RPG Works and Sundry

Fleeing and Pursuit Through the Eras of D&D

George Dorn — Sun, 22 Jun 2025 00:00:00 -0700

As part of recovering old ideas from earlier editions, to build out my next homebrew OSR campaign, I came across quite a few reviewers mentioning how picking and choosing fights, and knowing when to run away from an encounter were critical skills in surviving early D&D. But that got me wondering... how did running away actually work back then? And how has it change since?

Original D&D (c. 1974)

The very first D&D book doesn't even mention movement, much less pursuit. As a supplement to Chainmail, it assumed those movement rules could be used. In Chainmail, however, you controlled units of troops, not single heroes, so each round consisted of moving units, then missile fire, then artillery fire, then melee exchanges; at this point, fleeing a battle simply means moving away from other units. Some morale checks could force your troops to retreat (or be routed), which could get you 1 1/2 moves away, though with your backs to the enemy (a flanking bonus) and no ability to counter or move again until they spent a turn rallying.

It's not until the 1974 "Blue Book" edition that rules for fleeing are even mentioned:

If combat is broken off, the fleeing party must accept an attack without any return on his part, the attacker adding +2 to his die roll for hit probability, and the armor class of the fleeing party can not include a shield.

Further details are left to the DM; presumably turn-by-turn movement resumes, with the enemies opting to give chase either at the DM's whim or with an occasional spot rule:

If the party decides to flee they may be able to delay pursuit by discarding some of their possessions. Unintelligent montsers will stop to pick up food half the time (roll 1-3 on a 6-sided die) and intelligent monsters will stop for treasure half the time (roll 1-3).

Additionally, in the section on Time and Movement in the Dungeons (pg 9):

Faster speeds can be allowed for charging or a short sprint. If a character is being pursuid, however, he may have to throw away heavy treasure or armor in order to escape.

In this era, the speed of monsters was generally comparable to that of PCs; PCs (and humanoid NPCs) move 240 feet when unarmored, 120 feet in armor, and half if carrying a heavy load (e.g. treasure). Humanoid monsters followed those rules, with smaller monsters like goblins significantly slower than all but the slowest PC. Some monsters, though, would be nigh impossible to escape as a typical adventuring party, forcing dire choices like abandoning slower members or fleeing ahead of the party to improve your own odds. Wizards would be particularly good at this, as would the henchmen, hirelings and retainers likely accompanying you. Some of the particularly frighteningly difficult-to-escape monsters:

	HD	Move
Displacer Beast	6	150'
Dragon	5-11	240' (flying)
Manticore	6+1	120', 180' flying
Spectre	6	150', 300' flying
Wraith	4	120', 240' flying

These are all essentially impossible for a party of adventurers to escape over open terrain, such as a long, open, wide tunnel, leaving only situational escape options (get to the door and close it before getting overrun) or "run faster than the next guy" escape techniques. There's no "rules as written" mechanisms for adjudicating these, but round-by-round combat might work well enough.

Several monsters are quite a bit slower than PCs; oozes, molds and some insects are either immobile or quite slow. This leaves the majority of monsters situationally faster or slower than the party; if an ogre (90') attacks the party while it is loaded down with treasure (60'), it may very well come down to ditching treasure.

Curiously, zombies (120') are significantly faster than skeletons (60') and thus harder to run away from. Odd.

Basic (1977 - 1983+)

Starting with the first Basic edition (sometimes called Moldvay Basic or B/X), movement speeds vary by situation. Encumbrance rules become more complex, as well. Within combat, PCs will move between 40'/round (unarmored) and 10'/round (metal armor and/or heavily burdened). Running speeds are also given, as 3 times normal combat movement. Evasion and pursuit are far more detailed:

EVASION: Sometimes one side wishes to evade (avoid) an encounter. If the evading side has a faster movement rate than the other and combat has not yet begun, evasion is automatic as long as the evading side is not forced to stop.

PURSUIT: If either side wants to pursue the other, time is counted in rounds, and both sides are RUNNING. Monsters will chase evading characters only as long as the characters are in sight.

Dropping things to distract pursuers also carries over from the blue book.

Once combat has started, a FIGHTING WITHDRAWAL (1/2 normal movement) or full RETREAT (above 1/2 normal movement) are the main ways to flee, and must be declared before each combat round.

RETREAT: Any movement backwards at more than 1/2 the normal movement rate is a retreat. If a creature tries to retreat, the opponent may add +2 to all "to hit" rolls, and the defender is not allowed to make a return attack. In addition to the bonus on "to hit" rolls, the attacks are further adjusted by using the defender's Armor Class without a shield.

While it isn't said explicitly, the retreat option seems to be one you can use at your running speed. It also appears that if your side wins initiative, it is possible that your retreat results in no attacks by the enemy. If your side doesn't win initialive, however, you've already declared you are retreating, so the extra risks above apply.

Monsters are listed with combat movement speeds, and as with the blue edition, most are between the speed of an unencumbered party and a heavily encumbered party. The same extra-fast monsters produce inescapable encounters (outside of situational options), along with a couple new ones.

The Expert book (the X of B/X) adds rules about pursuit in the wilderness, providing a table of random evasion chance, in which a smaller party trying to evade a larger party has increasingly good base chances, and some extra rules about the order of events:

Pursuit: If the party fails to evade, they must fight or move away in a random direction (no mapping). If the other group is faster, there is a 50% chance the party will be caught. If the party is not caught, they may try to evade again. Repeat the procedure until the party evades or is caught.

Rules Cyclopedia mostly restated these rules (sometimes at great lengths) but does add a couple wrinkles that could significantly change the flow of combat:

The Evasion table is formalized to apply to all encounters, not just wilderness encounters.
Initiative is determined every round during pursuit, with each side moving accordingly. This could result in one side moving twice in a row, if they lost initiative one round and won the next.
Monsters make a morale check every five rounds of the pursuit or give up.
Fighting Withdrawal is changed significantly; now it isn't really a means of escape, just an opportunity to move the fight around at 5' per round or offer a truce.
Retreat has been rephrased to make it clear that you can't use your running speed unless you start a movement phase without being in hand-to-hand combat; the retreat maneuver uses your encounter speed instead. On the other hand, the order of events has changed a bit:

Any enemy attacking him later in the combat round (that is, either an enemy who followed him during the enemies' movement phase or an enemy attacking with a ranged weapon) receives a + 2 attack roll bonus this round.

AD&D First Edition (1979)

Initially, First Edition had very little to say on the subject of fleeing an encounter. The Player's Handbook notes:

If the party is fleeing, all movement -- excluding encumbered movement, is 10 times faster, so each move takes only 1/10 of a turn, or 1 round.

This seems important, but that 10x faster is just the combat movement speed; anybody moving in combat does so at 10x the speed they normally do. Later, in the Melee Combat section:

fleeing means as rapid a withdrawal from combat as possible; while it exposes the character to rear attack at the time, subsequent attacks can only be made if the opponent is able to follow the fleeing character at equal or greater speed.

The Dungeon Master's Guide explains avoiding encounters in some detail (pg 63), but then leaves much of it up to the DM:

It is always possible to flee from an undesired confrontation if the other party is surprised. It is never possible to flee from an encounter where the opponent party is in strinking range. A party can always flee an encounter if it gains the first initiative.

It then lists some factors the DM should consider in determining whether evasion is successful, such as monster motivations, relative speeds, deterrents, etc, and gives a flat 50% chance of successful evasion in all other cases. Later, in "PURSUIT AND EVASION OF PURSUIT" it lists a different set of criteria for whether pursuit happens, and whether evasion is successful is detailed in a rather complex table. If the pursuers are slower than the pursued, evasion happens at 100' of distance, or 50' of distance if line-of-sight was broken, or after 5 rounds if the pursuer is not gaining. If they're the same speed, it's 150' of distance, or 80' with broken line-of-sight, or 1 turn. And if the pursuer is faster, 200' and out of sight, or the pursuer is exhausted.

How a party is supposed to get more than 200' away from a monster that is faster than them is left as an exercise for the player, but an entire page of modifiers is suggested, from distracting monsters with food or treasure, placing barriers, taking a forking path out of sight of the pursuer, etc, but this mostly boils down to DM's judgement and though it isn't stated outright it seems Gygax expects groups to continue round-by-round actions until the chase is over.

In the combat section, the rules for fleeing after combat has started is explained much more simply:

Breaking Off From Melee: At such time as any creature decides, it can break off the engagement and flee the melee. To do so, however, allows the opponent a free attack or attack routine. This attack is calculated as if it were a rear attack upon a stunned opponent. When this attack is completed, the retiring/fleeing party may move away at full movement rate, and unless the opponent purseus and is able to move at a higher rate of speed, the melee is ended and the situation becomes one of encounter avoidance.

So running away gets you one or more free atacks (at +6 and with no shield of Dex modifier to AC), and if the monster is faster than you and chasing you, it gets to keep attacking.

While it seems like Gygax wanted to facilitate dramatic chases through dungeons, the rules-as-written don't seem to support it. If you're faster than the monster you're running away from, in the best case (surprise or first initiative) you can just run away, and at the worst your front ranks suffer some attacks. But if you're slower? Not only is it nearly impossible to achieve the distances and conditions needed to end pursuit, those conditions are even harder than normal. Worse yet, if combat has already started, the monster can simply continue following and attacking.

Given this, which monsters are a dire, save yourself, drop-everything-and-run encounter? Most PCs will be moving between 120' and 60' per combat round, determined mainly by armor, with most fighting classes around 90'. Most humanoid monsters will be comparable, but numerous beasts exceed 120', and most flying monsters move at 180' or more. Notable new hazardous encounters:

	HD	Move
Ghast	4	150'
Banshee	7	150'
Intellect Devourer	6+6	150'
Lamia	9	240'
Leucrotta	6+1	180'
Nightmare	6+6	150' (360' flying)
Rust Monster	5	180'

For the vast majority of encounters, armored PCs are simply going to be unable to flee from a combat once it has started, so long as the monster wants to pursue. For many, many more, even unarmored PCs will be easy to run down. I don't think Gygax meant for these rules to be used, but for rulings to override them almost all of the time.

At least Gygax fixed the Zombies vs Skeletons problem. Skeletons now move 12", while Zombies move 6".

AD&D Second Edition (1989)

Much of Second Edition works the same, at least for a single PC trying to flee a combat:

A withdrawal is up to 1/3rd of a PC's normal movement rate, not 5'.
A situation that likely caused an ad hoc ruling is codified: if a front rank fighter withdraws, the rest of the rank can block the opponent from following.
The free attacks a monster gets on a fleeing PC are clarified to be instantaneous and not count against their other attacks.
Once a PC has fled, the monster can give chase on its next turn. Another rule allows characters to attack only if they've moved no more than half their movement, though charging triples that distance with some additional risks, making it simple for most monsters to continue closing with and attacking a fleeing party.

However, rather than trying to rework the hodge podge of First Edition's pursuit and evasion rules, Second Edition simply omits them (aside from a rather extensive rule set for Aerial Combat). There are no rules to adjudicate what happens when one party immediately tries to flee when an encounter is sighted, apart from a brief discussion in the Player's Handbook:

Your character's success at evading capture will depend on movement rates, determination of pursuit, terrain, and just a little luck.

Even less helpful than Gygax.

D&D Third Edition (2002)

Third Edition was the first to codify combat actions to such a degree that the ability to flee was simply an emergent property of the round-by-round rules. At the same time, the need to flee from a random or unbalanced encounter has never been lower. Outside of tactical errors, most parties can assume that it is safe to start any fight, and only consider fleeing if it starts to go wrong.

Movement

Each character has a speed measured in feet. You can move that distance as well as attack or cast a 1-action spell, and you can move before or after attacking or casting.

You can also make a double move, which lets you move double your speed, or a run, which lets you move quadruple your speed.

Similarly, the risks for fleeing from an ongoing combat are made explicit by the Attacks of Opportunity rules:

When you move in or away from an area that an enemy threatens, you provoke an attack of opportunity from that enemy.

Player characters can also avoid the penalties of running from combat by performing a Withdrawal action, essentially allowing a safer double move:

Withdrawing from melee combat is a full-round action. When you withdraw, you can move up to double your speed. [...] visible enemies do not get attacks of opportunity against you when you move [out of your starting square].

