D&D 5e Spell & Magic System Design Questions for Dungeon Masters

A flexible magic system works best when it defines the decisions magic creates before it defines a catalog of effects. Decide what magic costs, what makes it risky, how precise it can be, and which limits keep one declaration from solving every problem.

You do not have to choose between rigid spell lists and total freeform casting. A modular structure such as verb + target + scale can give players expressive room while keeping each parameter understandable. “Move,” “reveal,” “transform,” or “protect” can combine with range, duration, and intensity, with stronger combinations demanding more time, resources, risk, or exposure. Then playtest situations rather than isolated spells. Give several players the same obstacle and watch whether they produce different solutions, whether one combination dominates, and how long adjudication takes.

If dangerous magic is part of the premise, make failure create consequences that change the fiction rather than merely wasting a turn. If magic is meant to feel professional and reliable, place uncertainty in cost, collateral effects, or preparation instead. The system is balanced when its limits repeatedly create interesting choices.

Change one layer of the magic system at a time unless the campaign is prepared to rebuild classes, monsters, items, and encounter assumptions together. Spellcasting touches too many parts of D&D 5e for a total replacement to be a cosmetic house rule. The safest redesign begins by defining what experience the new system is meant to create.

Separate access, cost, risk, and effect. You can make magic feel stranger by changing how spells are learned, what casting consumes, what consequences accompany powerful magic, or how reliable certain effects are. Altering all four simultaneously makes it difficult to identify what caused a balance problem.

Protect character identity. If a new universal magic procedure lets every character reproduce the strongest class features, classes built around spell access lose much of their reason to exist. Likewise, if every spell gains severe random backlash, characters whose primary contribution is spellcasting may pay a tax others do not.

Test the replacement against ordinary play: exploration, social scenes, a routine combat, a boss encounter, and recovery afterward. Record exceptions on one clear rules sheet.

A successful alternate magic system changes decisions in a deliberate way. If it only adds rolls, tables, and penalties without creating different choices, it has increased complexity rather than transformed magic.

Magic “stunts” can work in D&D 5e if they extend the fiction of a spell without turning every spell slot into permission to reproduce any other spell. The safest approach is to judge the proposed stunt by effect, cost, and precedent rather than by how imaginative the description sounds.

First ask whether the request is cosmetic, a small situational extension, or a new mechanical benefit. Cosmetic changes are easy to allow. Small benefits can often be resolved with an ability check, a reduced duration, a limited target, or a tradeoff. If the stunt duplicates a different spell, bypasses a major limitation, or creates a reusable combat advantage, it should usually require a more substantial cost or become a formal homebrew option.

Keep rulings local unless you want them to become precedent. “The fire spell can ignite this oil-soaked rope” is different from “this spell can always destroy carried equipment.” State why the stunt works in this situation.

The goal is to reward players for interacting creatively with the environment while preserving recognizable boundaries. Flexible magic feels powerful when circumstances matter; it becomes unstable when every limitation can be negotiated away.

No, but fixed lists solve an important usability problem: they reduce the number of choices a player must generate and evaluate at the moment of casting. A completely open magic system can be elegant in theory and still become slow if every spell requires fresh invention during play.

You can preserve flexibility without returning to a traditional list by constraining the search space. Use domains, verbs and nouns, effect templates, power bands, or a small set of parameters. “Control + fire” gives a player room to improvise while narrowing the decision far more than “create any magical effect.”

Separate design-time freedom from play-time speed. Players can customize a small repertoire between sessions, prepare a few signatures, or record successful improvised effects for reuse. The system remains open, but common decisions become familiar.

Also define limits before adding possibilities: what magic cannot do, how scale increases cost, and what risks appear when someone pushes beyond a safe effect. The goal is not avoiding lists as a matter of purity. It is giving players enough structure that creativity produces decisions instead of cognitive overload.

Start by defining what new decision the point system is supposed to create. If it only converts spell slots or another resource into a different number without changing when, why, or how players spend it, the redesign may add bookkeeping without adding meaningful play.

Use one unit, clear costs, and a predictable recovery rule. Then identify what prevents a player from spending the entire pool on the strongest available effect. Possible constraints include escalating costs, per-effect limits, preparation requirements, diminishing returns, or incentives to reserve points for reactions, defense, utility, or recovery.

Test the extremes before the average case. Build a character who hoards every point, one who spends everything immediately, and one who repeatedly buys the most efficient option. Run each through a short adventuring period and a long one. If one behavior consistently dominates, the resource economy needs another pressure or a different cost curve.

