On Identification

In my most recent AD&D 2e play report, the player characters wanted to discover the properties of a strange alien helmet they found in an old tomb. The helmet had seemingly been used to torture someone so badly that they became a ghost, so of course none of the characters dared to try it on and find out for themselves what it was capable of. The players are not quite ready to embrace the old-school approach that is bribing or bullying their henchmen and hirelings into being magic item guinea pigs, either. So, they decided to find the most capable wizard in town and ask her to identify the item for them. 

The problem? According to Table 69 of the 2e DMG (page 154), getting an NPC to cast identify should cost at least 1,000 gold pieces per item or function of an item to be identified. The DMG states that "the costs of purchasing a spell are such that it is far better for someone in the party to learn the spell. In general, the mercenary use of NPC spellcasters should be discouraged whenever possible. The player character are supposed to face challenges on their own!"

Clearly, there is an intended playstyle justification for the high cost of getting an NPC to cast an identify spell. But there is also a mechanical justification - identifying magic items in 2e is hard! Below is the description of the identify spell in 2e, with important parts bolded (PHB, page 175):

Range: 0

Components: V, S, M

Duration: 1 rd./level

Casting Time: Special

Area of Effect: 1 item/level

Saving Throw: None

When an identify spell is cast, magical items subsequently touched by the wizard can be identified. The eight hours immediately preceding the casting of the spell must be spent purifying the items and removing influences that would corrupt and blur their magical auras. If this period is interrupted, it must be begun again. When the spell is cast, each item must be handled in turn by the wizard. Any consequences of this handling fall fully upon the wizard and may end the spell, although the wizard is allowed any applicable saving throw.

The chance of learning a piece of information about an item is equal to 10% per level of the caster, to a maximum of 90%, rolled by the DM. Any roll of 96-00 indicates a false reading (91-95 reveals nothing). Only one function of a multifunction item is discovered per handling (i.e., a 5th-level wizard could attempt to determine the nature of five different items, five different functions of a single item, or any combination of the two). If any attempt at reading fails, the caster cannot learn any more about that item until he advances a level. Note that some items, such as special magical tomes, cannot be identified with this spell.

The item never reveals its exact attack or damage bonuses, although the fact that it has few or many bonuses can be determined. If it has charges, only a general indication of the number of charges remaining is learned: powerful (81% - 100% of the total possible charges), strong (61% - 80%), moderate (41% - 60%), weak (6% - 40%), or faint (five charges or less). The faint result takes precedence, so a fully charged ring of three wishes always appears to be only faintly charged.

After casting the spell and determining what can be learned from it, the wizard loses 8 points of Constitution. He must rest for one hour to recover each point of Constitution. If the 8-point loss drops the spellcaster below a Constitution of 1, he falls unconscious. Consciousness is not regained until full Constitution is restored, which takes 24 hours (one point per three hours for an unconscious character).

The material components of this spell are a pearl (of at least 100 gp value) and an owl feather steeped in wine; the infusion must be drunk prior to spellcasting. If a luckstone is powdered and added to the infusion, the divination becomes much more potent: Exact bonuses or charges can be determined, and the functions of a multifunctional item can be learned from a single reading. At the DM's option, certain properties of an artifact or relic might also be learned.

The key takeaway is this: In 2e, the caster must spend 8 hours preparing to cast the spell, as well as at least 8 hours recovering from casting the spell, and because the caster must handle each item, they are vulnerable to any curses which might befall them from doing so. Additionally, unlike in D&D 5e, for example, material components of spells are consumed unless stated otherwise (PHB, page 114) - in 5e, they're not consumed unless stated otherwise (PHB, page 203) - meaning that every casting of identify in 2e consumes a 100 gp pearl.

In 2e, when the player characters pay 1,000 gp to an NPC wizard to cast identify, they are paying for an entire day of the wizard's time (8 hours of prep and at least 8 hours of recovery), as well as reimbursing the wizard for the lost pearl, the risk of curses, and the risk of falling into a brief coma. Suddenly, 1,000 gp sounds more fair.

Also worth noting is that identify in 2e does not automatically reveal information about the item (the spell always has a chance of failure). A higher-level wizard has a better chance of determining an item's properties, and a wizard with access to powdered luckstone can discern more specific item properties like exact bonuses and number of charges. If anything, identify should be more expensive depending on the level of the wizard whose services are employed, as well as any additional components they might be using.

This guy will risk getting cursed and falling into a coma for 1,000 gp.

