Eero Tuovinen is crowd-funding a discussion about old school play: Muster: A Friendly Primer to Challengeful Dungeoneering. Why, you might ask, is another primer useful, given the existence of Finch’s Old School Primer, Principia Apocrypha, and various fragmentary discourses scattered across the blogosphere? Well, Eero comes at this project from a slightly different perspective, probably, from most readers of this blog. Most obviously, he prioritizes focused design over general toolboxes, though he has a long history thinking about the older style of tabletop fantasy roleplaying games (exhibit 1). While focused design is relevant to how he thinks about game design broadly, for this particular project, I would also draw your attention to the idea of wargaming around which he positions the project.
“Wargamey” is only one of five principles, the other four being challenge-focused, egalitarian, creative, and authentic. I, personally, am particularly interested to see what implications he works out from the last of these principles, given my somewhat ambivalent take on genre in tabletop roleplaying games:
authentic rather than ritualistic; the grand purpose of a D&D campaign is to game out the implications of the campaign set-up rather than replicate the GM’s expectations of what D&D adventuring is “supposed to be”. We discover the truth via play.
(As a side note, I briefly played in a few early sessions of the online Greyhawk campaign he mentions, though scheduling and other obligations got in the way of extended participation.)
One final point that I think makes this project worth supporting: part of the goal is open community access:
An essential vision for this project is that Muster will be a free, probably Creative Commons licensed booklet that anybody can pick up off the ‘net and do whatever with.
There is about a week left in the campaign, and he is only a few hundred dollars away from the funding goal at the time of posting, so this is a project where a few more backers might actually make a difference to viability. Check it out.
I follow a few ongoing manga series on Comixology—Delicious in Dungeon, Made in Abyss, Goblin Slayer. Ubel Blatt, though that is finished and I am way behind. Berserk, though that comes out so infrequently it is hard to think about as following. As you can probably see, I have a type: occidentalist fantasy, often influenced (perhaps self-consciously) by tabletop roleplaying tropes. Last time I signed in, I noticed this new series in my recommendations—The Undead Unwanted Adventurer—and it seemed to hit a number of those notes so I decided to give it a try. As it turns out, it is something like lighthearted Dark Souls pseudo fan fic, which has the potential to be, ahem, exactly my sort of thing. In terms of common manga genres: shounen isekai, leaning seinen, with a bit of harem going on, though though the isekai aspect is a rebirth rather than a transportation. (This post mostly avoids plot spoilers, apart from the initial event alluded to in the manga title, but it does contain some minor setting spoilers. Also, for the panel excerpts, keep in mind you should read from right to left.)
First, let’s get the judgments out of the way. This is a smaller-scale, more niche offering than some of the other series listed above. Nakasone’s art is pleasant enough, but lacks the precision and emotion of Miura’s work on Berserk, the richness and complexity of Nihei’s work on Blame!, or the saturated lightness of Tsukushi’s work on Made in Abyss. Comparisons with such accomplished artists may be unfair, but I want to set expectations. Further, the best art in The Undead Unwanted Adventurer is probably in the fan service panels—more on that below. The writing, as well, is good enough to get the story across, but otherwise unremarkable. The story began as a light novel in Japan, before spawning a manga manifestation, and the light novel medium is comparable in some ways with “young adult” fiction in the Western market. So, what we are left with is a story that will rise and fall based on the world building and manipulation of genre touchstones, and how it rewards or surprises reader expectations. Overall, the series kept me curious and entertained enough to plow through the three available volumes in a matter of days.
A common aspect of this subgenre is in-world justification or representation of RPG game elements. This can be crude—such as the magical HUD that characters consult in Rise of the Shield Hero—or clever—such as the “adventure sheet” adventurer guild registration form in Goblin Slayer. When done well, such aspects can suggest ways to connect game elements to setting elements. There is a kind of appreciation seeing in-setting explanations for something like a the Monster Manual hierarchy of undead, or similar game artifacts. Undead Adventurer has a lot of this, including different kinds of “mana” (roughly corresponding to character classes), adventurer guild ranking with exams, mega-dungeon exploration serving almost as borderlands industry, and so forth.