Also compare the new system with the structure it replaces. The important question is not whether the point totals look mathematically elegant, but whether the players face more interesting choices without gaining an obvious way to bypass the pacing, specialization, or trade-offs the rest of the game assumes.

Vancian-style spell preparation shapes play by making access to magic a planning problem: characters must decide which tools to make available before they know exactly what the day will demand. Replacing that structure changes more than convenience; it changes uncertainty, specialization, resource pressure, and how distinct different kinds of spellcasters feel.

Before designing an alternative, identify the experience you want. Spell points create flexible expenditure but can encourage repeated use of the most efficient effects. Fully free casting increases responsiveness but reduces the cost of poor preparation. Cooldowns make individual spells easier to reuse and shift attention away from the adventuring day. Fatigue or risk-based casting can make magic feel dangerous even when the character still has options.

Change one axis at a time: selection, frequency, cost, or risk. Then examine what happens to high-impact spells, healing, utility magic, and encounter pacing. The replacement should also have a fictional explanation—memorized formulae, spiritual favor, accumulated strain, unstable channels—so the new mechanic changes how magic feels in the world as well as how often buttons can be pressed.

Custom spells are most useful when they express something distinctive about the campaign's magic rather than simply supplying stronger versions of existing effects. A Dungeon Master should define what the spell accomplishes, why existing spells are insufficient, and how its mechanical cost compares with similar options.

Begin with the spell's level, casting time, range, duration, components, concentration requirement, and targeting rules. These constraints often influence balance as much as the immediate effect. Compare damage, control, mobility, information gathering, or utility with established spells of approximately equivalent power.

Consider whether the spell belongs to an NPC, a villain, a magical tradition, or the player-accessible spell lists. An ability designed for a particular antagonist does not always need to become a general spell available to every character.

Playtest combinations with ordinary class features, metamagic, multiclassing, and repeated casting. Pay particular attention to effects that bypass major obstacles, deny actions, or create advantages lasting well beyond their initial casting.

Give the spell a recognizable fictional identity. Unusual gestures, material requirements, side effects, or cultural associations can establish its place in the world without introducing unnecessary mechanical complexity.

Revise the design when actual play reveals disproportionate effectiveness or unintended interactions.

Make magic feel harder by increasing its predictability in the fiction before adding new mechanics. A hard-feeling system needs understandable limits, repeatable causes, and consequences that players can anticipate; D&D's existing spell rules already provide many fixed effects, so the larger change can come from how the world interprets, regulates, and supplies magic.

Choose a few constraints and apply them consistently. Spell components can matter as logistics, visible casting can attract witnesses, certain effects can leave recognizable traces, and institutions can restrict or license particular practices. Define what trained characters know about these patterns so players can reason from them instead of discovering exceptions only after committing to a plan.

Avoid adding a separate subsystem for every spell. A handful of stable principles will usually create more coherence than dozens of special-case tables. For example, if teleportation always requires a known anchor, necromancy always leaves a detectable residue, or rare components create procurement problems, those rules should produce both advantages and obstacles consistently.

The system feels harder when players can form reliable expectations and exploit them. Complexity is optional; dependable limits are the essential ingredient.

Charm Person is a good example of a spell that can look weaker than it is if “Friendly” is treated as meaning almost nothing, but it becomes too strong if Friendly is treated as mind control.

Under the current rules, a creature that fails its save is Charmed and Friendly toward you; the effect ends if you or your allies damage it, and when the spell ends the target knows it was charmed. That leaves meaningful social space without making the creature obedient. Before numerically improving these bad spells, I would check what the table is actually doing with their terms. For Charm Person, a favorable attitude should affect willingness to listen, negotiate, or grant something reasonable, while preserving the target's interests, loyalties, and self-preservation.

If it still underperforms in your campaign after that, I would favor a narrow change, such as making the social benefit more explicit in appropriate situations, rather than increasing duration or turning it into compulsion. A useful tune-up should create better decisions, not remove the scene the spell was meant to enrich.

Rules reference: D&D 5e SRD 5.2.1 — Charm Person; Charmed condition; Friendly attitude.

Do not weaken a spell until you can identify the repeated situation in which it is causing a problem. A spell can look overwhelming because encounters keep presenting its ideal targets, ranges, formations, or resource conditions. Changing the spell first may treat the symptom while leaving the campaign pattern untouched.

Track what actually happens for several uses. Does the spell end encounters before other characters can contribute? Does it dominate only when many weak enemies cluster together? Is the party reaching every fight with most resources intact? Does the effect create a rules problem, or does it simply feel stronger than expected? Those diagnoses lead to different solutions.