Identification, Made Easy

Over the editions of D&D, magical item identification has gradually become faster, more reliable, and less costly. In 3e, identify takes 1 hour to cast and requires no recovery time, and it automatically reveals all properties of a single item, including the means of activating the item and how many charges it has (if any).

I don't remember much from my brief time with D&D 4e, but from Googling around it looks like this was the edition which introduced the concept of automatically identifying items on a short rest, which brings us to 5e (DMG, page 136):

"A character can focus on one magic item during a short rest, while being in physical contact with the item. At the end of the rest, the character learns the item's properties, as well as how to use them."

The identify spell still exists in 5e, for those players who don't want to have to take the time to identify items the more accessible way. 5e's identify takes 1 minute to cast if a spell slot is used, or 10 minutes when cast as a ritual (avoiding the use of a spell slot). The 100 gp pearl is no longer consumed. The spell allows the character to "learn [the item's] properties and how to use them, whether it requires attunement to use, and how many charges it has, if any" as well as "whether any spells are affecting the item and what they are" and "which spell created [the item]." (PHB, page 252)

So, at worst, the player character can spend an hour to identify an item without error (without even being a spellcaster or having any knowledge of magic). At best, they can spend a spell slot to identify it in 1 minute (if they know identify and/or have it prepared). In the middle, they can spend 10 minutes and no other resources to identify it (if they know identify and can cast ritual spells).

From 2e to 5e, we went from 16 total hours of time spent on a single casting of identify to as little as 1 minute, from a 100 gp pearl which is consumed with every casting to a 100 gp pearl that can be used indefinitely, and from an uncertain prospect of success and a one-property-at-a-time process of discovery to an automatically thorough and accurate reading of an item with a single casting.

Much like the intended playstyle of 2e (player characters should overcome challenges on their own) informs 2e's approach to identification, the intended playstyle of 5e informs 5e's approach to identification. In 5e, player characters aren't meant to struggle with mundane adventurer tasks like figuring out what their new toys are and how to use them. Player character abilities simply work, they work instantaneously, and they do not usually have drawbacks, penalties, or restrictions.

This is especially true of spells - players can only learn a handful of spells as their characters progress, so they're unlikely to choose those which they deem "underpowered" or unreliable Thus, all spells must attempt to be equally compelling in order to attract players to choose them. (Wizards can technically learn every spell available to their class, but even their potential, from the player's perspective, is throttled by the greedy DM's draconian vice grip on scrolls and spellbooks found in treasure hoards.)

The modernization of magic in D&D comes at the expense of the fiction which the game was originally intended to emulate - conjurers can pick and choose exactly what otherwordly creature they summon, and have perfect and minute control over such creatures, transmuters can spend lifetimes polymorphed into monsters with no risk of ever truly becoming monsters themselves, and rare, powerful, and alien artifacts readily and reliably yield all of their secrets and powers in as little as a minute (or at most, in an hour, to any non-magic-using schmuck who handles them and thinks really hard about it).

This guy doesn't even know magic, but even he can force the cube to yield all of its most occult secrets during his lunch break!

Identification, Made Boring

The consequence of the modernization of magic item identification is that there's simply no reason for player characters to experiment with magic items ever again. Why would they, when, in a worst case scenario, they can simply sit for an hour and think really hard about the item to unerringly learn all of its properties?

The 5e DMG (page 136) does shoutout experimentation as a means of identification, for some reason: "Wearing or experimenting with an item can also offer hints about its properties. For example, if a character puts on a ring of jumping, you could say, "Your steps feel strangely springy." Perhaps the character then jumps up and down to see what happens. You then say the character jumps unexpectedly high."

Why special mention is given to experimentation when the mechanics almost completely eliminate the need for it is anyone's guess. In some scenarios, the party may have no spellcasters capable of casting identify, and may not have time to stop for an hour, or may have need of the item urgently, but this has never happened in one of my games. It's probably because one of my players almost always plays a wizard.

The DMG also has a variant rule which eliminates the ability to identify magic items on a short rest: "If you prefer magic items to have a greater mystique, consider removing the ability to identify the properties of a magic item during a short rest, and require the identify spell, experimentation, or both to reveal what a magic item does." 

Of course, there's no example given as to what a combination of identify and experimentation looks like, which would be helpful, given that identify, as written, is 100% accurate 100% of the time - so in what scenario would experimentation also be needed? Even in a campaign where short rest identification is eliminated, the identify spell is still foolproof. 

