving throw list like: “Poison, Gaze Attack, Madness, Transformation, Divine Wrath”. Carcosa’s list could be: “Mutation, Madness, Cosmic Fate, Tech Weapon, Energy Burst”. For a Paleolithic setting: “Fear, Weather, Disease, Nature Spirits, Evil Eye”.