Before rewriting mechanics, vary encounter structure. Use different numbers of opponents, distances, terrain, objectives, time pressure, and opportunities for multiple encounters between full recoveries. Do not design every scene specifically to counter one character; the goal is variety, not negation.

If the same spell still produces an unwanted pattern across varied situations, make the smallest house-rule change that addresses that pattern and tell the players before it matters. Test one variable at a time so you can tell whether the change actually solved the problem.

Experienced players may need more adaptation than new players because they already have strong instincts about when spell resources are scarce, which effects are efficient, and how much power can be concentrated into one encounter. A mana system can invalidate those habits even when the individual spells remain unchanged.

Introduce the conversion with explicit examples. Show the cost of a routine low-level effect, an expensive high-impact effect, and what a typical adventuring sequence looks like under the new economy. Then let players rebuild choices that depended heavily on the old resource structure if the conversion changes their role unexpectedly.

Watch for two opposite problems during the first sessions: players hoarding mana because they still think in slots, and players discovering one repeatable effect that becomes disproportionately efficient under a shared pool. Both are adaptation signals, not necessarily proof that the entire system is broken.

A short trial period helps. Track which spells are repeatedly ignored or spammed and ask players where their old expectations misled them. That feedback is especially useful with veteran players because they can identify exactly which familiar resource assumptions no longer match the new system.

Spell slots distribute magical resources into separate levels, while spell points let a caster spend from a more flexible pool. That difference changes planning, efficiency, and how frequently powerful spells can be repeated. D&D's long association with leveled spell resources also gives the system a recognizable identity, although attributing one definitive motive to its designers would be speculative.

With slots, a character may have several lower-level resources but few higher-level ones. The allocation constrains choices before a spell is selected and encourages using different spell levels across an adventuring day. A point pool lets players decide more directly how much of their budget to concentrate on expensive effects.

Flexibility can be appealing, particularly for players who dislike unused low-level slots. It can also increase the attractiveness of repeatedly selecting a particularly efficient spell. A viable point-based alternative must therefore review casting costs, maximum spell levels, recharge frequency, and the interactions of class features with the new resource model.

Neither system is inherently better for every campaign. Slots encourage planning within predefined resource bands; points emphasize flexible allocation. Before replacing one with the other, decide which decisions the table wants to make and playtest against varied encounter patterns. The important design question is how the resource structure shapes player choices, not which accounting method looks simpler.

Dual concentration can work as a defining homebrew feature, but it should be a tightly constrained exception rather than permission to combine any two concentration spells. Normally, starting concentration on another spell ends the first, so bypassing that limit opens combinations the individual spells were not balanced around.

Separating the spells into one beneficial effect on allies and one harmful effect on enemies is not much of a safety valve. Bless combined with Bane already changes a large number of rolls at once, while control, summoning, defensive, or mobility spells can create stronger combinations without dealing any direct damage. A ban on damaging spells therefore does not address the main risk.

I would make the second concentration slot curated. Let the item hold only a short list of approved spells, or impose both a spell-level limit and narrow categories that you can actually test. If this is a major class feature, limited uses or a sufficiently high class level can also give its power an appropriate cost.

Define the failure procedure before play. Specify who counts as concentrating on the stored spell, what happens when the caster takes damage, and how either effect can end. A conservative version is to require separate concentration checks for the two effects and end both if either concentration is lost.

Most importantly, test spell pairs rather than spells individually. If one combination becomes the obvious choice in nearly every encounter, narrow the allowed list instead of adding more complicated universal restrictions.

The safest way to broaden this spell is to preserve its command economy and control cost while varying the undead’s role. The baseline spell turns suitable remains into a Skeleton or Zombie, lets the caster command creatures made by the spell with a Bonus Action while they are within 60 feet, and requires control to be renewed after 24 hours.

Higher-level casting expands the number controlled. For an “Animate Dead spell variant,” I would not simply open the door to arbitrary undead stat blocks. That makes the spell’s power depend on which corpse happens to be available and can produce huge swings. A better approach is to define a small set of templates such as skirmisher, brute, guard, or crawler, then keep their expected durability, accuracy, and action economy in the same neighborhood while changing movement, damage type, or utility.

That also helps the fiction. Different corpses can look and behave very differently without forcing you to rebalance every monster entry. Playtest the variant by comparing total actions, effective hit points, and damage across several rounds, then ask whether each option creates a different tactical choice rather than a strictly superior summon.