If the player characters can't cast identify themselves, they can probably find an NPC to do it for them, and certain official Wizards of the Coast products list the price of magic item identification being as cheap as 20 gp per item, which is trivially cheap even at 1st-level. There is still no incentive to experiment.

Identification, Made Interesting

Ultimately, I was inspired to write this post by two things - my above reading of the AD&D 2e description of identify, and Prismatic Wasteland's post on Potion Clues. 

Warren's post specifically made me think about how in 5e, the DMG provides a brief description of what every potion looks like. A potion of healing is red and glimmers when agitated. A potion of fire breath is flickering orange, and smoke fills the top of the container and wafts out whenever its opened. A potion of animal friendship is a muddy liquid containing a fish scale, a hummingbird tongue, a cat's claw, and a squirrel's hair. 

Details like this invite players to speculate as to a potion's properties, take notes on potions they've encountered in the past, and perhaps most importantly, experiment. The characters have had potions of healing before, so they are able to recognize the glimmering red liquid. They are able to intuit from the appearance of a potion of fire breath that the potion has something to do with fire. In en encounter with a wild animal, they might risk quaffing a potion of animal friendship, having deduced that it has at least something to do with animals, and might give them an advantage. 

Hilariously, the DMG specifically says "Potions are an exception [to the usual identification rules]; a little taste is enough to tell the taster what the potion does." On the one hand, we have an interesting way for player characters to deduce a potion's properties just by looking at them. On the other hand, the DMG says, "Nevermind all that - just take a little sip, which has no consequences whatsoever, and all will be revealed!"

Warren writes that making the effects of a potion less immediately transparent "rewards characters who can identify potions, encourages seeking out sages who identify them, and creates interesting decision points when they have unidentified potions in their bags but their backs are up against the wall." I would go a step farther and apply this approach to all magic items.

In fact, the 5e DMG even suggests that all magic items should be visually distinct (page 136): "The command word to activate a ring might be etched in tiny letters inside it, or a feathered design might suggest that it's a ring of feather falling."

A few pages later, the 5e DMG has a great section on "Special Features" of magic items (page 141). That section includes tables to determine who or what created the item or was intended to use it, details from the item's history, and minor properties and quirks of the item. 

There's probably enough material here for a separate post which goes into more detail on making magic items themselves interesting, so that the process of learning their properties is in turn interesting. Discovering that a war hammer is a +1 war hammer and only a +1 war hammer is never going to be interesting no matter how the DM adjudicates it. So, a DM should never simply give the players a +1 war hammer - they can buy that in a shop, if all they want is a +1 bonus (magic item shops are already almost completely devoid of wonder, so the items to be found there might as well be accordingly wonder-less). 

Instead, that magic war hammer the player characters find by chance in an actual dungeon on an actual adventure is the Black Earth Bludgeon, crafted of solid stone and bejeweled with polished rock. It was made by cultists of Ogrémoch to slay beings of elemental air and the worshippers of Yan-C-Bin. It grants the wielder uncanny senses when underground, but makes them slothful and lethargic on the surface world, beneath the open sky.

So, how do the player characters learn all of this about the seemingly humble +1 war hammer? The purpose of my approach, which I'll detail below, is for a magic item's properties to be communicated entirely within the fiction of the game, and to encourage experimentation by the player characters as the single best way to truly discern an item's various properties.

Identification, Made Complex

First, we have to entirely eliminate identification via a short rest. It's both easy and reliable, which makes it boring. Then, we have to make identification via identify less informative. I don't want identify to reveal an item's properties at all - only fictional information about the item's identity. Casting identify should only reveal the +1 war hammer's name, origin, and purpose.

At the DM's option, identify could also be made more costly, and less reliable. Let's say that each casting of identify requires an Intelligence (Arcana) check to properly identify the item, and that the DC is determined by the item's rarity, like so:

Item Rarity    Intelligence (Arcana) check DC
Common         10
Uncommon     15
Rare                20
Very Rare        25
Legendary       30
Artifact            35

Now, a 1st-level wizard, assuming that they have proficiency with Arcana and start the game with an Intelligence of 16 (+3) - although I don't mind player characters starting with higher ability scores - can somewhat reliably identify common and uncommon items, and can less easily identify rare and very rare items. By 13th-level, the same wizard probably has +10 to their Arcana skill, meaning they can reliably identify rare items, and can occasionally identify even legendary items. Only a character with expertise in Arcana can ever properly identify an artifact.