An opponent could then charge on their next turn, granting them double movement speed (in a straight line) plus an attack, so withdrawing is often only capable of allowing a full party to flee if the situation permits.

The Dungeon Master's Guide spells out the general process of adjudicating pursuit:

Evasion and Pursuit

In round-by-round movement, simply counting off squares, it's impossible for a slow character to get away from a determined fast character without some sort of mitigating circumstances. Likewise, it's no problem for a fast character to get away from a slower one.

When the speeds of the two concerned characters are equal, there's a simple way to resolve a chase: [...] have them make opposed Dexterity checks to see who is the faster over those rounds.

In the case of a long chase, an opposed Constitution check made by all parties determines which can keep pace the longest.

The Dungeon Master's Guide II offers some details for those mitigating circumstances, to make a chase encounter more interesting, but the round-by-round procedure doesn't change.

As for the movement speeds of characters, they could range from 15' (for short folk in armor) to 40' (for unarmored characters with certain class advantages, like the Barbarian's fast movement.) Most low-level monsters are within that range, especially humanoids, though top speeds for other monster types tend to grow a bit with hit dice. Those speeds are expected to be balanced for level-appropriate parties having additional mechanisms for fleeing, such as teleportation or haste.

D&D Fourth Edition (2008)

Apart from returning to movement rates measured in miniature scale (given in squares), much hasn't changed from Third. Fleeing from melee usually results in Opportunity Attacks, and rarely puts you far enough away to avoid enemies charging and attacking the following turn.

A new Move Action, Run, was added, allowing for faster movement while still leaving a Standard Action available, with some additional risks. Taking the Run move action grants you an extra 2 squares of movement (compared to a PC's normal movement of 5-7 squares), at a cost of -5 to attack rolls, provoking Opportunity Attacks, and granting Combat Advantage to enemies.

Additionally, running away from an encounter could be made into a Skill Challenge, such as the Urban Chase example given in the Dungeon Master's Guide. This is entirely optional, but gives players a menu of applicable skills (Acrobatics, Athletics, Perception, Streewise) to roll to collect enough successes to resolve the encounter. This is, notably, the first time skills are used directly in resolving a chase scene abstractly. Curiously, this approach does not make any use of the various creatures' movement speeds. There's also not much guidance given for when or how to switch to this system when PCs opt to flee an ongoing combat.

Similar to Third edition, lower-level monsters tend to be slower than or similar speed to PCs, and this scales up with level as PCs also gain more movement powers. Unlike Third edition, nearly every class will gain access to such abilities, due to the Powers system shared by every character.

D&D Fifth Edition (2014)

Much of Fifth Edition's movement rules are a restatement of Third Editions. For example, instead of the "Combat Withdrawal" maneuver allowing slow retreats without free attacks against you, Fifth Edition has the Disengage action. This is treated the same as an action ("standard action" in Third), which means you can move your normal encounter speed safely.

Similarly, the "double move" action has been replaced by the "dash" action, which converts an action into an additional movement but otherwise behaves nearly the same. However, many classes gain the ability to perform a Dash using their once-per-round Bonus Action at relatively low levels, such as the Monk's Step of the Wind or the Rogue's Cunning Action, resulting in some characters being able to safely Disengage and move double their speed in the same round.

Pursuit and evasion are detailed in more depth in the Dungeon Master's Guide:

Chases

Strict application of the movement rules can turn a potentially exciting chase into a dull, predictable affair. Faster creatures always catch up to slower ones, while creatures with the same speed never close the distance between each other. This set of rules can make chases more exciting by introducing random elements.

The round-by-round resolution process is mostly unchanged, with each participant having an action and a move on each turn, resulting in a running combat. Some additional optional rules are suggested, such as controlling the number of times a Dash action can be used with a Constitution check, preventing Opportunity Attacks between participants, and providing a Stealth mechanic to escape a chase by hiding.

The Dungeon Master's Guide also suggests adding Chase Complications to Urban or Wilderness chases, listing a bunch of mini-challenges in a random table that can drastically change the progression of a chase, such as having to treat a crowd as difficult terrain, or being blinded by blowing dust, both avoidable with a skill check.

Summary

Early editions of the game emphasized the need for players to pick and choose fights and flee from ones they cannot win, and there was no expectation that encounters would be balanced. At the same time, there were no detailed systems for handling fleeing combat or escaping pursuers, leaving this aspect of the game up to spot rulings and house rules. Later editions provided more systems and options for escaping combat, while simultaneously reducing the need by adding an expectation of balance to the majority of encounters.

Antisemitism and the Origin of the Lich's Phylactery

George Dorn — Wed, 03 Feb 2021 00:00:00 -0800

Accusations of D&D's bigotry are raised from time to time, often with merit. From dwarves and orcs carrying racist baggage from Tolkien to calling monster clerics "shamans" to Oriental Adventures to naming a class "Warlock" to even using the word "race" instead of "species".

And don't even start on the origins of the half-orc.

But one that keeps popping up as an example of malicious antisemitism is the lich's phylactery:

"For a more explicit example of villainizing Jewish objects and folklore, take the Lich, a powerful being who has cheated death by becoming something unholy. Liches separate their souls from their bodies and put them in special places called a “phylacteries” so they can never die [...] [Gygax] made the choice that an undead wizard king would keep his soul in something that Jews use for daily prayers."

-- https://www.heyalma.com/dungeons-dragons-has-an-antisemitism-problem/

"Gygax straight up took the word phylactery from Judaism and made it something used for evil."

-- https://technoskald.me/2020/06/28/christian-mythology-in-gaming/

"the antisemitism in [D&D] was deliberate and malicious [...] the creators of [D&D] thought it would be a good idea to name the evil undead spell casters evil box with their soul in it, after a Jewish ceremonial object."

-- twitter thread I'm not linking

What Is A Lich?

A lich is a powerful, intelligent, undead monster, the result of a powerful spellcaster turning to necromancy to unnaturally extend their age.

Over the years, liches have become a classic, go-to villian monster, from Acerak in "Tomb of Horrors" (1974) to Vecna to the obscure Githyanki queen Vlaakith. They can live forever, control other undead, cast powerful magic, and are generally well-suited to being Bond villian masterminds. They've also changed and evolved continuously in the nearly five decades since their creation, until what we have today. As they appear in Fifth Edition and Pathfinder, along with dozens of derivatives and retroclones:

liches are villianously evil undead
liches use necromantic magic to bind their own souls and (sometimes) the souls of others
liches use a specific item to do that magic, an item called a "phylactery"

What Is A Phylactery?

In modern usage, outside of Dungeons and Dragons, "phylactery" most commonly is used as another word for "tefillin" - a small box containing Torah verses that many Jews wear tied to their arms and heads while praying. "Phylactery" itself is from an ancient Greek word (phulakterion), though that word was likely an ignorant misunderstanding; it literally means "amulet", particularly protective in nature, but it referred even then to the Jewish practice. Ironically, many current D&D players make that same mistake, though Gygax clearly understood the word when he adopted it:

Phylactery -- An arm wrapping with a container holding religious writings, thus a form of amulet or charm.

—Dungeon Master's Guide (First edition, 1979)

So at face value, we have a powerful, unholy creature using an item of Jewish religious significance to perform evil, necromantic magics. Not a great look.

How did we get here? Let's follow liches and phylacteries through the history of D&D.

Supplement 1: Greyhawk (1976)

The first appearance of the lich appears in the first published supplement to the $5 pamphlet edition of D&D, often referred to as the "Brown Book" or "Original" edition. Like all monsters of the era, their description is brief. It makes no mention of phylacteries or soul-trapping:

LICHES: These skeletal monsters are of magical origin, each Lich formerly being a very powerful Magic-User or Magic-User/Cleric in life, and now alive only by means of great spells and will because of being in some way disturbed. A Lich ranges from 12th level upwards, typically being 18th level of Magic-Use. They are able to employ whatever spells are usable at their appropriate level, and in addition their touch causes paralyzation, no saving throw. The mere sight of a Lich will send creatures below 5th level fleeing in fear.

Original D&D was less interested in flavor text; it was not paricularly interested in explaining the origins of monsters in great detail, instead spending its few words on combat and other mechanical details.

That edition (in the main pamphlet, 1974) also contained Magic Jar, a 5th-level Magic User spell that uses any inanimate object to house the life force of the caster indefinitely and allows attempts at posession. While not directly tied to lichdom (and castable by any Magic User of 9th level or above), this effect will make additional appearances in later editions.

AD&D First Edition (1977)

Here we have the first appearance of phylacteries, both as treasure items usable only by clerics, and in posession of liches.

A lich exists because of its own desires and the use of powerful and arcane magic. The lich passes from a state of humanity to a non-human, non-living existence through force of will. It retains this status by certain conjurations, enchantments, and a phylactery.

No further explanation of the phylactery is given; it's flavor text, not mechanical, and hints at the recurring theme that liches use both religious and arcane magic to "live" forever. Most notably, liches have still not gained any specific powers from it; they do not return from the dead once destroyed, and no mechanical effects for stealing or destroying the phylactery is given.

In the Dungeon Master's Guide, in addition to the glossary entry above, Gygax uses the term for a religious item usable only by clerics and introduces three magical phylacteries as treasure items:

Phylactery of Faithfulness: warns the wearer if they're about to take an action that would offend their alignment or god
Phylactery of Long Years: slows the aging of the wearer to 75% of normal (unless cursed)
Phylactery of Monstrous Attention: a cursed item, causes the wearer to draw the attention of hostile creatures of the opposite alignment

(A side note on probability: while two of these are meant to be good and one is meant to be bad, ~82% of the phylacteries found at random are good, roughly on par with finding good vs cursed items in general.)

This paints a very different picture of Gygax. He did not adopt the phylactery as the item responsible for the lich, and the association between liches and phylacteries was a single word of flavor text.

Dragon Magazine #26, 1979

In "Blueprint For A Lich" (pg 36), Len Lakofka finally describes the ritual by which spell casters become liches:

The lich needs these spells: Magic Jar, Trap the Soul, and Enchant an Item, plus a special potion and something to "jar" into. [...] To get out again, the MU/Cleric must have his (or another’s) recently dead body within 90 feet of the jar.

Lakofka envisions the ritual as the creation of a special version of Magic Jar, and while this does seem to be the origin of some of the mechanical features of the lich's phylactery of later editions, he never refers to the it by this name. In fact, it can be any item of sufficient value; gems and jewelry are given as examples.

Far more detail in put into the mechanics of how the jar works (it's far more complex than later versions) and the preparation of the potion fed to the human to be sacrifice (requiring two dead infants slain in extremely and oddly specific circumstances and a virgin, though gender is not specified).

Basic D&D (Red Box, B/X, BECMI and Rules Cyclopedia)

Mentioned here only for completeness, the lich of the "Basic" branch gains some additional powers (mainly summoning other undead) and variations (mainly levels of Cleric or Magic-User), but they never gain the power of item-based immortality that their counterparts in the AD&D line do.

Also, the word 'phylactery' never comes up in the core rules; while amulets, scarabs, talisman and medallions are mentioned, no phylacteries ever made the jump, either as treasure items or as mechanical devices related to a monster.

AD&D Second Edition (1989)

1989's Monstrous Manual may be the first time the life-extending necromancy mechanics and the word "phylactery" appear together, 13 years after the creation of the Lich and long after Gygax's departure from TSR.

In all cases, a lich will protect itself from annihilation with the creation of a phylactery in which it stores its life force. This is similar to a magic jar spell. In order to ensure the final destruction of a lich, its body must be wholly annihilated and its phylactery must be sought out and destroyed in some manner.

For the first time we have the lich's phylactery as a major plot device, an item you must find and destroy to defeat the evil villian. Yet at the same time, the definition of the word in the Lich's description seems to have gotten lost:

The phylactery, which can be almost any manner of object, must be of the finest craftsmanship and materials with a value of not less than 1,500 gold pieces per level of the wizard.

The three magical treasure phylacteries are also retained in this edition, and in various magazines and modules added another three, but the glossary definition is lost, along with the specifics from the treasure phylacteries. One, the cursed and rare Monstrous Attention phylactery, refers it as an "arm wrapping" and the other two explain even less.

Later, the somewhat obscure Encyclopedia Magica (1995) collects these and clarifies the explanation from the 1e glossary:

Phylacteries are talismans that are usually worn on the forehead and wrist, but are occasionally wrapped about the upper arm or the thigh. They contain small black boxes of prayers.