Rules reference: D&D 5e SRD 5.2.1 — Animate Dead spell.

Make the warning turn part of the spell’s power budget. A delayed spell is interesting because the players can see danger before it resolves, so the response options need to be understandable and meaningful.

Define four things explicitly. First, what creates the threatened area or target when the spell is cast. Second, exactly when it activates. Third, what can interrupt, weaken, or avoid it before then. Fourth, what happens if the caster becomes incapacitated, moves, loses Concentration, or no longer meets the spell’s conditions.

The delay should buy something. If the caster spends an action now and the effect can be canceled by any trivial hit, the spell will rarely be worth using. If the later effect is devastating and cannot be influenced, the warning is mostly theatrical. Aim between those extremes: players might leave the area, break line of effect, disrupt an anchor, force the caster to move, or reduce the final effect rather than automatically cancel it.

Avoid adding Exhaustion merely to make the spell feel costly unless that cost is central to the design. Exhaustion affects much more than this one casting. A limited use, consumed component, self-damage, or temporary vulnerability can create a clearer price.

A good delayed spell turns the next round into a visible tactical problem.

A funny random spell-effects table works best when its outcomes are surprising, playable, and varied rather than uniformly disruptive. Comedy should emerge from unexpected consequences that characters can react to, while the table preserves enough mechanical clarity for the Dungeon Master to resolve effects consistently.

Decide when a roll occurs and whether the table supplements an existing feature or introduces a new house rule. Do not assume that every spell automatically produces an additional effect unless that procedure is explicitly part of the design.

Mix several kinds of comedy. A spell might briefly transform nearby speech into theatrical proclamations, cause harmless objects to imitate their owners, or make a dramatic magical manifestation wildly disproportionate to an otherwise ordinary effect. Some entries can create temporary obstacles or unexpected opportunities.

Give mechanical outcomes clear durations, affected areas, and limits. Avoid effects that permanently disable characters, invalidate critical decisions, or repeatedly punish the same participant without proportionate benefit.

Use the table's context to influence its contents. Fey magic, a malfunctioning academy, and an eccentric wizard's laboratory should generate different kinds of absurdity.

Test the most disruptive entries before play. An entertaining random result creates a new situation the players can engage with; it should not require abandoning the encounter or suspending the game while the DM invents missing rules.

A larger Wild Magic table should create more varied magical consequences, not simply increase the number of results. Expanding a surge table to one hundred entries is useful when the additional outcomes produce meaningful differences in tone, tactical circumstances, and fictional complications.

Begin by defining the campaign's interpretation of uncontrolled magic. It might be whimsical, frightening, unstable, transformative, or connected to forces that have recognizable behavior. That identity should guide the entire table.

Distribute results across several functions: cosmetic transformations, environmental changes, temporary advantages, disadvantages, creature interactions, movement effects, resource alterations, and persistent complications. Avoid filling most entries with damage or random spellcasting.

Consider the probability of particularly dangerous outcomes. A table used frequently during ordinary play should not make catastrophic consequences disproportionately common unless that is an intentional campaign feature.

Each entry should resolve quickly enough to avoid repeatedly interrupting play. Effects requiring complicated bookkeeping may be better reserved for rare results.

Finally, test results at low and high character levels. An outcome that is amusing for experienced characters might overwhelm beginners, while another may become irrelevant later.

Variety matters most when the consequences become events the characters can interpret and respond to.

The core idea can fit a 3rd-level spell, but the reaction trigger and damage-redirection rules need to be much more precise before its balance can be judged reliably.

The first problem is timing. A reaction needs an identifiable trigger, so decide exactly when the teleport happens. For example: a creature within 30 feet of the beacon can react when it is targeted by an attack, before the attack roll is made. The creature teleports to the beacon's space and the beacon moves to the creature's former space. That makes the positional exchange understandable at the table.

I would treat area effects separately. An explosion, cone, or similar effect does not become clearer merely because the beacon is declared its new target. Instead, let the teleport happen before the effect resolves. The creature avoids the effect only if its new position is outside the affected area. This also creates useful tactical decisions about beacon placement.

The spell's durability is another design lever. AC 5 and 1 hit point mean that using the beacon to intercept an attack will usually sacrifice both the beacon and the concentration spell immediately. That makes the defensive mode much more expendable than the reusable repositioning mode. If the intention is for the spell to support several coordinated moves during a fight, give the beacon enough resilience that destroying it is a meaningful enemy action rather than an automatic consequence of the first interception.