On a failed check, the caster has failed to magically divine the item's identity from the depths of time. On a successful check, the caster learns that the +1 war hammer is the Black Earth Bludgeon, crafted by cultists of Ogrémoch to slay beings of elemental air and the worshippers of Yan-C-Bin. There's still plenty to learn about this item, but the player characters have a place to start.

What's to stop a caster from simply attempting to identify an item over and over again until they succeed? Make the 100 gp pearl component consumable when used to attempt to identify a magic item. Now, every attempt comes with a cost.

Players and their characters can still intuit things about an item based on its appearance, as I described with regard to potions. In the case of the Black Earth Bludgeon, the players, drawing on their knowledge of the genre, might conclude that the bejeweled stone hammer was crafted either by dwarves or elementals. They might seek out these creatures, or NPCs who are knowledgeable about these creatures, to attempt to learn more about the item.

If the players don't intuit information themselves, they might lean on their characters' skills to do so. A character proficient in History, Religion, or jeweler's or mason's tools might recall reading about the item in An Encyclopedia of Elemental Evil, recognize certain symbols of the Black Earth carved into the hammer's head or haft, or recognize the magical stonecutting employed by Black Earth artisan-priests. 

Loosely interpreting a rule from Xanathar's Guide to Everything (page 78) - as I've done in the past - tool proficiencies can be combined with skill proficiencies to gain advantage on ability checks. In my case, I allow two proficiencies of any type to combine in this way, not just one tool proficiency and one skill proficiency - so a character with any combination of History, Religion, and jeweler's and mason's tools would have an increased chance of recognizing the origin of the Black Earth Bludgeon. I would use the same DC-by-rarity guidelines I outlined above for identify.

This eliminates the need for identify to some extent, but this is a good thing, as it rewards players for being able to draw their own conclusions, or for making characters with niche knowledge and skills. In a scenario where none of the players or characters have any idea what the nature of an item might be, identify still exists, but it's more costly, in addition to being somewhat unreliable.

So, either through player skill, character skill, leveraging knowledgeable NPCs, or falling back on the identify spell, the players and their characters now know a bit more about the item. How do they figure out what it actually does?

Identification, Made Fun

They experiment! 

The Black Earth Bludgeon was crafted by cultists of Ogremoch to slay beings of elemental air and the worshippers of Yan-C-Bin, so the player characters might try using it to fight air elementals, Howling Hatred cultists, and flying creatures. Against these foes, the hammer feels twice as heavy when delivering blows, but not any more difficult for the wielder to swing. These enemies crumble against the hammer's awesome might. Perhaps, against these enemies, it grants additional bonuses to attack and damage rolls, or advantage on attacks. It might cause flying enemies to fall to the ground immediately, turn enemy cultists to stone, or banish air elementals back to their home plane.

The Black Earth Bludgeon is clearly attuned with the earth itself, so the player characters might try bringing it underground. Suddenly, the wielder has an intuitive sense of how deep they are beneath the surface, and which direction is north. Secret doors and maybe even dungeon walls crumble beneath the hammer's blows. Perhaps the hammer's head pulls in the direction of precious stones.

When they return to the surface, the wielder notices that the hammer feels twice as heavy. Their every step is a burden. The endless, wide open sky feels like it's crashing down around them. They're becoming agoraphobic.

In this way, the magic item becomes another character in the game. The player characters have to spend time with the item, devote resources to uncovering its secrets, and most importantly, play with it. 

For this to be truly interesting and fun, the magic item should likewise be interesting and fun. It should have a unique appearance from which players can draw conclusions, an origin and history which the player characters and NPCs might be aware of, and multiple features, including both benefits and drawbacks, which are discovered only through continual experimentation in a variety of scenarios. Player characters cannot simply find an item, sit down with it during a coffee break, and immediately decide if they want to use it or throw it in a bag of holding to sell at the local Magic Mart.

This playstyle isn't for everyone. To some extent, it's working directly against the intended playstyle of modern D&D, which might be too much for some DMs and players. There's presumably a reason why WotC designed this element of the game this way. I'd guess that it's probably because players complained about identification being too onerous in the past. WotC loves to design by committee, rather than by vision.

I'm sure there are plenty of players who simply want to hit the identify button and be done with it, and get to using their new toy, but sometimes it's more fun to start playing with the toy without knowing what it might do.

On Specialists

I love wizards. It used to be that when I made a character for a Dungeons & Dragons game, I waited to see what everyone else was playing first, and picked a class which met some unfulfilled role in the party. Because, until recently, there was only one Intelligence-based class in 5th edition, this usually meant I'd be playing a wizard.