D&D Third Edition (2002)

WotC turned the Lich from monster into a template (introducing some variety and flexibility) and simplified some of the mechanics. Not only was the phylactery retained, the definition in the lich's description was adjusted to again reflect the real-world meaning:

The most common form of phylactery is a sealed metal box containing strips of parchment on which magical phrases have been transcribed. [...] Other forms of phylacteries can exist, such as rings, amulets, or similar items.

3e ditches two of the treasure phylacteries, retaining the Phylactery of Faithfulness and adding the Phylactery of Undead Turning, an important item for a cleric specializing in destroying powerful undead (such as liches.) It also describes the Phylactery of Faithfulness more carefully:

This item is a small box containing religious scripture affixed to a leather cord and tied around the forehead.

D&D Fourth Edition (2008)

This edition retains the lich's phylactery and mechanics:

The phylactery, [...] usually takes the form of a sealed metal box containing strips of parchment on which magical phrases have been transcribed in [the lich's] blood. [...] Other kinds of phylacteries include rings and amulets [...].

4e dropped all of the beneficial phylacteries. This was almost certainly not with deliberate antisemitic intentions, but rather a side effect of the overhaul of the treasure system. Many items that did not grant simple combat bonuses or specific skill bonuses for use in the extremely mechanized non-combat skill challenges didn't make the cut.

Pathfinder (2009 and 2019)

1st Edition Pathfinder's lich (and its phylactery) is a direct copy from D&D 3e. The of the treasure phylacteries, the Phylactery of Faithfulness is retained, but the Phylactery of Undead Turning is not. Instead it adds Phylacteries of Positive Channeling and Negative Channeling, improving damage to undead and healing to living, and vice-versa.

2nd Edition Pathfinder's lich retains the 1st Edition description and adds an additional power, Drain Phylactery, allowing the lich to cast spells from the item.

The Phylactery of Faithfulness is the only treasure phylactery in 2nd.

D&D Fifth Edition (2014)

Curiously, 5e liches are strictly wizards, removing the religious connection that made the phylactery even vaguely suitable. The lich's phylactery behaves much like 2nd through 4th editions:

A lich is created by an arcane ritual that traps the wizard's soul within a phylactery. Doing so binds the soul to the mortal world, preventing it from traveling to the Outer Planes after death. A phylactery is traditionally an amulet in the shape of a small box, but it can take the form of any item possessing an interior space into which arcane sigils of naming, binding, immortality, and dark magic are scribed in silver.

An additional evil twist is added, however:

A lich must periodically feed souls to its phylactery to sustain the magic preserving its body and consciousness.

All of the good phylacteries are also missing, probably the result of treasure being cut back drastically, but leaving no positive association with the word. For the first time in D&D history, a popular edition of the game says, perhaps accidentally, that a phylactery is only for liches!

Summary

When Gygax first introduced the word, he clearly understood its contemporary meaning. But it wasn't an item exclusive to liches nor an item of particular importance to the lich. Liches were as religious in origin (however twisted) as they were magical, but the significance of the lich's phylactery, as a macguffin you must hunt down and destroy to beat the villian of the game, simply wasn't there until long after Gygax left TSR.

In short, Gygax's liches just don't work the way later liches do.

Later writers, probably out of ignorance of the meaning of the word, combined the Magic Jar effect with the word they didn't know from Gygax's writings to create that macguffin. But even then, the phylactery was a religious item for good clerics, too. It was awkwardly appropriated and ill-suited to the polytheistic cultures of almost all of the game's settings, certainly, but not locked to a single villianous purpose.

But a newer player, with the 5th edition as their only Dungeons and Dragons experience, will never see the word outside of its use with liches. The only time it'll appear as treasure is as a quest item for defeating not just a lich sustained by it, but a lich that in turn sustains the phylactery by feeding it souls.

This is obviously a recipe for a bad impression, and confused arguments between new players who see only this version of the word's usage and older players who recall a longer and more complex history.

How's Your Campaign?

George Dorn — Fri, 31 Jan 2020 00:00:00 -0800

Alex Schroeder writes, How's Your Campaign? And since my 3.5 game has atypical longevity, it might be worth answering.

1. How many sessions have you been playing, more or less?

About 200. Ballpark.

2. How long have you been running this campaign?

Very nearly 10 years. Our first session was in April of 2010.

3. Have you had long breaks? If so, how did you pick it up again?

We frequently have scheduling issues around the holidays and sometimes for a month in summer. Our longest breaks were probably no more than six weeks, though.

4. How many people are at the table when you play?

This has varied as people have come and gone, but we've ranged from 5-8 people at a time, myself included.

5. How many characters are in the party when you play?

Currently we have 5 primary PCs, 1 semi-retired PC and 4 followers (two animal companions, a familiar, and a paladin follower of the cleric). There's often one or two NPCs that bamf in when I remember they exist. There's a lot of history to remember.

6. How many players have you had in total over that time period, not counting guest appearances?

Eight regular players have appeared over the years.

7. Have you had guest appearances? How did it go? Did you gain regular players that way?

We've had a few. One of the current regular players started as a guest who would play whatever NPC was available, or the archmage's familiar (to great effect).

8. What have the character levels been over time?

Most of the PCs started at level 1, and they're now 19 and 20. Along the way, we had to decide what to do upon reaching level 21; as a group we decided to nix the Epic Level Handbook and just allow multiclassing and prestige classes.

9. What classes did the players pick? Did you add new classes over time?

The original group was a human wizard, human rogue, human druid, human paladin and two fighters (one human, one dwarf). The paladin died at a level before Raise Dead was available (the only permanent death) and that player returned as a half-ogre cleric of Kord until she moved away. The dwarven fighter's player retired, a human cleric and paladin of Pelor (the paladin is technically an npc follower) and gnomish bard replaced those two players. Eventually the bard's player also retired (as did the original druid's player) and an elven druid joined relatively recently. The original druid comes back occasionally.

For the most part, everybody either stayed with their original classes or took obvious prestige classes. The wizard became an archmage, the human fighter became a devoted defender (ported over from 3.0) to better protect the archmage, and the rogue recently started taking levels in Shadowdancer.

We've also had two short side campaigns related to the main one; these have allowed for novel party compositions and trials of other systems. The first was to welcome the druid back from retirement and was for two sessions; three players and myself all played druid characters from a stack of weird druid pregens. There was an alienist, a Swanmay, a weretiger lecher nudist named Hunter and the returning druid.

The second was an all-warlock party of three using the 5e rules, set in the same dungeon the main party was exploring but in an alternate universe and at a different time. It was three sessions, had accelerated character advancement (to see how 5e played) and required each PC take a different warlock specialization.

10. Tell me about some adventures you ran over that time that I might enjoy hearing about?

Ah, game stories. So easy to reminisce about with players from the game, so easy to bore third parties to tears.

I'm working on a summary of the various modules and snippets of published works I've gleaned adventures and ideas from, but until the campaign is over readers will have to be satisfied with some quick highlights (polled from my players):

In Castle Caldwell, the party released a BBEG lich-deathknight-avatar-of-Vecna from a tomb in the basement were nearly TPK'd in the process. They got better; it was a tone-setting plot device more than a real encounter.
The dwarf fighter developed a reputation for delivering massive critical hits from a composite bow, specifically when fighting dragons.
The first druid had moments of pure metal awesomeness: chasing down a behir fleeing through a desert of thorns while shapeshifted into a giant eagle and calling lighting down from the heavens; leaping off a tower as a bobcat onto a gargoyle nobody could hit, knocking it to the ground and getting a grapple going; summoning a unicorn in a fight against a lesser lich only to have it impale the lich with its horn and destroy it...
The Red Hand of Doom module delivered a few noteworthy set pieces the players remember fondly: defending the city against multiple waves of varied attackers (red dragons, an assassin, a Tiamat-blessed hobgoblin half-dragon leading a small army); fighting running battle against a super-intelligent, extremely powerful blue dragon, in the Fane of Tiamat, using the small tunnels to evade its lighting breath and spells; the half-ogre cleric of Kord, enlarged and with a stack of other buffs, killed the avatar of Tiamat (at the end of Red Hand of Doom) with a punch.
A rematch with the BBEG some 15 levels later, in the fountain where all souls are created and sent to those being born... The human fighter, overconfident with his adamantium sword and Improved Sunder, went for the lich's magic staff, which turned out to be a nearly-fully-charged Staff of Power. It exploded, and I played the outcome by-the-book, explaining how the whole thing worked before announcing the damage (most of the PCs killed instantly) and rolling the lich's 50% chance of destruction (vs forced travel to another random plane) out in the open. The dramatic moment saw the lich survive, but he hasn't made another appearance since.

11. Have the rule changes over that time? Do you maintain a house-rules document?

This is a weird case where we're playing almost entirely rules-as-written. I did this to put as much power in the hands of the players (to make bad decisions as much as good) without relying on last-minute rules changes or favorable rulings to save them. This worked great up until about level 8; 3.5's complexity scales exponentially with level and so trying to stay on top of all of the rules instead of coming up with my own intuitive rubrics has been incredibly punishing.

If I had it to do over, I'd play by the E6 rules instead.

As for house rules, it's mainly a whitelist of published options; a ton of 3.x publications suffered from power creep. Many of the basic classes are made entirely obsolete by later classes that are as good or better than the original class in every way. Take Fighter, for example; nobody should play a Fighter if all of the splatbooks are on the table. Similarly, prestige classes are carefully selected to avoid power creep, especially when they build on already-powerful base classes (the "Tier 1 and 2"s - mostly spellcasters).

12. Has the setting changed over time?

We started in a sort of generic fantasy setting, creating just as much world as I needed to fit parts of modules I was borrowing and the big plot ideas that were developing. Over time, this grew into Greyhawk, as many modules were either from there or had good documentation about where to place them in that world.

Planar travel expanded on this; Greyhawk really only details the Prime Material Plane's surface and a little bit of the Underdark, so once the party started traveling to other planes there were whole other worlds to explore.

Regarding the footnote (adding cultures and places later, after the game's started) - the campaign has always been developed just-in-time; I'm generally not a fan of extensive worldbuilding ahead of time. I'll document some broad ideas, but all of the really interesting elements (to me, anyway) were developed during play, either as a response to players taking an interest in something (often an underdeveloped item from a published work), me having sudden inspiration, or both.

13. How much in-game distance did the party cover, how big is the area they have visited?

On Greyhawk, the party started roughly in the Yeomanry (their starting town and Castle Caldwell were retconned there), and the setting for Red Hand of Doom (the Elsir Vale, about 250 miles across) got roughly shoehorned into the southern half of the Kingdom of Keoland. Later modules would take them to Istivin, and as they became even better at rapid forms of travel (Shadow Walk and later Windwalk being favorites) went as far north as the Land of Black Ice on a whim (they needed diamonds for Resurrection spells and heard there was a dragon up there) and deep into the Dry Steppes to the west to find an ancient Baklunish ruin.

They've also been to the Underdark in several places, from a drow outpost to a svirfneblin village to a city of ghouls.

Once planar travel became an option, they've been all over:

A trip to the Outlands to join a battle, defending Plague-Mort against demons. Or was it devils? (This was a one-shot send-off for a player and her cleric of Kord).
A longer trip to the Outlands to find their way through Plague-Mort to Pazunia, into the Demonweb pits, and then into a tiny, frozen demiplane.
An excursion into Elysium, specifically Pelor's Fortress of the Sun on Thalasia, to meet with Pelor and Vecna to discuss the whole avatar of Vecna problem. (Turned out it's a different Vecna's avatar...)
A trek to Pandemonium (specifically Phlegethon and Agathion) to free an imprisoned demoted god. A short stopover at Windglum was rather entertaining.
A trip to a demiplane of dreams, though this wasn't under their control.
An excursion of the Positive Energy Plane (as mentioned above) to stop the avatar of Vecna from doing... something? to that fountain of souls.

Finally, the PCs have spent quite some time in the Far Realm, after following a trail of clues to Tovag Baragu, a network of portals leading to Vecna's transdimensional temple (c. mid-90s Second Edition). Most recently they've left their own universe and the Far Realm located near it, and hitched a ride with a space whale to the Outer Far Realm. Way, way, way off the map.

14. Have you used proprietary setting books? Like, could you publish your campaign or would you be in trouble if you did?