Once the trigger, teleport timing, area effects, and beacon durability are settled, playtesting will tell you much more about whether the 3rd-level slot is correct than numerical comparison alone.

Track a homebrew mana system with one visible resource pool, one clear maximum, and costs expressed in the same unit whenever possible. The bookkeeping should be simpler than the decisions the resource creates; if players spend more attention converting values or checking recovery exceptions than choosing how to use magic, the subsystem is doing too much administrative work.

Avoid separate pools unless they produce genuinely different play. A main mana total plus several color, school, class, or recharge pools can look expressive on paper while turning every spell into accounting. If distinctions matter, consider tags or permissions before creating another number to track.

Recovery should also happen at predictable moments. State exactly what restores points, whether partial recovery exists, and whether unused mana has any value at the end of an adventuring period. Put the current total somewhere that can be changed instantly: a counter, tokens, a large field on the character sheet, or a simple digital tracker.

Finally, test the system during stressful turns. If a player can see their current total, understand a spell's cost, and update the resource without pausing the table, the tracking layer is probably light enough. Complexity should live in strategy, not arithmetic.

Make every part of the sigil answer a different question, so combinations behave predictably before you start inventing exceptions. One useful structure is: outer sigils define substance or property, the circuit defines how power moves or changes, and the center sigil defines the final shape or delivery method.

Then design outer sigils as verbs and qualities, not only elements. Heat, pressure, adhesion, decay, reflection, growth, silence, memory, weight, division, attraction, preservation, or motion can combine with water, fire, stone, and similar substances. Two water sigils might increase volume; water plus heat might create steam; water plus pressure with a narrow-output center could produce a cutting jet.

Give combinations consistent conflict rules. Opposed sigils might cancel, dominate according to quantity, or convert into a third effect. Pick one logic and use it repeatedly so players can infer results.

Finally, connect power to cost. More sigils, unstable pairings, or tightly focused outputs can increase backlash, preparation, or endurance demands. The system becomes interesting when players can predict what is possible but still have to decide how much complexity and risk they are willing to accept.

Start by deciding whether the spell’s main payoff is damage or control. A 2nd-level spell that deals competitive damage, knocks a target down, frightens it, and can also incapacitate it is trying to buy several valuable effects with one slot.

For a single-target illusion attack at touch range, a sensible first playtest is 3d8 psychic damage on a failed Intelligence save, half on a success, plus one modest rider on failure. Prone is a cleaner starting rider than frightened or incapacitated because it changes positioning and follow-up attacks without potentially removing a creature’s whole turn. Increase the damage by 1d8 for each slot level above 2 if you want straightforward scaling.

The illusion should also explain its fiction. If the target takes psychic damage because its mind accepts the impossible hammer as real, describe the impact through perception rather than making the conjured object physically smash the environment.

I would remove a 24-hour immunity unless repeated casting creates a real problem in testing. A successful save already reduces the effect, and blanket immunity adds tracking.

Then test the spell beside other 2nd-level options at the same table. Watch total damage, how often the rider matters, and whether the spell becomes the automatic choice whenever one target is available.

Do not automatically make a single-target Silence lower level. The official Silence spell is a 2nd-level concentration spell that creates a fixed 20-foot-radius sphere for up to 10 minutes; creatures inside are deafened, thunder damage is suppressed there, and spells with verbal components cannot be cast. A version that follows one creature trades area for reliability against that creature, which can be stronger in the situation that matters most.

For an unwilling target, a reasonable playtest starting point is still 2nd level if the effect requires a saving throw, lasts no more than 1 minute with concentration, and allows another save at the end of each affected turn. Limit the homebrew effect to preventing sound and verbal spell components unless you intentionally want the extra thunder-immunity function too.

If the spell has no repeat save, automatically follows the target, or can be applied from long range without meaningful counterplay, treat those as power increases rather than conveniences. Raise the level, shorten the duration, or add another limitation before testing.

Judge the result by play: can one failed save erase an enemy caster for an entire fight? If so, the single-target version is doing more than its smaller area suggests.

Rules reference: D&D 5e SRD 5.2.1 — Silence.

To make psionics feel genuinely different from spellcasting in D&D 5e, change the procedure and the decisions, not just the names. Decide what psionics spends, how it is resisted, what risks come with using it, and what information opponents can perceive. If every power still behaves like a spell slot with different flavor, the distinction will disappear in play.