One thing I've never liked about 5e wizards (and a lot of 5e's spellcasters in general) is how same-y they all feel. Gone are the days when the Dungeon Master picked which spells the wizard had in their spellbook. Players now have total freedom over their spell selection. The scope of the game has narrowed at many tables to be primarily about combat - some might argue it's always been that way, but early dungeon-centric campaigns like Castle Greyhawk seem to have been at least equally about exploration. There is a proliferation of content on YouTube which concerns itself primarily with ranking the "best" spells, which means experienced players who want to optimize their characters tend to pick the same "must-have" spells every time - mage armor, magic missile, shield, misty step, counterspell, fireball, etc. 

(As an aside, this is why I like playing with new players, and try not to steer them towards the "good" spells unless they ask for my advice - they haven't yet closed themselves off to the possibilities of less "optimal" spells, and it's often a delight to watch them find ways to use those spells.)

Certain subclass abilities incentivize picking spells from an Arcane Tradition's school (Arcane Ward is recharged by casting abjurations, diviners can regain spell slots from casting certain divinations, Sculpt Spells only works with evocations, etc.), but even the Savant feature that each specialist gets only incentivizes learning spells from one's specialized school of magic outside of the leveling system (the feature reduces the gold cost of scribing spells, so a specialist wizard is better off learning other spells on level up and then getting the ones they want from their specialized school by other means).

All wizards tend to have the same spells, aside from those oddballs picked up from spell scrolls in treasure hoards and from enemy wizards' spellbooks. An abjurer's Arcane Ward might make them play slightly differently than a diviner with their Portent or an evoker with Sculpt Spells, but by and large, they're all probably going to have very similar repertoires.

Specializing Specialists

Advanced Dungeons & Dragons 2nd edition did specialist wizards differently. Specialists gained additional spells per level from their chosen school, received a bonus to saving throws against spells of their own school, imposed a saving throw penalty on their enemies when they cast spells of that school, had an improved chance to learn spells of their school, and could more easily create new spells within their school. The trade-off was that each school had a school of opposition, from which the specialist was forbidden to learn spells.

Later editions did away with this aspect of specialists along with most other restrictions on player character choice, and wizards became meta-chasing casters like everyone else, casting whatever spells D&D YouTubers told them were most optimal, albeit with a few bells and whistles of varying usefulness.

In my 5e games, I offer a house rule for wizards who choose the abjuration, conjuration, divination, enchantment, evocation, illusion, necromancy, or transmutation Arcane Traditions:

Starting at 2nd level, whenever you gain a wizard level, you can add one wizard spell to your spellbook, in addition to the two you already learn (as described in Spellcasting). This additional spell must be of the school associated with your Arcane Tradition. Additionally, when you make a saving throw against a spell of the school associated with your Arcane Tradition, you have proficiency on that saving throw attempt, even if you normally lack proficiency. When you cast a spell of the school associated with your Arcane Tradition which forces a saving throw to resist its effects, you can give one target of the spell disadvantage on its first saving throw made against the spell.

Additionally, you have advantage on Intelligence (Arcana) checks to learn spells from your Arcane Tradition’s school from spell scrolls.

It's a small bonus, but I think that it, along with the subclass features already included in each Arcane Tradition, gently nudges wizard players towards leaning into their specialization, if they have one.

Side Note: I also imported a spell research downtime activity based on 2e's system, which benefits specialists who choose to create new spells within their school of magic, but that is probably beyond the scope of this post.

Schools of Opposition

Another way to..."encourage" a player to lean into their wizard archetype is to reintroduce schools of opposition - that is, schools of magic from which a specialist cannot learn any spells. This is not something that fits nicely with new school playstyles, and more experienced players who are familiar with "must-pick" spells and the "meta" of wizards may chafe at the restriction. It may be worthwhile to offer the aforementioned house rule as an optional bonus if the player opts-in to locking themselves out of opposition schools. If that isn't to their liking, the DM can just leave specialist wizards as they are.