About the harem thing. Our protagonist is a largely misunderstood and unappreciated loner who, throughout the course of adventures, accumulates a number of mostly female sidekicks. Our hero has the opportunity to save or otherwise assist these supporting characters, who come to appreciate his quirks and become fiercely loyal. Sound familiar? (I am somewhat surprised that I was unable to find an overly specific subgenre term for this dynamic given how common it is, but here we are.) This sort of escapism can range from playful wish-fulfillment to bitter resentment, and Unwanted Undead Adventurer lands firmly in the range of playful for me, though it occasionally deserves some eye-roll.
One of the most enjoyable qualities of occidentalist manga in this vein, for me at least, is seeing the common tropes generally taken seriously rather than constantly subverted, but also filtered through the lens of another culture, or idiosyncrasies of a particular creator, which lends novelty and the occasional (but regular) instance of surprising, substantial deviations from common expectations. For example, orcs in the world of the Undead Adventurer are brutish, pig-faced humanoids—that all seem to be terrifying giants, hunted because the flesh of orcs is particularly savory. Recognizable, but warped, and somewhat ghoulish. Perhaps this is a way to approximate the New Sincerity turn, but in genre fantasy, for those of us that grow bored of constant genre irony and deconstruction?
(Panels excerpted here are an assortment from volumes 1 through 3.)
Every once in a while, but with some regularity, someone will ask for module recommendations, often as a way to get into old school or classic play. I am somewhat conflicted regarding the best response to such requests because, while there is nothing fundamentally wrong with modules, and they have some particular utility, modules are also limiting, often have poor handling, and are poor examples of the form’s potential.
My inclination is to instead suggest that referees, even new referees, create a simple custom scenario. This is unnecessarily intimidating, partly due to some unfortunate ingrained assumptions, including seemingly upward comparison to professional offerings with illustrations, cartography, edited prose, and so forth. In reality, crafting a satisfying classic scenario is often less complicated than the process of building a player character in a contemporary system, but has become obscured by a shroud of world building complexity and the detritus of expectations related to literary story structure. There are a few examples of procedures that a referee can follow to create a scenario, such as Moldvay’s double spread in 1981 Basic D&D (pp. B51-B54) which explains how to create an adventure. The recent Old School Essentials SRD provides similar information, available for free online (adventure scenarios, designing a dungeon, designing a wilderness). Other resources exist as well, though often buried in other materials.
These resource are useful, but still often seem to be pitched at the wrong level, lacking sufficiently concrete set of recommended actions, or attached to bland thematic content. As an example of the kind of referee rules that I think would be an antidote to module recommendations, I want to highlight two recent blog posts about crafting megadungeons:
The Two Week post in particular is a practical and clear example, though perhaps still slightly intimidating for a starting referee. Unfortunately, the trappings of a megadungeon is probably mildly counterproductive, despite the likely truth that following Nick’s megadungeon building guidelines would probably be easier and more enjoyable than skimming B2, or just about any published module scenario. (In reality, the observant and experienced reader will also note that a megadungeon is really just a series of linked scenarios that happen to be subterranean.)
If you are aware of any other similar scenario building procedures worth highlighting, drop them in the comments. Procedural rules rather than something like a collection of tables would be most useful. I am sure I am neglecting some other good resources. This is also a call for what to include in a game’s referee book or section.
Dark Souls encodes a number of fundamental play dynamics that produce a particular kind of satisfying play. In this post, I am going to discuss a way the game supports and rewards an enterprising but focused player stance toward the unknown.