One workable model is a small pool of points that can be spent flexibly on a few powers, with stronger effects requiring more points or accepting strain. Another is a push-your-luck system where the character can exceed a safe limit at the risk of conditions, lost control, or narrative consequences. You can also differentiate detection: a psychic effect might lack visible components but leave mental traces that specialized creatures can sense.

Keep compatibility with the rest of 5e. Decide explicitly how Counterspell-like effects, antimagic, concentration, resistance, and rest recovery interact with the new subsystem. Do not create exceptions one scene at a time. Write a short interaction sheet and test it with several classes. A distinct subsystem succeeds when it creates different choices while remaining understandable to players who still use the normal magic rules.

Change one part of Shield at a time and preserve the reason a player chooses it: spending a reaction and a spell slot to survive a dangerous moment. As written, Shield is a reaction after an attack hits, gives +5 AC until the start of the caster's next turn, and also negates Magic Missile. Those three pieces are the baseline you are changing.

If the concern is the size of the AC swing, the cleanest modification is a smaller bonus, such as +3, while keeping the duration and trigger intact. If the concern is that one casting protects against many attacks, another option is to keep +5 but apply it only to the triggering attack. Do not reduce the bonus, shorten the duration, add a roll, and remove the Magic Missile protection all at once; that makes it hard to tell which change solved the problem.

Test the revision in the situations where Shield matters most: several incoming attacks, a lightly armored caster, and a character who already has high AC. Watch whether the spell still creates a meaningful defensive choice rather than an automatic reaction.

Finally, describe the altered version clearly to players before character choices are locked in. A homebrew fix is easier to judge when everyone knows exactly what changed and why.

Make the world reward different spell functions before weakening Fireball or restricting how often it can be cast. A spell-point system makes resource spending more flexible, so players may repeatedly choose an efficient area-damage spell if nearly every problem is a cluster of enemies waiting to be defeated. The encounter design may be reinforcing the behavior.

Introduce situations where damage alone is insufficient. Spread adversaries across elevations, mix foes with civilians or fragile objectives, provide escape routes, or establish a countdown that requires movement and interaction. A web of hazards, a negotiation with consequences, or an objective that requires information can make control, mobility, protection, and utility spells essential rather than decorative.

Review the actual spell-point conversion separately. Check how its costs compare across spell levels, whether high-level casting has limits, and how many encounters occur before resources are restored. A generous pool with no meaningful opportunity costs may encourage repeated use of the most efficient effect even in varied encounters.

Avoid designing every battle specifically to nullify Fireball. Let it excel when tightly grouped enemies create the opportunity, then present different challenges that reward other choices. A successful resource system gives players reasons to diversify voluntarily instead of making one preferred spell mysteriously ineffective whenever it appears.

Yes, but treat it as a redesign of character progression rather than as a simple compensation package. Removing player spellcasting changes combat, healing, travel, information gathering, and problem solving. Giving every character a dragon can replace some of that capability, but only if the dragons are designed as part of the characters rather than as extra full NPC turns.

Define what each dragon contributes at the same milestones where the character would otherwise gain important options. An elemental attack can cover damage, but decide how the campaign will handle healing, flight, scouting, control, defenses, and utility. Do not try to reproduce every spell; choose the capabilities that matter to the campaign’s intended style.

Keep the dragon mechanically attached to its rider where possible. Shared actions, limited commands, or rider features are easier to balance and faster to run than doubling the number of independent creatures in initiative.

Build encounters around the altered party, not around assumptions from an unchanged D&D game. Test with a short adventure before committing to a long campaign. If one dragon ability solves every obstacle or another character loses too much of their class identity, revise that specific piece.

The concept can work well, but the dragons should become the campaign’s character system, not a bonus layered over a system you have partially removed.

Revealing a monster's exact hit points can be useful, but its value depends on how much the party can act on that information. Precise numbers enable efficient target selection and timing of high-damage resources; a descriptive condition already supports most ordinary tactical decisions. Exact health knowledge is an information advantage, not automatically an encounter-breaking power. For a homebrew tracking spell, decide what the feature costs and what uncertainty it removes. A target must first be marked or injured, the caster spends an action or resource, and the information remains accurate only at the moment of use. Those restrictions create meaningful decisions without adding multiple random accuracy rolls. If the effect consumes a spell slot and an action, a medicine check to receive a deliberately inaccurate number may make it feel unreliable rather than exciting. Allow players to communicate what their characters observe. Artificially forbidding them to share a number they know rarely improves play. If precise numerical knowledge undermines the atmosphere, reveal a health band or worsening condition instead, and tell the table that this is the feature's intended resolution.