The schools of opposition are as follows:

• Abjuration: Illusion and Transmutation
• Conjuration: Divination and Evocation
• Divination: Conjuration
• Enchantment: Evocation and Necromancy
• Evocation: Conjuration and Enchantment
• Illusion: Abjuration, Evocation, and Necromancy
• Necromancy: Enchantment and Illusion
• Transmutation: Abjuration and Necromancy

I'll admit there's not much rhyme or reason to this aside from being how it was done in 2e, but one can derive some meaning from it: Wizards schooled in the more subtle arts of enchantment and illusion don't have the skill or inclination to invest time and gold into learning crude battle magic like abjuration and evocation, or the dark arts of necromancy - why reanimate a corpse when you can trick the living into doing what you want? Inversely, necromancers are much too isolated in their gross crypt laboratories to learn magic that relies on understanding people - and why would a wizard preoccupied with the undead choose to learn spells which largely don't affect them? Summoned creatures get in the way of evoker's blasting magic, and they'd rather blow someone up than charm them. Diviners know too much about the nature of the multiverse to start conjuring creatures from who-knows-where. Conjurers use refined, artisanal magic to create more sophisticated effects than the instantaneous blasts of evocation, and they have no need for divination, as they can simply acquire information from the extraplanar creatures they summon, or traverse the planes and find knowledge for themselves.

When one looks at the schools of magic and which spells belong to which school, it becomes apparent how blurry the lines are - why is necromancy, the magic of life and death, not responsible for curative magic? Why is fear, a spell which triggers an emotional state in its targets, an illusion and not an enchantment? 

I think of this as a feature of D&D's magic system, not a bug. The schools are not indicative of what kind of magic is possible within each school, but rather the approach used to achieve the magic's effects. Necromancy is not responsible for healing magic because it primarily deals in "dark" magic. Fear is an illusion because it works like that scene in Fellowship of the Ring where Gandalf makes himself appear very big in order to scare Bilbo back to his senses ("I am not trying to rob you! I'm trying to help you.").

When one thinks of magic this way, its various schools become almost like monastic traditions - wizards are disciplined in a certain approach to magic, and stand in opposition to incompatible disciplines. An illusionist might see a necromancer as an affront to magic itself. The two might engage in battle whenever they meet, and their approaches would look very different - the illusionist may use their magic to confuse and misdirect the necromancer's simple-minded minions, then shatter the rival wizard's mind with powerful fear effects, or the necromancer might come prepared with undead specifically created to see through such magic, forcing the illusionist to resort to trickery and obfuscation to escape the enemy's minions. They're not both going to stand 120 feet apart slinging fire bolt and blocking with shield.

I don't care much for D&D YouTube - content there is often preoccupied with player character builds, ranking character options into tier lists based on their "power" in a featureless white room combat, or trying to be "funny" ("This is what it would be like if Skyrim players played D&D!"). One of the only channels I really enjoy is WebDM, and they have a great series on the schools of magic in D&D, which is a good place to start thinking about these things.

Other Specialists

It isn't just wizards who need to work within the confines of their discipline in 2e. Priests' spells were broken into spheres, with each deity granting and barring access to specific spheres, while druids were simply priests of natural spirits and old gods, with access to spheres that suited their style of worship.

There's a world in which every Divine Domain, Druid Circle, Sorcerous Origin, Otherworldly Patron, and maybe even Sacred Oath and Ranger Conclave has its own, totally unique spell list. I like Domain, Circle, and Oath spells because they offer a happy middle ground, wherein spellcasters automatically have convenient access to the spells that fit their theme, but can still prepare the "optimal" spells they might prefer (although this usually means that the subclass-specific spells don't see much use unless they're particularly "good").

Even then, this method isn't applied consistently. Otherworldly Patron spells are added to the warlock spell list, but the warlock still has to choose them as known spells (foregoing potentially "better" vanilla warlock spells), and some rangers and sorcerers get bonus spells, but not all.

One solution to this is to give all Ranger Conclaves and Sorcerous Origins bonus spell lists (which I've done in my games). An option for warlocks would be to give them their Otherworldly Patron spells for free (which I've done as well). But how do you encourage players to actually use these spells over the "optimal" ones?

Like with wizards, some subclasses have moved towards solving this problem by granting certain benefits when casting subclass-specific spells (and others that fit with the theme) - Life domain clerics heal better, Circle of the Shepherd druids have more durable summons, Storm sorcerers can cause damage to enemies near them whenever they cast a lightning or thunder spell, etc.

I encourage players to lean in to the theme of their subclass's specialization. When I play an Aberrant Mind sorcerer, I'm only taking spells which can pass for psionic effects. When I play a Divine Soul sorcerer, I'm only taking cleric spells. When I play a Draconic sorcerer, I'm playing an elementalist. My Shadow sorcerers are necromancers. My Storm sorcerers only use weather magic, etc.

I find that the world feels much more rich when spellcasters of a given type use different approaches to magic and a different selection of spells. Each wizard, cleric, druid, or the like feels more unique and memorable. Ultimately, this is something players most likely have to decide they want for themselves. The DM can't force this kind of playstyle, but they can offer incentives for players who embrace it. 