First, the game penalizes both reckless and overly cautious play. A player of even a relatively high level character can still be messed up substantially by some of the weakest and simplest enemies if the player fails to pay attention and take the threat seriously. In contrast, a player that is too cautious will always be on the back foot, and less able to take advantage of opponent weaknesses. In this way, courage in the face of danger and a spirit of enterprise serve the player best for developing their own skill, improving their fictional avatar, and deepening the complexity of the imagined world. In short, the game entices player engagement and rewards persistence in the face of failure (amplified by many other design decision that are beyond the scope of this post), rather than providing participation trophies for just showing up or presenting a passive media entertainment experience. (Post continues after 80 second video fighting an Old Knight in Dark Souls 2.)
Second, this reward and penalty structure creates a cybernetic feedback loop. Feedback loops are powerful, but can also topple into degenerate cases. The most engaging forms of feedback loop for gameplay involve adaptation rather than return to static equilibria. The system improves itself. If a feedback loop reaches a static equilibrium, the “game” ends, even if the players continue to engage due to habituation (perhaps analogous to how a thermostat maintains static conditions). If the feedback loop is negative, at some point the game ends because the activity as pattern destroys itself (similar to how a democracy can collapse into tyranny, a market bubble can pop, or a virus can burn itself out by killing all the hosts). In an adaptive feedback loop, each iteration of feedback produces a more complex, satisfying, resilient whole.
The play iterations of a tabletop roleplaying game can manifest similar dynamics. In one example of a degenerate case, players learn to avoid play by creating elaborate scripts to mechanically deploy when presented with any challenge. In OSR/etc. games, this sort of script may manifest as something like: always look at the ceiling, tap the floor with a ten foot pole, listen at the door. Disconnected from any context or cost, such scripts represent rote mechanization rather than adaptive learning. Players indicate by such behavior a way that their own motivational architecture is to some degree incompatible with or in tension with the game mechanisms or the particular campaign instance.
In the adaptation case, players learn how to navigate challenges by developing skills, along with the contextual knowledge of when to use the skills, and in so doing increase competence. This allows a player to face more complex challenges that incorporate, but move beyond, the existing skill. A simple example trajectory of several iterations might be fighting one troll successfully (to learn troll weaknesses), then fighting many trolls, then tricking other opponents without access to troll weaknesses into fighting trolls, and so forth. Ideally, the player would learn such details through play rather than outside of play by memorizing facts from an official rulebook.
Dark Souls, being a computer game, might be a purer example of this dynamic compared to a tabletop roleplaying game, since options are more constrained, but the template applies to some kinds of successful tabletop roleplaying games. The various versions of Dark Souls, and specific challenges within each, embody this ideal design to greater or lesser degrees; sometimes the game misses the mark. Though Dark Souls rewards certain kinds of creativity when approaching challenges, it has a basic combative frame which limits the learning potential. Additionally, sometimes the challenges degenerate into tedium—such as some long approaches from bonfires to bosses which require the player to pay a relatively boring tax to attempt a previously failed challenge again. But the abstraction is still clear enough to be a useful exemplar. (Post ends with 112 second video fighting The Lost Sinner in Dark Souls 2.)
(Videos are personal play recordings, all Dark Souls 2—I apologize for inflicting my poor technique on the viewer.)
Because people talking about classic tabletop roleplaying games have become balkanized over various social media platforms, and Twitter is a hellscape, and I may want to link or refer to this collection of threads in the future, I am spotlighting them here. I also used Thread Reader App to create thread pages so that the threads can be read in a more bloglike format, for those that so desire:
It is probably worth noting, since this is the Internet, that my spotlight means “useful and worth reading” rather than that I agree with all threads on all points and in all particulars. For example, I would say that alignment as adventurer allegiance (rather than adventurer moral commitment) can function quite apart from moral essentialism, to say nothing of less savory interpretations. Law and Chaos as depicted in OD&D (and the refined “Basic” rulesets) is fuzzy in this regard, easily pushed in either direction. To be more explicit in notion, consider using alignment categories such as Unseelie, Neutral, Seelie, or Rebel Alliance, Neutral, Galactic Empire rather than Chaos, Neutral, Law (or, God help us all, the baroque murk that is the AD&D 3×3 matrix).