With some very heavy editing, perhaps, but it borrows very heavily from parts of so many published works it's probably not worth it.

Brightwater: the Meatgrinder

George Dorn — Mon, 21 Oct 2019 00:00:00 -0700

Over the weekend, I hosted a one-shot kickoff funnel game. I used the module Meatgrinder, a free zine-style adventure for Dungeon Crawl Classics.

Systems

The rules are cobbled together from Basic (especially B/X) and Dungeon Crawl Classics. The idea of a funnel is from DCC; a funnel is a module for many low-level characters, often level 0 commoners, in which most die or are grievously wounded. The few survivors become first-level adventurers. Despite being a relatively recent invention, a funnel module provides several helpful features for starting OSR games. They're the opposite of modern D&D; PCs are nearly powerless, believable people with ordinary lives. Funnels set the tone of danger and desperation of low-level adventuring and are story generators, creating backstories for heroes in-game. Combined with a simple character generation system, funnels can be as low-friction as starting a boardgame, except heroes emerge from the other side.

Stats are the usual six from B/X, rolled 3d6 in order with B/X's non-linear stat mods (e.g. 16-17 is +2, 18 is +3).

I'm not using skills, at least not yet; instead I'm borrowing DCC's occupations. Every commoner has a job of some sort, providing just enough information to know what the character might be good at. When a roll is needed, a simple roll-under-stat mechanic is used, generally on a d20. Meatgrinder kills off the King's Army and town militia in the opening scene, leaving only a small band of irregulars, the PCs. While the DCC backgrounds include things like guards, soldiers and demihumans, I came up with my own collection from several giant lists of medieval occupations, leaving out any military or combat-competent jobs.

Character Generation

I streamlined the process of creating a character, pre-rolling 30 characters and filling out custom index-card character sheets.

Ability scores were simply 3d6, in order, and space was left for name, occupation/background, AC, Hit Points and notes. Occupations were charted up roughly on a d200 table, with each associated with one or two ability scores.

Character creation was relatively fast:

draw a PC card from the deck
name the PC (a list of suitable names cribbed from Dungeon World was provided, but most players preferred making up silly names)
roll on the d200 occupation chart, rerolling if the ability scores were an outright awful match, or if unrealistic duplicates happened (how many professional egg candlers can a small frontier town support?) The spreadsheet of occupations I used can be downloaded here.
come up with a tool or implement related to that occupation that, in a pinch, the PC could fight with. DCC provides these but I left it wide open for players to invent their own. The butcher had knives, the wainwright had a mallet, the barkeep had a club for rousting drunks, the nightwatch had a staff to lean on and a lantern, etc.
Finally, I gave the option to age the character, giving them a roll every 5 years to become better at their jobs. The roll was a simple d20, rolling under a related stat (and alternating every 5 years).

"Leveling up" in a job had no concrete game effects, but I did provide a new title (often Journeyman or Master) and some ideas about what getting better at the job meant. I also used critical successes (rolling very low) to sprinkle in some folk magic; one character had a brooch they believed kept wild animals away while they collected medicinal herbs in the woods, and another (the "dog leech" or town veterinarian) learned a song that seemed to calm distressed animals.

All-told, 10 players created 20 characters in about a half hour. There was room to front-load even more of this; I had plans to create occupation cards, each with three occupations to pick from (and their stats), along with concrete tool/weapon suggestions and leveling up options. I still plan to develop this, but ran into issues with my laser printer. Another six characters joined the party after being rescued (replacing other fallen PCs) and that only took a minute or two while other players were planning their next steps.

Actual Play (Spoilers)

The players are a band of irregulars told to hold a mountain pass outside of town against invading beastmen. I had the players select somebody to carry the absent Duke's backup war banner (which quickly became a spear). Fires brought the PCs back to town, where they found much of it burnt down. The stone church where the villagers too young, old or infirm were hiding had been breached and a bonfire in the town commons had "a couple dozen" horrible beastmen cooking and eating people. Meatgrinder goes hard on the child-murder, but I toned it down, deciding that beastmen stole children to raise as more beastmen.

The first fight began with a lot of careful planning. For obvious reasons, the players were tentative about a frontal assault, so they picked three volunteer runners to harass them and try to draw them into a fight. After two rounds of hilariously bad attempts to throw stones at the beastmen, six decided to give chase. With slightly fewer beastmen to deal with, the irregulars reluctantly charged. The opening barrage was barely effective, reducing the beastmen in number to match the PCs, so I had the players roll attacks for beastmen against their own characters. It was brutal, with every hit scored killing or maiming a PC. A second round of battle went only slightly better, but I decided to check morale, realized morale wasn't a provided stat for the beastmen, and decided they weren't in the mood to lose more of their number, what with their leader having left them behind for cleanup duty. The beastmen withdrew, giving the PCs a minor victory and setting the stage for surviving NPCs to tell the story of what happened.

The three runners rolled CON to see how fast they could run, outpacing half of their pursuers. Two PCs ran to the tanning pits (a feature invented by the players, since one was a tanner) and dove in (ick), and the last ran to the woods with two pursuers. They caught up with him, but one fumbled badly, throwing a meat cleaver into the other's head. Some desperate wrestling later had the PC (the village tinker) smashing in the skull of the beastman, the first truly heroic moment.

The survivors regrouped at "the old old mine", where an NPC told the players about the attack: the small band of King's Army soldiers sent to defend the town (and lead by the local duke) were routed at the river (with cries of "dragon!", though no survivor witnessed a dragon) and the militia fell to the beastmen. The leader of the beastmen, a 10-foot-tall boar-headed man, ordered the town burned, the survivors slaughtered, and then took the children and marched south along the river.

The PCs considered ransacking the Duke's Manor, but decided against treason (too bad, given the handful of useful weapons and armor they'd have found there). Weapons from the fallen beastmen and one unburned King's Army soldier were gathered, mostly crude meat cleavers but one shortsword and a suit of banded mail (I ruled that it would fit best on PCs with physical stats closest to the dead soldier).

Eventually, the players worked up the courage to go after the beastmen, trailing them to their lair in The Mountain Pass Caves (of Doom!). They sent scouts up the stone steps to the cave mouth, beginning the actual dungeon crawl proper after two hours into the session. They scouted the first room, full of bedbug-ridden cots and furs, waffled at the next tunnel and after their eyes adjusted to the darkness realized the tunnel contained some beastmen.

Too much planning later, the PCs assembled into cot-wielding pairs and lured the beastmen into the first room. Befitting the pathetic aesthetic, the cot-wielders had marginal success barricading or pinning the beastmen, but this was enough to kill the first two without casualties. Five more approached, but the scouts had discovered a drainage area in middle of the tunnel and suspected a pit trap, and sure enough the beastmen had to cross along the edges in single-file. Three more were dispatched before the rest began to flee; slowed by avoiding the pit trap, they fell to numerous thrown spears and a miner's pickaxe.

The tunnel lead to a room lit by firelight, with two side tunnels. These were briefly investigated; one had treasure on a pedestal (surrounded by eerie gem-eyed stone faces) and the other had some terrifying wall of spiky death gears, so the party continued into the dining hall.

They charged into the dining hall, where the feasting beastmen were surprised. Only a couple fell, with hilarious fumbles on both sides, but after a second round of fighting the four remaining beastmen fled further into the caves. The dining hall contained a pen with ten or so villagers waiting to be eaten; six were able-bodied enough to join the irregulars, replenishing some of their ranks and picking up weapons from the fallen.

While the party regrouped and carried their casualties to the cave mouth, one PC found the gold on the pedestal too enticing and decided to climb up to collect it. He died from the poison gas before he could finish saying "hey, something smells weird". The tinker studied the terrifying spiked gear wall from a distance, realizing it might move around the room and deciding it was too terrifying to approach.

This left one avenue; a narrow tunnel in the back of the dining room leading into the darkness. Two scouts investigated tentatively, finding a pale blue glow at the other end. Eventually they worked up the courage to find its source, a pyramid-shaped room with a single chanting beastmen in robes working a ritual sacrifice of a villager. The scouts tried distracting the beastman summoner, who only sped up the ritual, sliced open the sacrifice and called up some chaos demon (Hegoredulthu). The cultist fell a round later to the advancing party, but the demon started emerging from the chest of the sacrificed villager. Many attempts to wrestle the body off the crude stone altar were made, but the demon latched on with thin, bony tentacle-fingers (I was imagining John Carpenter's The Thing at the time) until somebody thought to throw the night watchman's lantern on the corpse to set it on fire. With the gateway into this realm on fire, the summoner dead, and surrounded by angry villagers with meat cleavers, the demon opted to retreat.

At this point, it was four hours into the session and I decided to add more tunnels to make getting to the final fight easier. The players were also focused on finding the children, so the two tunnels out of the summoning room were inspected and one (back into the meatgrinder room) was clearly dusty, so the other was followed. The simplicity of the dungeon was a benefit here, along with listening for what the players wanted.

The more-used tunnel lead into a large room with a dragon! A dragon! The PCs were rightly terrified, but the dragon seemed largely disinterested in them. This scene was mostly for comic effect (what color dragon is it? it's dragon-colored! how did it fit in here? maybe they raised it from an egg?) but finally one player got the idea to feed it the cultist and I decided the dragon-shaped pig would be content with that. I had the pig carry the cultist to one corner, leaving two tunnels to choose from: one next to the dragon and one not, so they chose to stay far from the dragon.

That tunnel lead to the head Pigman's room, where the children of the village were formed in a circle around him, staring blankly. There was room here to find out what was going on, but the players were having none of that; they charged in with everything they had left. One PC did decide to drag some kids out, but the rest surrounded the Pigman, who started goring and hewing PCs while they poked and prodded him with sharp implements. He finally fell to a blow from a heavy staff (I think a pole from the BBQ pit?). It would have been an anticlimactic fight but for the viscious followup blows the players narrated "to make sure he's dead."

The PCs then escorted the wounded, the children, and carried the dead back to town, stopping only briefly to use some jury-rigged ropes and poles to pull the gold off the pedestal without being close enough to the gas to die from it.

Notes from Running the Game

Morale rolls didn't work, at all. I suspect in a more typical dungeon crawl, where reaction rolls are also used, the pair might be relied on to make combat avoidable or end quickly, but in this scenario reaction rolls largely didn't make sense. I had to overrule the morale rolls constantly, having the beastmen flee at the first sign the PCs might be a threat, to keep the game moving and to avoid a full TPK. The feeling of desperation was unharmed by this; a majority of PCs fell in battle, some killed outright and others permanently maimed.

Another missing mechanic was the option for the PCs to flee. The players never considered running, aside from the impromptu chase scene near the start of the session, simply because running might mean their children died. In future sessions, where the goal is a little less pressing and failure is an option, I'll need to telegraph unwinnable fights more.

Initiative was very free-form. I was going to have each side roll once, but then I realized I could gain some of the benefits of individual initiative, even with ten players, by rolling once for the monsters. The combatants ended up in three groups - those that beat the monsters, those that tied the monsters (and thus went simultaneously) and those that lost initiative (or had giant two-handed weapons, like the farmer with the scythe). Each group went in any order, though by the end of the night I just picked one side of the table at random and went around.

Combat was pretty boring, as it often is, with an occasional fumble, critical hit, or death/injury roll to change things up and provide room for narration. This is entirely on purpose, though; the drama comes from the die rolls, not tactical decisions.

I gave the PCs descending armor class, but Meatgrinder used DCC's ascending armor class. It confused me briefly, but I think I may just use ascending AC after all. Mostly it didn't matter; everybody had an effective THAC0 of 20, the beastmen had a to-hit bonus baked in and players were rolling and adding Str or Dex bonuses accordingly. There was rarely a question of whether a PC got hit; only two had armor of any major value, and Dex bonuses/penalties were rarely more than 1 (using BECMI's stat table), so most of the time it was simply a question of whether the roll was 11 or above.

In any case, there were very few mechanics to track, and any weird idea the players came up with were easy to make quick rulings on. The cots-as-weapons I waffled on for a bit but decided an opposed roll-under-Strength sufficiently modeled the situtation and gave the PCs the advantage they were looking for (since it was 2-on-one). No combats took longer than ten minutes to run; the majority of the session's time was spent by the players planning, scouting, and examining things.

What Now?

From here I hope that a group or two will form from the players wanting to continue the story, in a West Marches-style free-form schedule.