On Augury

“What are your fees?" inquired Guyal cautiously.

"I respond to three questions," stated the augur. "For twenty terces I phrase the answer in clear and actionable language; for ten I use the language of cant, which occasionally admits of ambiguity; for five, I speak a parable which you must interpret as you will; and for one terce, I babble in an unknown tongue.”

― Jack Vance, Tales of the Dying Earth

In my long-running 5th edition Dungeons & Dragons campaign, which ended just a few months ago, one of the player characters was a Path of the Ancestral Guardian barbarian (Xanathar's Guide to Everything - side note: perhaps my favorite barbarian Primal Path, and arguably more busted than the infamous Bear Totem Warrior). At 10th-level, the Ancestral Guardian barbarian gains the feature "Consult the Spirits", which allows them to cast augury or clairvoyance once per rest without using material components or a spell slot.

I don't think I've ever had a player choose to take augury, but in this case the barbarian got it for free. Understandably, they wanted to use it. A lot.

"I hate this spell, but I'll try," I said.

Babbling in an Unknown Tongue

The problem with augury, more so than other divination spells, is that is asks the Dungeon Master to predict the future of a game which relies heavily on dice rolls.

In 5e, the spellcaster "receives an omen...about the results of a specific course of action" and the DM chooses between Weal ("something good will happen if you do this"), Woe ("something bad will happen if you do this"), Weal and Woe, or Nothing. Given the nature of dice, I find it hard to give any answer other than "Both...maybe?"

The spell is much the same in Advanced Dungeons & Dragons (2nd edition) - along with 1st edition, my fallback when something in 5e needs more advice, context, or (dare I say?) mechanics - except that the spell is, in classic AD&D fashion, slightly more difficult to cast effectively, having a much greater chance of giving a "meaningless reply", and requiring 1,000 gold pieces-worth of components!

Interestingly, the 2e description includes this example of use:

"...if the question is 'Will we do well if we venture to the third level?' and a terrible troll guarding 10,000 sp and a shield +1 lurks near the entrance to the level (which the DM estimates the party could beat after a hard fight), the augury might be: 'Great risk brings great reward.' If the troll is too strong for the party, the augury might be: 'Woe and destruction await!'

Interpret As You Will

Two potential uses of augury come to mind which don't fundamentally change the mechanics of the spell: telegraphing danger and enhancing a piety system (as described in Mythic Odysseys of Theros, page 36).

The above play example included in 2e demonstrates the spell's potential to telegraph danger. If the DM is running a sandbox, in which challenges and locations of varying difficulty and danger are usually scattered about the area of play somewhat randomly, and the players can well and truly try to go anywhere and do anything, a savvy cleric, druid, wizard, etc. might use this spell to effectively assess the "appropriateness" of a given expedition for characters of their level.

Unfortunately, the spell only applies to courses of action that the party plans to take within the next 30 minutes. It's unclear if, for example, the party plans to set out to the incredibly dangerous Kandalifi Jungle region in the next 30 minutes, whether or not the spell accounts for the entirety of that journey or only the first 30 minutes of it. If it's the latter, then the spell is good only for assessing the danger of exploring the next few rooms of the dungeon (which can also be useful). I'd lean towards interpreting the spell as applying broadly to the currently-planned course of action, so long as the party plans to take that action imminently. Here, I'm favoring leniency and generosity because this is an oft-neglected spell which could use some love from the referee.

5e's optional piety system is laid out in the Magic: The Gathering setting book, Mythic Odysseys of Theros (a supplement which, like Guildmaster's Guide to Ravnica, is useful but unfairly maligned and neglected merely for being a Magic book). Theros provides a comprehensive list of its gods, their goals, worship, suggested alignment, classes, cleric domains, backgrounds, and - most importantly - ways to earn and lose piety.

In an ideal world, this system would be a part of every D&D book going forward (given it details the gods in some way). If the DM is running a game in the Forgotten Realms or another established setting, it's easy enough to extrapolate ways to earn and lose piety - as well as piety's benefits - from Theros's blueprint and the descriptions of the gods in the chosen setting. If the DM is running a custom setting, it's worth it to consider these elements for each of the setting's gods (or at least the ones that are relevant to the campaign).