In case you want to create similar thread pages using Thread Reader App, here is how to do so A) without cluttering up someone else’s thread and B) without cluttering up your own timeline.
Quote-retweet the thread in question with only the text: @threadreaderapp unroll
Wait for the @ reply from @threadreaderapp (which in my experience takes a few seconds)
Save the URL provided in this reply (which you could also derive from the original Twitter thread URL based on the tweet ID)
Delete your quote-tweet that was originally addressed to @threadreaderapp
Now the unrolled thread page persists for future reference
(You may need to follow @threadreaderapp first as well, but I am unsure.)
You can also easily print the unrolled thread pages to PDF, perhaps with the help of extensions to exclude distracting images, and so forth.
Rules Cyclopedia and BECMI D&D provide rules for advancement up to level 36, with guidelines that adventurers should gain one level every five adventures (Rules Cyclopedia, page 129). I generally take “an adventure” to entail multiple game sessions, but even if you complete one adventure every session, assuming no setbacks such as adventurer death or level drain, and play once per week, a group must play for almost 3.5 years to reach level 361. It seems reasonable to assume that few actual campaigns have followed this procedurally proposed trajectory. This is an extreme, but other TSR rulesets, such as B/X or AD&D, seem to also imply an implausible level of commitment for most people playing most campaigns in most situations. One interpretation of this fact is that the rules are flawed, and it would be better to make endgame rules that people will, in fact, actually play. But what if the main function of late game rules material in TSR D&D, and descendants, is more to entice players into longer games?
A design trend exists around more focused games, which values explicit and transparent description of both rules and intent. According to this sort of evaluative criteria, the late game TSR D&D rules are ineffective, and perhaps even misleading or disingenuous. They promised me a castle, and all I got was this lousy longsword +2! However, there is no inherent reason that the most effective rule, functionally, must articulate its intent and function, or even that the designer must understand the likely function. Surely there are exceptions, but in general focused games tend toward one-shot, mini-series, or shorter campaigns. It seems possible that the lack of enticing phantasm in such designs may partly explain this difference.
Goals are only motivating prior to being reached. Following goal accomplishment, it is on to the next goal. Perhaps the TSR endgame is a form of ultimate imagined goal that works exactly because it is unlikely to be obtained. Lack of complete consummation does come with some drawbacks, such as the anecdotal observation that campaigns often end, following Eliot, not with a bang, but a whimper. As the duration of a campaign increases, an anticlimactic ending becomes more likely, even if players do maintain interest and attention, as changes in life circumstances will implacably conspire against the campaign of unusual ambition’s persistence. Whether a longer, complex campaign, often lacking closure, or a shorter, more contained campaign is better seems like a matter of taste, but either way having rules that players almost never use could still shape gameplay in substantial ways. This is another example showing how taking a text at face value may be a poor, or at least incomplete, guide to game functionality or game meaning.
1. A fighter requires approximately 3.4 million XP to reach level 36. A magic-user requires approximately 4.3 million XP to reach level 36. ↩
People often use different words for the same (or highly similar) concepts, or the same word for radically different concepts. Academia in particular has driven home for me the mismatch of jargon and meaning across silos, but the same thing happens in nontechnical discussions. The confusion around creative agenda modes shows an example within discussion about tabletop RPGs. In general, I suspect actual communication happens far less often than people assume, in any domain. In this post I will highlight a few more terms that scenes seem to use in different ways, specifically simulation and exploration.