With over half the town dead, the Duke and his soldiers missing and harvest season just starting, the PCs now must decide how to spend the gold they found to best survive the upcoming winter and rebuild the town. There are orphans to care for, no militia or patrols left, and probably no help coming. They've also witnessed real horrible eldritch magic: magic traps, hypnotized children, demon summoning, and pigs transformed into dragons. There's some more of the cave system to investigate and clues to other adventuring locales to find.

Not to mention becoming level 1 adventurers.

On-the-fly Multiclassing Rules for BECMI / BX

George Dorn — Mon, 08 Jul 2019 00:00:00 -0700

As much as I like Basic's simplicity, I miss the flexibility of multiclassing that 3.5 and 5e had. 5e's is even more reasonably balanced than 3.5's.

The key to the flexibility is not having to decide to be multiclass until after you've played a while. 1e and 2e's multiclass systems failed in this regard; you have to decide exactly what classes you were mixing when you start the character, and if you change your mind later you have to create a new character.

This optional multiclass system builds on Breeyark's excellent Building a More Perfect Class series; that spreadsheet / calculator is essential for this additional feature. In fact, traditional 1e/2e multiclassing is entirely possible using just that calculator, by picking the class features you want for your character as it is created. This system adds the ability to buy additional features after the PC has been adventuring for a while, by spending XP to gradually become a new custom class.

How Multiclassing Works

At any time, a PC may decide to start learning a new set of class features. A cleric might want to train like a fighter to gain the ability to use swords or have a d8 hit points. A halfling might want to start learning thieving skills.

When this happens, use the class xp calculator to build in those new features, and generate another XP table for the character. They don't change levels or classes yet, but new XP may be spent to level up those class features that were initially missing. But until both classes are equal, buying up the new class features costs one level higher. This also means that getting to level 1 in the new class costs the xp it would have taken to get from 1 to 2 if the character had always been this set of classes.

An example: A 5th-level fighter decides to pick up spellcasting and turn undead like a cleric, and doesn't want to wait until level 9 to become a paladin. His old, new, and difference chart looks like this:

Lvl	Old	New	Diff	Cumulative
1	0	0	600	600
2	2000	2600	1200	1800
3	4000	5200	2400	4200
4	8000	10400	4800	9000
5	16000	20800	---	---

He's already at 5th level, having obtained 16000xp. To gain the powers of a level 1 cleric in addition to his existing fighter abilities, he'll need 600xp. To gain the powers of a level 2 cleric, he'll need 1200xp more, and so on. Once both classes are the same (level 5 in this case), his XP is set to 20,800 (5th level for the new class) plus any extra he had earned in Fighter before starting this exercise. It cost 9000xp to get there, instead of just the 4800xp difference between level 5 fighter and level 5 fighter/cleric, but that's the cost of the flexibility.

It's not mandatory to advance in the new class continually. When XP is earned, the player must decide whether to allocate it to the old class or the new class. However, so long as the player's classes are unequal, a 10% penalty is applied to all earned XP. (This is optional; it applies only if prime requisite xp penalties and bonuses are in use.)

Improved Hit Dice

If the new class has a higher hit die than the old class, the player should replace the rolled hit points with a roll of the new hit die. However, since most players don't track the hit points rolled at each level, they can instead opt to add the difference in average expected results, using the following table:

Old HD	New HD	+hp/level
1d4	1d6	1
1d4	1d8	2
1d6	1d8	1

If using the option d10 / d12 HD option from Breeyark, the pattern (half the difference in sides) holds; upgrading from a d4 to a d12 adds 4hp/level.

These extra hit points are only added when a level of a new class with a higher HD is being purchased; once the character's levels match, the new higher HD is rolled as normal.

Multiple Saving Throws

In most cases, the player should choose which class's saving throws to use when creating the multiclass. In the fighter/cleric example, I assumed the new class would keep the fighter saves. Optionally, he could have used the cleric's saves, which are mostly better, though occasionally worse, for slightly more xp (2800 base). If choosing a more expensive set of saving throws, the PC keeps any saves from the old class that were better than the new class, until advancement in the new class provides better saving throws.

For example, when the 5th-level fighter decides to gain levels of cleric, his saves are 10/11/12/13/14. If he opts to pay the extra XP for cleric saves, the fighter saves are better until he reaches 5th-level in cleric. At that point, he takes the best save for each class, or 9/10/12/13/13 (where a normal cleric of level 5 would have 9/10/12/14/13). Then his saves continue improving as a cleric.

Overlapping Spell Levels / Level-based abilities

If our example fighter/cleric reaches 9th level and decides to become a paladin, he'll be gaining cleric spells from two sources, as well as gaining the turn undead feature from two sources. There are several options worth exploring, but it's unclear which is the most balanced:

Ignore overlaps: Ignore all but the best source of spells / class feature. In the case of a paladin/cleric, the paladin still contributes detect evil and hirelings, but doesn't provide cleric spells or turn undead.
Allow overlaps where possible: Keep the best source, add extra memorization/castings per day from the other source, to a maximum of 9 in each level. A paladin/cleric will get his level in cleric spells, but also his level / 3 in additional cleric spells. For a 9th-level paladin/cleric, this just means an extra 2 first-level spells per day. Around level 21, the 9-per-day limit starts to take effect for level 1 spells. This is probably relatively balanced. Because turn undead has no daily limit, this option ignores the paladin's turn undead contribution.
Add sources together: Take the best source of spells / class features, and for each other source, add levels equal to that source divided by the number of sources. So our 9th-level paladin/cleric will cast spells or turn undead like a 10th-level cleric (9 from cleric, (9/3)/2 for paladin). At level 12, he'll cast spells like a 14th-level cleric. Around name level this isn't too bad (since it'll be adding 1-2 cleric levels but the PC has earned almost 1-2 levels of fighter in XP) This is probably unbalanced at much higher levels, however.

Demihuman Limits

The class calculator makes it trivially easy to remove level limits from demihuman classes. The Breeyark calculator suggests this adds 100xp to the base amount. The DM should use caution when allowing this option, in particular if Magic-user spells are involved in the class. The class feature of casting magic spells should be worth a lot more if the class can reach level 36, compared to the vanilla elf's level limit of 10. This accounts for the main discrepancy in the Breeyark calculator, which thinks elves should have a base xp of 4100 (vs 4000) and magic-users should have a base xp of 2050 (vs 2500!). Having unlimited access to magic-user spells is probably worth closer to 2000 base xp, instead of the calculator's 1600.

A similar adjustment might need to be made for demihuman clerics with no level limit.

It's also possible for a demihuman PC to reach maximum level and then decide to start spending that extra XP to remove the level limit; this will end up costing close to 50% more xp per level than just starting with that option, so it's not absurdly unbalanced.

Another consideration is demihuman hit dice. The class calculator also makes it easy to change the hit dice or saving throws of a demihuman. For instance, it's possible to make an elven fighter/magic-user with a d8 hit die instead of a d6. It's up to the DM whether to allow this; it may make sense to require demihumans to stick to the hit die, saves and to-hit chart of the original vanilla class.

Some More Examples

Some common 1e multiclass options include dwarven fighter/thieves and elven fighter/mages. Gnomish wizard/thief is also a textbook example in 1e, but we'd need to start by inventing the gnome class.

A dwarven fighter/thief has 2850xp/level, assuming the demihuman to-hit chart and weapon mastery, dwarven saving throws, all armor and dwarven-style limits on weapons. The DM might choose to penalize the use of some thief skills in heavier armor, but there's no existing rules for it yet. If the PC started as a vanilla dwarf and just wanted to add in thief skills, the chart up through level 5 looks like:

Lvl	Old	New	Diff
1	0	0	650
2	2200	2850	1300
3	4400	5700	2600
4	8800	11400	5200
5	17600	22800	---

To extend the chart, the Diff column is New - Old for the next row down.

An elven fighter/mage played this way must have started as a custom class to begin with, before mixing in the other class. The new XP chart is the vanilla elf, but assuming the elf started as a fighter, with a base XP of 2200 (using Elf saves, Demihuman to-hit, fighter specials, elven racial abilities, name level limit).

Lvl	Old	New	Diff
1	0	0	1800
2	2200	4000	3600
3	4400	8000	7200
4	8800	16000	14400
5	17600	32000	---

This is a very expensive way to advance to a textbook elf, but the survivability of a fast-advancing fighting elf might be worth the extra cost in XP.

Conclusion

This is still a work-in-progress, but I think it captures the spirit of what I'm trying to accomplish. It adds much flexbility for long-lived characters who want to change up what they're doing, without needing to port in the somewhat wonky dual-class options from later editions, or forcing players to design their multiclass setup from the beginning.

There might be some balance issues I haven't considered; I've mostly thought about levels 1-5 since that is my favorite level of play. It's entirely likely that something breaks badly around name level or higher.

How to die in D&D

George Dorn — Tue, 07 May 2019 00:00:00 -0700

Avoiding death is a primary motivator for most adventurers. Whether by avoiding combat and traps or defeating enemies and subverting dangers effectively, most decisions PCs make are heavily influenced by the risk of death. Obviously, then, the rules by which characters die (whether a temporary inconvenience or permanent affliction) have an enormous impact on the style of play at the table, and these rules have changed many times over the years.

Brown Book D&D (1974)

As an addon to the wargame Chainmail, in which heroes were considered units expendable in pursuit of victory, the original D&D publication barely mentions death at all, merely adding a note to the Dice for Accumulative Hits (Hit Dice):

This indicates the number of dice which are rolled in order to determine how many hit points a character can take, [...] the number of points of damage the character could sustain before death.

Hit points were low, at slightly more than 1d6 per level for Fighters, slightly less for Clerics, and roughly 1d6 every two levels for Magic-Users. Death was a pressing concern at all times, as all weapons also did 1d6 damage and monsters attacked once for every HD. An evenly-matched fight almost certainly meant deaths on both sides.

Compared with later editions, death was also instantaneous. At 0hp, you died. No time for the rest of the party to win the fight and come to your aid. This simplified the survivors' decision to flee the scene to a simple question of the value of the dead PC's equipment...

On the subject of healing, only magical healing and recovery is discussed. Clerics and healing potions were essential, and natural healing left to the DM to adjudicate. A ring of regeneration, however, would bring the user back from death unless destroyed like a troll, a theme that would come back in Gygax's other work. Raise dead existed, with a brief mention that Constitution determines the chance of success - if we are to assume this meant 'adversity' as mentioned in the Abilities section, a PC of average Constitution seemed to have about 70% chance of survival after being raised.

Supplement 1: Greyhawk (1976)

Original D&D being incomplete, the supplements that followed are often considered part of that "edition" and here some things are clarified:

Hit point bonuses from high Constitution were expanded from the previous +1 for high scores to +1 for 15-16, +2 at 17 and +3 at 18.
Hit dice were determined by class (d8 for fighters, d6 for clerics, d4 for others) instead of being 1d6 occasionally gained for each level.
Probability of Resurrection Survival was given, based on Constitution scores, and a rule added that your Constitution is also the number of times you can be resurrected.

Supplement 2: Blackmoor (1975)

An odd supplement (with an odd publication date), I only mention it because it includes the first rules about hit locations. They're complex and I suspect rarely used, but it did open up more possibilities for random instant death. A kobold throwing a dagger could kill any 1st-level character, and quite a few higher-level characters, just by doing a point or two of damage to a PC's head.

Basic (1977 - 1983+)

Basic continued the tradition of death at 0hp. The boxed Mentzer red book player's guide explains that as part of the single-player adventure, including a special note explaining atypical rules for that adventure:

If you are struck down to zero hit points or less, you can grab your potion - if you still have it - and drink it before you pass out. It will cure you somewhat, but only back up to 4 hit points. If you don’t have the potion left - sorry, but you are dead! (Special note: In group games, you will not be allowed to do this. Zero hit points indicates death, with no extra time to do anything.)

And the frequency of death is even addressed early, if the player dies in that demo dungeon:

Your character has been lost in the dungeon. Don’t be upset; it can happen in any DUNGEONS & DRAGONS game, and often does, through no fault of yours.

While natural healing isn't explained, resting at an inn for "a day or two" seems to be enough for most low-level PCs. As for magical healing, at this low level only potions and spells of cure light wounds are detailed.