So how does this relate to augury? Well, a character who is concerned with piety could use the spell to determine whether a given course of action would gain or lose their god's favor, or leave it unaffected. Piety works just as well for a cleric as it does for a druid concerned with placating nature spirits or a wizard who is particularly devout towards a deity of magic or knowledge. If the DM was to introduce a piety system and tie augury into it in this way, I would recommend opening the spell up to godly paladins as well (perhaps the spell could also telegraph whether the paladin is likely to be put in a situation where their morals or oath is put into jeopardy, although this might be asking too much of the DM).

Admitting for Ambiguity

5e's description of the augury spell has another element I didn't mention, which further complicates the issue:

"The spell doesn't take into account any possible circumstances that might change the outcome."

This is basically the DM's "get out of jail free" card for the spell. The description specifies the casting of additional spells or the loss or gaining of a companion. It's obvious enough that if the player casts this spell before proceeding to fight the fearsome troll beyond the stairs to the dungeon's third level and gets the "Weal" result, and then the party's fighter abruptly decides to go home, that the odds of defeating the troll have probably drastically changed. Similarly, a result of "Woe" might be significantly altered by the druid casting flame blade immediately afterwards. But what if the troll is accompanied by an ogre mage who wins initiative and opens the encounter by devastating the party with a cone of cold? The DM can anticipate the added difficulty posed by the ogre mage's presence, but they cannot foresee bad luck such as the ogre going first in combat, or a majority of the party failing their saving throws against its spell.

One way to account for this uncertainty is to tie a certain benefit to the spell. Perhaps a "Weal" result carries with it a Portent, akin to the School of Divination's 2nd-level feature, which the caster then gets to utilize at a time of their choosing within the next 30 minutes. A "Woe" result might grant the DM a Portent instead. Instead, maybe the caster gains inspiration from an auspicious omen, or - one of my personal favorite house rule mechanics - the DM gains reverse inspiration from a bad omen, i.e., the DM can force a player to have disadvantage on a roll or reroll a result of their choice at a later time.

These are two ways the DM can gently nudge an uncertain outcome towards greater certainty, if so inclined. The question is, is it fair for the DM to then have say over the result of the spell's casting when it's now being tied to a meta-currency? Is it more fair to roll a d4 to determine the result and apply the benefits and drawbacks based on that? Should the spell remain a ritual? Should the components be consumed?

For Twenty Terces

I included the Jack Vance quote from Tales of the Dying Earth at the top of this post because it made me think of augury as transactional, which made me think of material components. The spell requires 25 gold pieces-worth of "specially marked sticks, bones, or similar tokens", which are not consumed in the casting of the spell. What if greater components could garner greater portents? What if augury could be cast at higher spell levels? When viewed this way, augury is less a neutral reading of the situation at hand and more a ritual performed and offering made to gain a god's favor.

My revised version of augury is as follows:

Augury

2nd-level divination (ritual)
Casting Time: 1 minute
Range: Self
Components: V, S, M (specially marked sticks, bones, or similar tokens worth at least 25 gp and up to 2,500 gp)
Duration: Instantaneous, or 30 minutes

By casting gem-inlaid sticks, rolling dragon bones, laying out ornate cards, or employing some other divining tool, you receive an omen from an otherworldly entity about the results of a specific course of action that you plan to take within the next 30 minutes. The DM chooses from the following possible omens:

• Weal, for good results
• Woe, for bad results
• Weal and woe, for both good and bad results
• Nothing, for results that aren't especially good or bad

The spell doesn't take into account any possible circumstances that might change the outcome, such as the casting of additional spells or the loss or gain of a companion.

Alternatively, you can use this spell to determine whether a given course of action will be looked upon favorably or unfavorably by any one god.

If you cast this spell using a 2nd-level spell slot, you can instead choose to roll a d20 and record the number rolled. You can replace any attack roll, saving throw, or ability check made by you or a creature that you can see with this foretelling roll. You must choose to so before the roll. The foretelling roll can be used only once. After 30 minutes, you lose the foretelling roll if it is unused. This use of the spell consumes the material components.

If you cast the spell two or more times before completing your next long rest, there is a cumulative 25 percent chance for each casting after the first that you get a random reading. The DM makes this roll in secret.

At Higher Levels: When you cast this spell using a 5th-level spell slot and material components worth at least 250 gp, you can choose to roll two foretelling rolls, and you lose any unused rolls after 8 hours. When you cast this spell using an 8th-level spell slot and material components worth at 2,500 gp, you can choose to roll three foretelling rolls, and you lose any unused rolls after 24 hours. Both of these uses of the spell consume the associated material components.

Classes: Cleric, Druid, Paladin, Wizard