Forge-speak simulationism seems close to what OSR/etc gamers (myself included, I suppose) mean when we speak of exploration. When a big model theorist discusses simulationism it has almost nothing to do with whether, for example, sailing rules in a game accurately capture wind dynamics as might be modeled by a physicist. Recall that the subtitle of the original Edwards essay on simulationism is The Right to Dream. And Forge-speak uses exploration to denote a completely different concept, as should be superficially obvious when you look at the diagram/logo for the big model wiki, something more like engaged play. Here is a big model explanation of Exploration:
What’s Exploration? Exploration is playing the game. If you are imagining stuff, rolling dice, talking in character, or paying attention to someone else’s scene in the game- you’re playing and you’re Exploring. If you’re busy playing videogames on your DS and not paying attention, if you’re on your cell phone, reading comics, or otherwise in any common sense way not playing the game, you’re not Exploring.
The lack of knowledge is thus a source of peril and uncertainty. It is both an obstacle to be overcome and a hazard to be dared. There is also a thrill of discovery, of uncovering things that are hidden. If the game is being played well, what is uncovered are not boring things. Every dungeon is initially a mystery, every artifact a hidden wonder, every faction an unknown quantity. To explore a dungeon is to unravel the mystery of the place.
A pair of Dark Souls analogies may help further illustrate this distinction. OSR exploration has to do with the feeling you get when you hear the new area sound, which has been burned into my mind as the sound of a new vista opening up:
This feeling is tied heavily to interaction with the fictional context as a coherent and meaningful place rather than as game mechanisms or story progressions. (Tangentially, I would also add that this is more about imagined possibility than it is about immersion in the method acting or narrative transportation sense.) In contrast, big model exploration would be closer to what is going on when the player learns and exploits the pattern of an enemy’s move set:
This is more like a structural fitting in of the player’s voluntary actions with the mechanisms and interfaces that the game presents. You could get the first kind of exploration without engaging many of the Dark Souls game mechanisms at all. And you could get the second kind of exploration in a wireframe test arena totally disconnected from the stages or levels of the game (which would be the campaign setting or shared imaginary space in tabletop roleplaying).
Recently I was thinking to myself: people have written some words about tabletop RPG theory (understatement of the decade). This work is often “empirical” in the loose sense that theorists base conclusions on observing their own play experiences and the experiences of others (often second-hand) but it is not empirical in the sense that a social scientist would generally use that term. And so I was thinking, I have some quantitative data about player preferences, would it be possible to see whether this data could substantiate some of these claims using quantitative methods, or perhaps suggest alternatives based on something beyond anecdote? Specifically, I was thinking about that venerable Forge concept “creative agenda” and the various creative agenda modes proposed as incompatible: gamism, narrativism, and simulationism. These four concepts belong to a broader collection of ideas called the big model, though the GNS subcomponent has had the largest impact on general thinking about RPGs, of the big model concepts.
Before exploring creative agendas in data about player preferences, I had to make sure that I understood the concepts, beyond surface associations. This means taking the theorists seriously, and accepting definitions, even (perhaps especially) when the naming seems counterintuitive or disconnected from the lay understanding of terms. It is unproductive to approach the theory of electromagnetism with the stubborn idea that a field is properly only an expanse of ground covered by grass, flowers, and so forth. Ideally, jargon will involve some congruence between intuitive and technical meanings, but that is perhaps unrealistic to expect in all cases. Sometimes it is best to just use the imperfect, established term, and move on with your life.
After several conversations with people who have engaged heavily with these concepts and find parts of the big model personally useful, I have come to believe that I (and probably many people outside of the Forge and story games diasporas) misunderstand both the specific terms and the broader goal of this collection of ideas. Rather than explanation, I think the purpose of big model thinking generally, and the more specific ideas around creative agenda, is to present particular ideals of possible game practice. This is in many ways closer to a therapeutic framework in counseling psychology, such as psychoanalysis, person-centered therapy, CBT, or DBT. Therapeutic frameworks generally involve a greater focus on particular interventions and some alignment with evidence based practice compared to the big model, but once you understand the theory like this, the way people use and discuss it makes a lot more sense, at least to me.