Expert adds the previously-mentioned raise dead spell, though the risk of failure is removed and no limit to resurrections is mentioned. This significantly changed the flavor of the game, once a cleric reached 10th level, or access to the spell was obtained through other means. Where in OD&D a raise dead might bring a PC back to life a couple times, now so long as the party cleric survived, anybody could be raised indefinitely. Not only did this make combat far less risky for individual PCs, once death had occurred in a combat, it became vital to reclaim the body before fleeing, preferably in one piece!

Expert also added the ring of regeneration but explicitly stated that it ceases to work when the wearer's hit points drop to 0 or less.

Companion and other books extended access to raise dead - the first Gazetter suggests the cost of employing a 10th-level cleric as a Chaplain at a stronghold should be 1,000gp!

AD&D First Edition (1979)

The rules of death, dying, and healing in AD&D First Edition's Player's Handbook have much more in common with Gygax's early work than the Mentzer/Holmes Basic line.

Damage is meted out in hit points. If any creature reaches 0 or negative hit points, it is dead. Certain magical means will prevent actual death, particularly a ring of regeneration (cf. MONSTER MANUAL, Troll).

Gygax kept the lethality of the original game intact; low-level AD&D characters were expected to die just as frequently. He also kept his curiously powerful ring of regeneration which must have resulted in a common trope of the party fighter dying, the rest of the party fleeing, and the fighter waking back up an hour later, alone in a room full of dangerous monsters...

For the first time, rules for natural healing were detailed: PCs heal 1hp per full day of rest up to day 30, and then 5 per day after. This option was really only intended for parties without a cleric, given that even a 1st-level cleric had access to faster magical healing.

Raise dead continued to be a risky proposition, retaining the saving throw to survive the ordeal and the limit on the number of times a character can be brought back to life (equal to the character's initial Constitution, regardless of magic). It also added a new drawback: every time a character is resurrected, they permanently lost a point of Constitution, making future resurrections more difficult.

The Dungeon Master's Guide

Confusingly, the Dungeon Master's Guide also had rules for dying from hit point loss and natural healing that conflicted with the Player's Handbook. It might be assumed that the curt rule statement of the Player's Handbook was meant to express the serious risks involved in combat and that the DM would only bring up the rules of the DMG when they first applied. In any case, they made significant changes:

When any creature is brought to 0 hit points (optionally as low as -3 hit points if from the same blow which brought the total to 0), it is unconscious. In each of the next succeeding rounds 1 additional (negative) point will be lost until -10 is reached and the creature dies.

Healing a character through wound-binding or magic would stop the bleeding, but the character would remain in a coma for up to an hour and then require a full week of bed rest (unless a heal spell is used). The DM is also advised to apply permanent scars or loss of limb to PCs reduced to -6 or lower...

The DMG's natural healing rules mostly matched the PHB's, but noted that four weeks of bed rest will return any character to full hit points, making the jump to 5-per-day moot.

AD&D Second Edition (1989)

Second edition also kept the tradition of instant death at 0 hit points.

Natural healing through rest comes with two options; resting while traveling regains 1hp per day, while a full day of bed rest regains 3hp (adding Constitution bonuses each full week).

Being raised from the dead retained the saving throw and Constitution loss, though the hard limit was removed, allowing PCs to regain lost Constitution and be fully restored through items like a manual of health or even a wish.

Hovering on Death's Door

The Dungeon Master's Guide, however, adopted the "unconscious and bleeding" rules from the First Edition DMG as an optional rule called "Hovering on Death's Door." It behaved much like First Edition's, though without the coma, but made special note that a character forgets all spells when dropped to 0hp or below, regardless of the method of restoring the character.

D&D Rules Cyclopedia (1991)

While normally I group this under the Basic heading, this appears to be new rule created for this book. In Chapter 19: Variant Rules, two optional rules are provided.

First, it presents an option to simply remove all magic that raises characters from the dead. This helps to restore the threat of death to parties with a cleric level 10 or above, and reduces the odds of a PC reaching truly staggering levels.

Second, it suggests that if using the first rule, the DM should adopt a change to the death rules, such that a character doesn't necessarily die at 0hp. Instead, a save vs death ray is made immediately and every 10 minutes; so long as the saves are made, the PC is still alive and can be revived with healing magic or the Healing skill (added in Gazeteer).

D&D Third Edition (2002)

Third edition canonized the optional rule from second edition and elaborated on it further:

When your hit point total reaches 0, you’re disabled. When it reaches –1, you’re dying. When it gets to –10, your problems are over — you’re dead.

At 0hp, a character can limp around, but not take any standard actions (like attacking, casting a spell, or anything strenuous) or suffer 1 point of damage and become unconscious.

Below 0hp, a character starts bleeding, losing 1hp per round, until -10hp, at which point the character dies. There's a 10% chance each round that the bleeding stops (or at least slows to once per hour), and any external healing has the same effect. From there, magical or natural healing can recover the character, though without aid it is unlikely.

Natural healing yields 1 hit point per level for eight hours of full rest; a full day of rest doubles this.

A ring of regeneration ends up behaving quite similarly to original D&D as a consequence of these rules, as it will stop the bleeding and return the character to consciousness as it regains hit points.

Magical resurrection is less chancy than previous editions, requiring no special save mechanic to survive, but all but the most powerful forms result in the loss of a single level. The most powerful form, the level 9 true resurrection, has no downside, but costs 25,000gp in diamonds. Very high-level parties may scoff at that tiny expense to bring back a party member, leading to the cheapening of death as seen in the Basic lineage. In my own game, I've addressed this by making diamonds a somewhat limited resource; even a big city can only provide so many diamonds for PCs to purchase at one time.

D&D Fourth Edition (2008)

Fourth edition does away with the disabled condition, but keeps the "down and bleeding" rules for when a character reaches 0hp. Without help, the character makes saving throws each round; three failures means death, but one significant success allows the spending of a "healing surge" to recover hit points (from 0) and wake back up. While dying, any healing restores hit points as if at 0hp, resulting in recovery.

Healing surges are a new resource created in 4e; every character has a number of them per day based on class and Constitution. The surges can be used by the character themselves (once per encounter as an action called "second wind" or while taking a short rest) or used by other characters casting healing spells on the character. This resource was reportedly meant to both remove the reliance on dedicated party healers, but also to regulate the pacing of adventures by treating it like any other expendable resource. However, it's also very abstract, not representative of anything within the game fiction, and many felt it was too unlike any previous edition.

Natural healing is simply a matter of resting and spending healing surges; a full night's rest fully heals any character, removes almost all negative status effects, and recovers all powers.

Raise dead made use of the new Ritual system, in which anybody could purchase ritual books from a marketplace and use them. A newly-raised character suffers some minor penalties (-1 to most rolls) until the characters "reaches three milestones" - another new concept pertaining to adventure progression. This reduces recovering a character to a matter of going to town, spending some gold (based on the 'tier' or level of the character) and spending 8 hours casting a ritual. Even a level 1 party can do this, so long as they have 500gp to spend.

D&D Fifth Edition (2014)

Despite the lukewarm reception of Fourth edition, many of the death, dying and healing rules were retained to some degree.

Being reduced to 0hp knocks a character unconscious. Each round thereafter, a roll with 55% chance of success is made; collect three successes before three failures to become stable and avoid dying. During this time, any healing (magical or skill) returns the character to normal. There's some other rules surrounding the death saves process, but on average most characters will recover rather than expire unless further damaged.

Healing is primarily done via a 'short rest' mechanic similar to Fourth edition; rather than spending healing surges, however, the character spends recovery hit dice (the same number and type the character rolled for hit points) to recover hit points; these hit dice are recovered after a long rest, which also heals all damage and most status effects. Magical healing remains useful for combat situations and doesn't rely on the recovery hit dice mechanic, however.

Returning characters from the dead is available at a surprisingly low level; 5th-level clerics can cast revivify to raise somebody that died in the last minute with no negative effects. Characters dead for longer need a more powerful spell like raise dead or resurrection, which (like Third edition) require some gold expenditure to obtain diamonds and levy some temporary penalties on the subject and/or caster that can be removed with rest.

Summary

The rules for how to die in each edition reflects that edition's philosophy of play. In some editions, death is meant to be easy and cheap, reflecting a pathetic aesthetic or suggesting combat as brutal and dangerous and to be avoided when possible. In other editions, combat is glorified, and the roles of PCs within combat is even more glorified, so much so that random and pointless death no longer fits the gameplay.

Other quirks, like the availability, cost and risk of magics used to raise dead characters, point to finer nuances of play; death came easy for Basic D&D characters, but after the "character funnel" of low-level play, death became a trivial annoyance for all but the worst cases, and an annoyance you could keep a cleric on retainer back home to deal with.

All of this also influences play style for PCs; players who know death is hard and resurrection is easy will behave much differently around powerful monsters. The same players will have a much different view of combat and risk if they experience six months of character-funnel play.

A history of gaining experience in D&D

George Dorn — Tue, 23 Apr 2019 00:00:00 -0700

A lot of ink has been spilled on the subject of murder hobos, whether they're the pinnacle of D&D play-as-it-really-happens, and what to do about it. Many OSR blogs point out (often correctly) that older editions downplayed the "kill" part of "kill stuff and take their loot" but when did the change happen?

Murder hobos behave according to incentives, generally of three types: experience (power gained through levels), treasure (power gained through money and magic items) and fame (power gained through narrative reward).

Brown Book D&D (1974)

Originally, D&D awarded XP for treasure gained and monsters defeated. The actual XP rewards were not included in the books, though an example of calculating XP for an encounter was:

Experience Points: Experience points are awarded to players by the referee with appropriate bonuses or penalties for prime requisite scores. As characters meet monsters in mortal combat and defeat them, and when they obtain various forms of treasure (money, gems, jewelry, magical items, etc.), they gain "experience". This adds to their experience point total, gradually moving them upwards through the levels. Gains in experience points will be relative; thus an 8th level Magic-User operating on the 5th dungeon level would be awarded 5/8 experience.

Though the actual rewards are never listed definitively (an ommission that would be repeated by the "Blue Book" Basic of 1977), we can infer some practices:

XP for monsters was based on monster level; many groups assumed 100xp per HD (see Greyhawk, below).
XP for treasure was 1xp per gp.
XP was scaled for each character, based on relative level to the monster, who typically matched the dungeon level.
XP was scaled for each character based on their Prime Requisite scores; higher scores gave an XP bonus and was the primary means by which exceptional scores benefitted characters.

The first rule was contentious enough to require errata, but even groups playing under these rules would see a gradual shift in the primary source of XP as characters gained levels.

Low-level characters facing a typical group of low-level monsters, say a lair of kobolds, would earn XP primarily from slaying the monsters themselves, and a comparatively small amount of XP from the treasure obtained. A small (!) lair of 40 kobolds would yield around 200gp in treasure (140 from the 1d6-per-enemy, and around 60, with wide variance, from the level 1 lair treasure expected to appear in any room containing a monster). That's 4000xp from monsters and 200xp from treasure.

By the time the party was fighting 5HD enemies, though, this was a different story. A half-dozen Mummies or Cockatrices (both treasure type D) would yield 3000xp from the monsters, but another 3000xp from treasure (~700gp from the "lair" treasure, and ~2300gp from treasure type D). From here, XP from treasure outpaced XP from monsters, with dragons earning tens of thousands of GP while only being worth perhaps two thousand XP for simple slaying.

Supplement 1: Greyhawk (1976)

The incompleteness of the original D&D printing led to Gygax making drastic changes in the form of a supplemental pamphlet. In this, he and Kuntz address the XP system:

Guidelines for A warding Experience Points for Monster Slaying: (Addition)

The awarding of experience points is often a matter of discussion, for the referee must make subjective judgments. Rather than the (ridiculous) 100 points per level for slain monsters, use the table below, dividing experience equally among all characters in the party involved.

This change would be canonized in the rest of the Basic lineage, reprinted verbatim in the Blue book (1977) as well as influencing the AD&D rules. The table image above is from Expert, but the numbers remained largely unchanged throughout the Basic lineage.

It's clear that Gygax's intent would continue to emphasize XP from treasure over XP from killing monsters, only now this was true even at very low levels - those 40 kobolds are now worth 300xp from slaying, but the same 200xp from average treasure found; the cockatrices worth 1800xp from slaying but 3000xp from average treasure.