To be (somewhat) more concrete: the big model theorist looks at a collection of play goals aligned in some way, such as learning and challenge (“gamism”) and wonders whether purifying a play experience around only these play goals might be uniquely satisfying for the collective experience of a group of players. Similarly for other modes. This is what the big model means by coherent play. Whether this matches any particular experience is almost beside the point, much like the way a three-minute sprint mile is an ideal that has never been reached for the sprinter (though obviously the activity of sprinting is far simpler than tabletop roleplaying).
So what about the player preferences data? Based on this understanding, it would be a poor fit for big model concepts, even accepting that measurement at the individual level could approximate a group-level analysis consistent with big model claims. People will have to think a lot harder about ways to measure big model concepts before that will become possible, and from what I can tell there has been little enthusiasm for that kind of work to date. The data I have still has potential to reveal connections between what players enjoy during play and what players find useful in games, which I may explore in a future post.
Some simple takeaways for outsiders when trying to make sense of big model ideas: substitute “play goal” (or perhaps “ultimate play goal”) whenever you see the words creative agenda, and realize that the point of the theory is presenting possible ideals, rather than explanation. You may lose some nuance, particularly involving individual versus collective experience, but that is probably closer to the intended meaning than running with intuitive associations.
A brief taste of the feast to come. All the spells are written, the layout is done, and we are just messing with some final details. Update: it is done! Get it here for free. Four samples:
[Earthquake] Mood of Gravity
By smearing the face with mud, dust, or jagged rock shards, the sorcerer’s soul fuses with a greater earth spirit, binding the emotions of the sorcerer to the movement of faults deep below. For the duration of the spell, annoyance and minor pains result in tremors while wrath and great pain yield proportional restlessness of the earth. When the spell ends, the sorcerer becomes numb to the experience of any emotional intensity until the spell can be prepared again.
The sorcerer must prepare a dish from flesh of a four-legged animal, spiced well and thoughtfully. Upon speaking the magic words, and consuming the food, the sorcerer grows to the size of the animal consumed, persisting in that size until passing the animal’s remains. Only the sorcerer’s body changes, and upon the spell’s end there is a 1 in 6 chance of the sorcerer permanently manifesting some physical aspect of the creature, such as goat horns, horse hooves, or cat eyes. If the animal consumed was of a kind to lay eggs, the sorcerer will lay an egg within 1d6 weeks, birthing a creature of wonder or terror.
[Fear] Beacon of Terror
By prostration, tracing forbidden signs on the ground, and slapping the ground three times, the sorcerer causes an infinitely high beam of light to spring forth from the ground like a pillar rising to heaven. For any creatures of level four or less witnessing the radiant beam, taking any action other than fleeing the beam pillar requires succeeding at a saving throw. When the spell ends, the beam flickers out, but if the beam was called under open sky there is a 1 in 6 chance of attracting attention from spirits of the air, dragons in transit, or a flock of giant eagles.
[Message] Sending of Bats
During the dead of night while in a cave, earthy hollow, or dead tree, the sorcerer invokes the ancient pact of Chirops, which obligates bats to serve as envoys for the wise. In 1d6 turns, a colony of bats prostrates themselves around the sorcerer to receive a message, which the colony will then carry to a destination or recipient provided by the sorcerer, to sing the message thrice in a harmony of bat voices. There is a 1 in 6 chance the bats will be the size of condors—bat champions—only willing to serve following negotiation of terms.
Here is a JRPG rules hack. I think this one is tighter than my previous attempt, and may even be playable as is.
First chose a base chassis (B/X D&D, Old School Essentials, Labyrinth Lord, whatever), and then apply the following rules modules.
Every player character gets a signature weapon. Fighters get sword (because fighters are the magic sword class). For other classes, choose a non-sword signature weapon (or determine randomly): 1 axe, 2 bow, 3 crossbow, 4 dagger, 5 mace, 6 spear, or 7 war hammer. Adventurers can use weapons afforded by class or signature weapon, and attack with advantage when using a signature weapon. (This list of possible signature weapons matches possible magic weapons from the classic treasure tables; if you choose some other kind of weapon, such as revolver, you might want to modify the treasure tables accordingly.)
Rather than traditional spells, adventurers draw power from magic crystals called homunculiths. Replace magic weapon plusses with homunculith sockets (so a spear +2 means a spear with two homunculith slots). Any character can use magic afforded by a slotted homunculith if the character can use the weapon and can supply the necessary magic points.
Magic-users can slot a number of homunculiths in various magical paraphernalia equal to character level. This can be hat clips, belt buckles, cane handles, whatever (describe the slots; the stone has to go somewhere, and has to be visible). Working homunculith slots into equipment requires a haven turn or downtime action for a magic-user. Only magic-users can make use of homunculiths in magical paraphernalia. Characters that would otherwise begin with spell slots start with one randomly determined homunculith.
Determine treasure using the treasure tables with some degree of strictness, but: replace magic scroll results with spell homunculiths, replace magic ring results with nexus homunculiths (used to summon daemonotheurgic entities; see below), and read arrow or bolts as bow or crossbow with homunculith slots (by bonus), respectively.
Since hit points come from hit dice, magic points must come from magic dice; adventurer MD by class: fighter = d4, thief = d6, magic-user = d8. (Generally, classes with high HD should have low MD and vice versa, so infer MD for other classes based on that principle.) Determine MP total similarly to HP total (so a third-level fighter gets 3d4 MP). Additionally, use the MD when determining damage from magic that calls for dice (so magic-users roll with pools of d8 and fighters roll with pools of d4). Characters recover spent MP during haven turns/downtime.
Choose a spell list. Determine the spell associated with each homunculith randomly. Ignore results with “summon monster” type effects (because nexus homunculiths handle summoning). You could use the traditional spells, the spells from Pits & Perils, Wonder & Wickedness, the spells I drafted as part of my previous JRPG Basic musings (black magic spells, white magic spells), some other source, or some combination. Here is the list of spell names from Pits & Perils: Bolt, Call, Calm, Cure, Fade, Fear, Find, Foil, Gaze, Glow, Heal, Hide, Know, Link, Load, Mend, Mute, Null, Pass, Rise, Ruin, Send, Stun, Ward. Determine the MP cost of each spell randomly by rolling 1d6. Once determined, the cost is set (so it is possible to discover a better homunculith with the same spell).
Nexus homunculiths are bound to summonable daemonotheurgic entities. Generate the entity linked to a nexus homunculith by rolling on a table of monsters, and then adding an elemental aspect: 1 fire, 2 ice, 3 lightning, 4 radiance, 5 shadow, 6 slime. For the table of summonable monsters, collect all the monsters in your rulebook of choice with HD of 6 or higher, crossing them off as adventurers discover homunculiths. Give each daemonotheurgic entity a name. Attach an action die to the daemonotheurgic entity. By default, this is d6:
Special (make this up when creating the entity)
Summoning an entity costs 1 MP. When summoning an entity using a nexus homunculith, determine entity HP using remaining entity HD.
Roll the entity’s action die to determine actions each round after summoning. The summoner can override the action die using a command, but this requires spending an action. Commanding the special attack will cause the entity to depart afterwards. The action die determines the monster’s action but the summoner’s player determines all other details, such as targets and so forth.
When determining summoned monster HP, roll the monster’s remaining HD and leave the dice on the table as they fall (or record the numbers per die). When the monster takes damage, the player may decide to which die the damage applies. If a die total is reduced to zero or less, remove the die and ignore any excess damage rather than process the spillover damage. Restore removed dice during downtime recovery. Healing a summoned monster allows rerolling some number of remaining HD rather than adding HP directly or restoring removed monster HD.
At the end of each combat round, spend 1 MP or the entity departs.