Basic (1977 - 1983+)

Moldvay Basic (the first "Red Book", TSR2014) lists treasure as the first source of XP, and then reprints the XP-per-HD table, expanding the rules and procedures for awarding XP:

XP rewards for encounters include "monsters killed or overcome by magic, fighting, or wits."
XP for treasure was allocated by the DM, even if the party decides to split treasure in a different way.
DMs were encouraged to increase rewards for exceptionally tough encounters.
DMs were permitted to reward partial XP for encounters the party doesn't win, but learns from anyway.
DMs awarded XP to surviving members of a party, at the end of an adventure. This nuance technically means survivors benefit from the deaths of other adventurers, though in practice most groups would allow replacement PCs who would recieve XP earned by the characters they were replacing.
XP was capped at one level per "adventure" - if you recieved enough to gain a second level, you earned just enough to be 1xp shy of that level. When followed, this rule lead to PCs seeking out short adventures to earn a few xp before diving back in to a longer adventure...

Mentzer Basic (the boxed "Red Book", TSR1011B) changed the rules in some small but significant ways:

XP for monsters was only given for slain or "conquered" monsters, but not those that flee.
XP for treasure now matched the split the party decided on, not the DM, so parties that gave more gold to PCs that didn't recieve magic items (which weren't worth XP) got more XP instead.

Both Basic books continued the Prime Requisite bonus XP, but reduced it to 5-10% from the original 10-20%. Both books also codified rules that NPCs recieved half a share of both the treasure and XP for monsters.

Companion (TSR1013, 1984), added some adventure design guidelines that revealed some further design intent. First, On most adventures, XP gained from defeating monsters should be 1/5 of the total XP. Second, it made the first mention of "bonus XP", suggesting ad-hoc rewards for "completion of a goal (rescue, retrieving an item, etc.), special individual actions (heroic performance, exceptional or frequent use of special abilities), alignment play, or other aspects of the overall adventure." This wasn't a new idea - such bonuses had already been mentioned infrequently in published modules. For example:

B2, Keep on the Borderlands, granted an ad-hoc reward for destroying evil writings (or keeping them if the party is evil).
B5, Horror on the Hill, granted an ad-hoc reward for destroying supplies of a mustering Hobgoblin assault force.
B9, Castle Caldwell and Beyond, has ad-hoc rewards for cleverness in dealing with a trap,
X1, Isle of Dread, offered the unique option to grant XP to players if they set up a trade route to the island.
An adventure in the Companion book itself offers a hefty XP reward to PCs that win a jousting tournament.

Gazetteer made significant contributions to Basic throughout the 80s. A few options for the XP system were added. For example:

GAZ2, The Emirates of Ylaruam (TSR9194, 1987) make heavy use of small rewards to reward roleplaying the distinct culture of the setting; PCs earn XP for behaving honorably, suggesting 1/100th of the XP needed to reach the next level, or a 5% bonus at the end of an adventure. Elsewhere, more design intent is revealed, suggesting a PC's XP should come 1/5 from defeating monsters, 1/5 to 2/5 from treasure, and the remaining 1/5 to 2/5 from goals, honor challenges and good roleplaying.
GAZ3, Principalities of Glantri (TSR9208, 1987) introduces many new options for magic users, along with a new ways for them to earn XP: creating magic items, researching new spells, acquiring rare knowledge (a reason to adventure), etc. Along with these new options are suggestions that magic users earn less for treasure and monster-slaying, lending motivations other than traditional adventuring.

Rules Cyclopedia (TSR1071, 1991) collected many of these and standardized the ad-hoc awards, establishing five sources of XP:

Roleplaying Awards were 1/20th of the base XP needed to advance to the next level. (E.g. a fighter reached level 4 at 6,000 xp, and will reach level 5 at 12,000 xp; a 1/20th award is therefore 300xp). These could be heroism, good alignment play, etc, but the DM was advised to limit this to once per session.
Goal XP was equal to the XP value for all of the monsters defeated in achieving that goal. Confusingly, this wasn't divided among the party; the party would earn a share of monster XP for fighting monsters, but then the entire XP was awarded to each PC again if the goal of the adventure was achieved.
Monster XP - as in prior books, XP for fighting monsters, divided among PCs. Monsters fought but not defeated were worth 1/4th their value.
Treasure XP - 1 per GP, as in prior books. This also included payments or rewards PCs earned for finishing an adventure.
Exceptional Actions - another 1/20th of next level award. A bit nebulous, but the examples given include saving other party members from harm, or pulling off exceptionally difficult or unusual uses of skills.

By this point it was clear that XP from sources other than killing monsters were the primary means of attaining higher levels, especially after the first few levels.

AD&D First Edition (1979)

1e brought back a lot of Gygax's early ideas about adjusting monster XP by PC level:

Dividing the total adjusted hit dice equivalent of the monsters slain by the total of all levels of experience of all characters who had a part (even if only 1 missile, blow, spell, etc.) in the slaying yields a fraction which is the measure of challenge. If the numerator is greater than the denominator, then full experience should be awarded. If the denominator is greater, use the fraction to adjust the amount of experience by simple multiplication.

XP for treasure was also retained, though the un-named challenge ratio mentioned also applied here:

If the relative value of the monster(s) or guardian device fought equals or exceeds that of the party which took the treasure, experience is awarded on a 1 for 1 basis. If the guardian(s) was relatively weaker, award experience on a 5 g.p. to 4 x.P., 3 to 2, 2 to 1, 3 to 1, or even 4 or more to 1 basis according to the relative strengths.

Magic items were also worth XP, beginning a tradition of listing both XP value and GP value for magic items that would last into second edition.

An amusing side-note was added, suggesting Gygax was already facing players annoyed by XP-for-treasure:

Note: Players who balk at equating gold pieces to experience points should be gently but firmly reminded that in a game certain compromises must be made. While it is more "realistic" for clerics to study holy writings, pray, chant, practice self-discipline, etc. to gain experience, it would not make a playable game roll along. Similarly, fighters should be exercising, riding, smiting pelts, tilting at the lists, and engaging in weapons practice of various sorts to gain real expertise (experience); magic-users should be deciphering old scrolls, searching ancient tomes, experimenting alchemically, and so forth; while thieves should spend their off-hours honing their skills, "casing" various buildings, watching potential victims, and carefully planning their next "job". All very realistic but conducive to non-game boredom!

It would seem that the Gazetteer authors agreed with the Gygax's premise, but not the conclusion...

Gygax's XP-for-monsters chart also continued to be used, though slightly modified:

While there was no mention of roleplaying awards or bonuses for achieving goals, Gygax did suggest an optional special bonus of 1,000 xp for any PC that was slain but then raised from the dead.

On the subject of good roleplaying, though, Gygax added training time for advancing in level, and provided extensive rules on how to determine the number of weeks spent training, based on how well the PC performed the "natural functions" of their class, as well as acting in accordance with their alignment. Given that each week of training also cost 1500gp/level and "Poor showing with aberrant behavior" could quadruple the time spent, this was all stick, no carrot.

Additionally, Basic's "XP cap" was kept, but in a new awkward form: when a PC earned enough XP to gain the next level, they stopped earning more XP until they trained for that level. The rare DM that enforced that rule would be wise to also only award XP at the end of adventures, to avoid PCs taking a month of downtime between excursions to the plot dungeon.

Despite this minimal core system, module authors were happy to add other ad-hoc sources of XP:

T1-4, Temple of Elemental Evil, had prisoners that could be rescued for XP rewards, as well as some small, non-magical treasures. Early encounters with prisoners weren't worth much, but later ones were nobles worth both XP and platinum.
U1, The Sinister Secret of Saltmarsh, has an intermission in which PCs consider collected clues. PCs making deductions and taking certain actions to further the plot are awarded experience points at the DM's discretion.
U2, Danger at Dunwater, has a half-page digression on how to calculate an XP reward for a pivotal decision made by the party.
S1, Tomb of Horrors - even Gygax's own module contained a significant ad-hoc award for defeating the final enemy.

AD&D Second Edition (1989)

By the time Second Edition was printed, many traditions from existing groups, tournament play, published modules and other games were collected and added to the rules.

A full page in the DMG suggests possible situation that warrant XP rewards, including awards for players making the game fun (and not being disruptive), PCs surviving, players improving their play style, and PC achieving story goals. These awards are nebulous and don't include XP values.

More concretely, the existing first-edition rules were expanded, including new ideas to replace XP-for-treasure:

XP for monsters included monsters killed, captured, routed, forced to surrender, or even persuaded to accept terms all count.
XP for treasure was limited only to Rogue characters.

Additionally, per-class awards were somewhat codified:

Now, with the exception of the Rogue, characters could expect to earn most of their XP from defeating monsters, taking specific in-class actions (presumably while defeating monsters) and nebulous ad-hoc XP bonuses.

D&D Third Edition (2002)

3e changed everything about levels and XP. For starters, the XP required to gain a level was standardized across all classes at 1,000XP per current level above what you needed for your current level.

Level	Additional XP needed	Total Xp Needed
1	0	0
2	1000	1000
3	2000	3000
4	3000	6000
5	4000	10000

This is a drastic reduction from prior editions, which doubled total XP needed every level for most levels, then settled into 100,000xp or more per level after.

To compensate, XP calculations depend heavily on the PC's current level; one ogre (CR 3) against a party of four 1st-level PCs earns each of them 225xp (23% of level 2), while the same ogre against a party of four 5th-level PCs is only worth 188xp (4% of level 6). Furthermore, this causes lower-level party members to earn more XP and catch up to the average over time.

The DMG also provides three other means of earning XP without slaying (or defeating) monsters. It refers to these as "Story Awards".

Non-combat encounters that are hazardous also have Challenge Ratings based on how dangerous they are. A trap might be CR3, just like that Ogre, if it is roughly as likely to cause injury or death.
Mission goals earn XP rewards as well, though the specifics are left vague: The mission award should be more than the XP for any single encounter on the mission, but not more than all standard awards for encounters for the mission put together.
Roleplaying awards are left up to the DM, but a suggestion of 50xp per character level is made. Due to the change in the XP-per-level chart, this is identical to 1/20th of the XP needed for the next level, as in Basic's Rules Cyclopedia.

However, there's an important note about these Story Awards; the DM is advised to either halve or remove the "Standard Awards" (that is, XP for defeating monsters) if Story Awards are used:

Don’t simply add story awards to standard awards (even if you compensate by giving out more treasure as well) unless you want to speed up character progression.

D&D Fourth Edition (2008)

4e squashed the XP chart further; there's a complicated algorithm, but the short version is that each new level requires just a little more additional XP than the last one, and that extra amount doubles each half-tier.

Level	Additional XP needed	Total Xp Needed
1	0	0
2	1000	1000
3	1250	2250
4	1500	3750
5	1750	5500
6	2000	7500

XP for encounters is fixed, with a standard monster of a given level granting 1/10th the XP a PC of that same level needs for the next level. So a level 3 Ogre is always worth 150XP (assuming a "standard" Ogre) regardless of PC levels.

Non-combat encounters with risk (traps, puzzles, etc) are counted as monsters defeated based on the type of Skill Challenge they represent. Skill Challenges are treated as combats, complete with initiative, and consist of one or more party members rolling skill checks, trying to collect a set number of successes before collecting a set number of failures. Published modules treat these as encounters, just like monsters, and specify their XP rewards for success, as well as consequences for failure.

This did not help defend 4e from accusations of being a tabletop MMO.

D&D Fifth Edition (2014)

5e modified the XP-per-level chart again, squishing the base significantly but bringing back a (somewhat) geometric progression similar to 3e's.

Level	Additional XP needed	Total Xp Needed
1	0	0
2	300	300
3	600	900
4	1800	2700
5	3800	6500
6	7500	14000

According to the DMG, this should work out to roughly one session (not adventure!) per level up to level 3, 2 sessions per level to level 4, and then 2-3 sessions per level thereafter. This is drastically faster than any prior edition, but is balanced by a slightly shallower power curve; 5th-level PCs are no longer superheroes when compared to 1st-level PCs as in any edition through 3.5.

Meanwhile, the XP-for-monsters works the same as 5e: monsters have pre-printed XP values, you add up the XP for the encounter and divide it among the PCs that participated.

Non-combat challenges are built like combat challenges, using the same rules to estimate the danger and consequences for failure, and award XP in the same way as combat encounters.

Additional awards can be made at milestones throughout an adventure and at its end, treating minor milestones as "easy" encounters and major ones as "hard" (there's encounter-building rules explaining difficulty levels).

Ad-hoc awards for decision making or good roleplaying are notably absent; a clear design goal of rewarding XP for overcoming challenges was closely adhered to. Given the purported "three pillars" design goals of Socialization, Exploration and Combat, the XP reward system feels under-developed.

Summary

Only recent editions of the game have been motivated to focus, through XP rewards, primarily on killing stuff. Basic, 1e, 2e and to some extent 3e have codified (or even enforced) sources of XP other than killing monsters; from acquiring treasure to roleplaying to accomplishing goals, older editions rewarded non-combat play and intelligent problem solving, often to a much greater extent that defeating monsters.

After 3.5 (including 3.5 to some extent), winning combat encounters became the primary means of gaining levels.

A history of hitting things in D&D

George Dorn — Tue, 16 Apr 2019 00:00:00 -0700

In working on the as-yet unnamed homebrew Basic [1] D&D derivative, I'm having to make some choices about what features from post-Basic editions I'd like to adopt. In doing so, I was reminded of an argument at the game table about which edition invented THAC0 - was it 1st edition AD&D or 2nd? The answer is far more complicated.

The Examples

Along the way, I'll be using two example scenarios:

a first-level fighter with a longsword, and just enough strength to gain a +1 to attack rolls (a 13 in many editions, a 12 in others), trying to hit a goblin
a first-level archer (fighter where applicable) trying to hit a enemy with significant armor, such as a dragon. For this example, a +2 to-hit bonus is used; it could be a bonus from Dexterity (where applicable), race (halflings) or a +2 magical bow.

Brown Book D&D (1974)

From the beginning, D&D relied on charts. You used your level and the AC of the monster and tried to roll over the number to hit. However, at this stage, AC ranged only from 9 (no armor) to 2 (Plate armor and shield) and level improvements to your target number occurred every 3 (fighters) to 5 (magic users) levels. This resulted in a very small chart.

There are several other oddities, at least to our modern sensibilities, but only worth mentioning briefly:

the only bonus/penalty to hit was for missile attacks from high/low Dexterity characters.
Monsters fighting "normal men" rolled an attack for every Hit Die; the +X part was a bonus given to one of those attacks.
the party was expected to consist of about 20 PCs; this makes the "number appearing" column of the monster chart make more sense - you typically encountered 40-400 goblins!

Our examples are both rudimentary and straightforward: the fighter needs to roll a m to hit a goblin (no strength bonus existed), the archer needs a o to hit a dragon (given +1 from dexterity and +1 from a magical bow).

Basic (1977 - 1983+) [1]

Red Book Basic (TSR2014) had an extremely simple Hit Roll Table:

Because the book only went to third level, all player characters had the same target roll. Our fighter needs to roll a l (and then add 1 to it for 13) to hit the goblin (AC 6).

Our archer example however, reveals the first interesting wart of the chart system: the cap of 20. A roll of 20 was the most that could be required to hit anything at the time, and most importantly this is not a natural 20 but one after bonuses. Our low-level archer can hit both a Red dragon (AC -1) or a Gold dragon (AC -2) on a total to-hit roll of 20, which means a natural roll of r or better.

This wart significantly benefits characters with bonuses to hit, whether magical or otherwise. Further books, from Companion to Rules Cyclopedia, standardized the table to repeat the 20 exactly five times before moving on to 21. The full chart, after expansion by subsequent books (from the DM's Screen, TSR9431):

So our archer example goes even further, also hitting the "large" versions of our Red (AC -3) and Gold (AC -4) dragons, and even up to a "huge" Red dragon (AC -5), all on a natural roll of 18 or better. It's not until he tangles with a huge Gold dragon (AC -6) that the difficulty increases again (needing a natural 19).

It's also worth noting that Basic had no concept of a critical hit. A natural 20 was not a guaranteed hit; our archer can do amazing things due to the +2 to hit, but has zero chance of hitting a monster with AC -8.

Two more oddities appeared over the years. After the 20-plateau, the chart increases steadily to 30, where we have another plateau for 5 ACs. And higher-level characters trying to hit poorly-armored targets find the to-hit number dropping to zero, and then they start getting extra damage added to every hit; a 36th-level fighter only misses a goblin on natural 1 (the first mention of this) and does an extra 5 points of damage.

AD&D First Edition (1979)

1e buried the to-hit chart in the DMG. Here's the fighter chart:

A bit counter-intuitive to read, so our fighter example can illustrate. To hit a goblin (still AC 6), our fighter needs a natural roll of m (to which +1 is added for a total of 14).

Our archer example reveals that not only has Basic's 20-plateau persisted, it stretched to 6 entries instead of 5! Now our archer can hit anything from a 0 to -5 AC on a natural r or better. At the same time, many monsters have seen their armor nerfed; of the dragons, only a Chromatic (AC 0), Red or Silver (AC -1), Gold (AC -2) or Platinum (AC -3) fall in this range. To see an AC -5 we have to look into the ranks of demons and devils.

Curiously, in Appendix E of the DMG, we have our first mention of a single figure meant to simplify the chart: a column called "To Hit A.C. 0".

Despite the name, this isn't a proper THAC0 as it was later known; there are no explanations of how to use this number, except presumably to compare monsters and perhaps make it easier to look up the right column in the monster to-hit matrix. The Monster Manual does not print the number at all.

AD&D Second Edition (1989)

2e brought about the first core rulebook to mention the THAC0 mechanic. The 1996 printing puts it simply:

Figuring the To-Hit Number The first step in making an attack roll is to find the number needed to hit the target. Subtract the Armor Class of the target from the attacker's THAC0 (Remember that if the Armor Class is a negative number, you add it to the attacker's THAC0.) The character has to roll the resulting number, or higher, on 1d20 to hit the target.

THAC0 officially removed the 20-plateau wart, for better or worse, and also made it easier to preemptively include your typical bonuses and penalties for any given weapon. We could say that our fighter has a THAC0 of 20 and a +1 to hit, or we could just say he has a THAC0 of 19. They're logically equivalent. To hit that goblin (AC 6, still), the fighter needs a natural m.

To compensate for the now-missing wiggle room at the edge of your character's abilities, for the first time we see rules stating that a 20 always hits. This is a small consolation for the archer with the magic bow, however. Where a natural 18 would hit a wide range of dragons, that r will only hit AC 0 targets now - Blue and Green dragons. Coupled with a de-nerfing of monsters in 2e, a natural t is needed for almost every other well-known dragon type.

The natural 20 rule also has interesting side effects; without additional countermeasures, a small army of scrubs (goblins, kobolds, bullywugs) can outclass the most absurdly armored high-tier enemy, simply by arming themselves with ranged weapons and hoping for natural 20s.

When was THAC0 Invented?

AD&D Second edition was the first core book to make use of THAC0. However, several other books made use of it before development of 2e began in 1987. The Basic line of books, for example, had a very inconsistent attitude towards its use:

AC1, The Shady Dragon Inn, 1983, Carl Smith: includes THAC0 for pregenerated PCs, but only says what the acronym stands for, not the mechanics it implies.
B10, Night's Dark Terror, 1986, UK: Includes THAC0 and how to use it, but says "in most cases, the roll needed to hit other armour classes = THAC0 minus AC," alluding to parts of the chart it cannot replicate.
B11, King's Festival, 1989, Carl Sargent: includes THAC0 for monsters and NPCs, and explains the mechanics, suggesting that you use it and ditch the tables entirely.
But even later, some adventure modules avoid using it, while earlier modules in the same series was happy to adopt it. For example, CM7 (1986) makes use of it, but CM9 (1987) avoids it.

AD&D First Edition also included it in some publications, though with similarly fuzzy attitude for whether it augmented or replaced the tables:

RPGA 1, To The Aid of Falx, Mentzer, 1982: Might be the first use of the THAC0 acronym ever published, but how to use it is left as unexplained as in the 1e DMG.
RPGA 3, The Forgotten King, 1983: Uses THAC0 for monsters and NPCs, with a special notation, seemingly an attempt to encode the 20-plateau.
UK4, When A Star Falls, 1984: Similarly uses THAC0 and explains the mechanics with the a caveat as B10, as well as occasionally including a star in the notation to suggest looking up the chart in specific cases.

I Have A Theory

I have a theory, then, that when the 1E DMG was published in 1979, the inclusion of the monster summary tables and the column called "To Hit A.C. 0" got groups of players thinking about how to simplify the mechanic. Perhaps within the RPGA, possibly at Gencon, DMs started ironing out the details of the alternate system, and by 1982 it had pollinated back to module writers for TSR. Then there was some disagreement about whether it belonged in the Basic line, with some authors/editors using it right away and others avoiding it all the way into the 90s. By 1987, though, the faults in a direct conversion to THAC0 (inability to hit things you could on the chart) had become obvious and were "patched" during the design of Second Edition by officially adopting the common house rule of always hitting on a natural 20.

Update 7/10/2019:

With help from assorted blogs, I've tracked down a fairly early mention of Thac0 to the UCLA Computer Club's Alarums & Excursions fan magazine. In issue #32, April 1978 editor Lee Gold writes:

"One of the devices we use is THAC0 (To Hit AC 0). This is a composite of the character's type, level, strength, weapon, and other bonuses. THAC0 must be recomputed when any of this changes. THAC0 greatly speeds things up all by itself."

This is reporting what her group did at the time; whether they invented it or adopted it from elsewhere is unknown. A later issue explains it in more detail, but given that it includes class, level and strength, it's clearly not just a reference to the to-hit chart as Gygax intended with the monster chart in the 1E DMG.

But this also invalidates my theory that the 1E DMG originated the notion; this publication predated the release of the DMG by a year. Rather, Gygax may have seen clubs using this folk mechanic and included that column to make things easier for those players using it.

Other sources for this info:

Playing At the World's To Hit Armor Class Zero

D&D Third Edition (2002)

3e brought about a drastic overhaul of nearly every die-roll mechanic, calling the new mechanic DC (Difficulty Class). Saving throws, skills rolls, and even to-hit rolls became a matter of rolling 1d20, adding your bonuses, and trying to exceed the desired DC. AC was divorced from its naval wargaming roots (1st-class armor being better than 3rd-class armor) and instead ascended as it got better.

On the one hand, this made the combat roll streamlined. Our fighter needs to roll the goblin's AC (15) to hit it, adding 1 to the roll; in other words, a natural n or better.

On the other hand, this also lead to a linear progression of power and a proliferation of tiny bonuses to track. That first-level fighter almost certainly isn't adding just +1; more likely +1 from Base Attack Bonus, +3 from Strength, +1 from a feat, etc. It's not uncommon for a mid-level fighter to be adding +20 to an attack roll, and monsters have been scaled accordingly. While it's mechanically identical to a low THAC0 combined with Strength and magic, it feels different during play to lump everything into a single bonus that overwhelms the variance range of the die it modifies.

Our archer example is quite similar to the THAC0 approach, but now the number and types of dragons has exploded (presumably to provide precise challenges for a party of any level).

This approach of roll vs AC has remained through Pathfinder, 4e and 5e, for better or worse.

Which was better?

The chart system was considered clumsy, particularly for the DM, who had to look up the right row for any given monster. But PCs could jot down their own target numbers (Basic even included it on the character sheet) for reasonable ranges. Hidden in the Basic charts, however, were interesting subsystems: diminishing gains at higher levels; grace ranges in the form of target plateaus; different advancement systems for monsters.

THAC0 played comparatively quickly, once you got used to the math. But advancement in both Advanced editions became linear and with THAC0 came the need for natural 20s, which opened the door for critical hits, then the need to balance those with confirming critical hits, and things got far more complicated than a table lookup.

By 3.x, the linearity was standardized into all areas; saving throws, bonuses from stats, costs to level, skill advancement, etc, etc. Some of the heroic epic of the early editions was lost along the way.

[1]

(1, 2) To avoid going down the rabbit hole of Moldvay vs Holmes basic, I'm treating everything from the 1977-ish Red Book onward as the same lineage. This is mostly irrelevant to this mechanical deep-dive, but it's worth noting that AD&D didn't really come after "Basic" - AD&D was in parallel development for almost the entire Basic lifespan. Even the Blue Book (1974) makes reference to the not-yet-printed AD&D game.