漫画 Shelf Elf

 Manga Index

"Shall I stack this up neatly for you?"

I'm going to try to draw up a few of these as illustrations for an essay I still need to write about my very successful Box Binning (and Bag Bunching) Project.


模倣子 Terrain Generation in mDnD

Initial mDnD Essay HERE 


We need a simple way to divide up space, so that an entire globe may be easily generated, preferably mostly as the party of adventurers makes their way through it.

I propose a set of nested increasingly large triangles that ultimately reach the size of an entire planet, in the form of an icosahedron (same shape as a d20).

A macromemetic-like system is needed to generate the next blob of terrain as the party enters it. Crossing the boundary from one triangle to the next, we could ask whether one needs to know where the given triangle (whatever its size) is located in the larger one. Perhaps this is not the case.

In a given terrain, such as town, kind of town, desert, forest, hilliness, mountains, the weather, winter, where is the seashore and so on, the next tiny step in any direction is going to tend to be more of the same.

So our requirements for this system are that:

  1. it be scalable from person size to whole world size
  2. it be easy to dynamically generate the next grid as the party approaches it.

The Icosahedron (d20) as Globe 

Quick overview: the Earth has a surface area of approximately 200 million square miles. Thus each face of an icosahedron (d20 solid shape) would contain 10 million square miles.

fig 1. icosahedron (20 faces) as approximate globe

This means that the edge of each face would be some 5,000 miles.

We can endlessly tile a gigantic equilateral triangle thusly:

fig 2. triangular face tiled into four, then each of four into four, etc.

In other words, the first "cutting" or "tiling" gives us four equilateral triangles inside the original one, and each edge is half the edge length of the original. This dividing by two property is intriguing. It makes things simple and clean if we just use multiples of two. In other words, if we just say that the largest edge is 4,096 miles, then we can go all the way down to one foot, which we call the "zeroth tiling level" (1).  

To extend from the scale of the very small to the very large, i.e., the level at which characters might experience a town or dungeon up to planet size, we have tiling level and edge length:

tiling levelfeetmiles

fig 3. triangle edge sizes from one foot to planet size

How to Generate the Next Bit of Terrain? 

We want to be able to generate new terrain as the party moves into it. The overall concept I want to follow is that the chance of a totally new terrain type starting is always at some level of possibility, and is also based on surrounding already decided terrain.
fig 4.1. triangles that abut upon the party's triangle

Every triangle is surrounded on three sides by three other triangles, as well as three triangles which only touch on the points. There must be continuity between triangles that abut one another. Or must there?

Another thing to think of is the next higher level, or even higher than that. The triangle we are currently in is within a larger triangle, and that triangle is within a still larger one. All of the surrounding triangles influence the triangle between them. Each one of those triangles are either the same kind of terrain, or a transition to some other kind of terrain.
fig 4.2. What terrain transitions look like

A terrain tile triangle is either solid, i.e., a continuation of the current terrain, be it grasslands, forest, town, dungeon, or whatever, or it's two-thirds one terrain and a third some new one, with the line of division going between the midpoints of two of its sides. When a party enters a new grid space, they will still have the same terrain, but they may or may not face a boundary ahead of them, as in figure 4.2.

Now comes the next big question: how to determine which kind of terrain happens next? The odds are that things will stay the same at any given moment. However, as in figure 4.2., if a neighboring tile changes terrain, then at least one, possibly two, and perhaps even three or four will change to the same new terrain. This is obvious since a change in terrain cannot take place on a single grid tile -- it must be continuous with the tiles around it.

The reality is that this needn't be decided for more than one tile if one is moving across an edge, and still even one is sufficient if one is moving through a point.

The problem is what kind of dice roll is needed to determine a change in terrain? Is this totally arbitrary, or is there some way to make it reasonable?
fig. 4.3. terrain of smaller area determined by surrounding area

Terrain Intersections and Boundaries

Two sections of terrain can border one another like in fig. 5.1a.

fig. 5.1a. two bordering terrain regions

The only kind of division of tiles that is "needed" is a triangle the tip of which may be cut off by one or the other of the bordering regions, see fig. 5.1b.

fig. 5.1b. A triangular tile shared between two regions

Now, if we have three regions intersecting we have two choices: 1. disallow certain terrains to border one another, e.g., no mountains next to sea shore, no forests next to deserts, and so on, or 2. have tiles that connect three different regions. This basically translates into things like needing a solid band of grassland that you have to cross to get from the ocean to the mountains and such. Figure 5.3. shows how two or more regions of terrain can meet. Note the red exes depict where more than two regions touch, and where we would need special tiles beyond the "tip bitten off" tiles from figure 5.2.

fig. 5.2a. Four sections of terrain and how they meet

fig. 5.2b. Triangular Tile Dividing between three regions

You might object that there has to be a tile that has a different region on each corner and a fourth type of tile in the center.

Maybe this is all just too complicated.

K.I.S.S (Keep it Simple and Spiritual, Dummy) 

It may be enough to say that each next segment of terrain, of whatever size, will tend to be of the same type as the current one, unless there is some kind of a change, and maybe this can be determined by a natural twenty roll on a d20. Or possibly rolling two d20s and seeing how far apart they are. If they exactly match, say, it might mean that the terrain changes drastically, and that can be up to the DM.

One way this might make sense is to think of the tile the party is in, and all of the tiles in the "super triangle." Think of this as the triangle that contains the party's triangle and the triangles around it, or two levels higher out. This is the first triangle that points in the same direction. Including the party's triangle, this is a total of 16 triangles. If you roll a 20, say, or a 17, 18, 19, or 20, this can be interpreted as "our triangle is special in this supertriangle." In other words, that the next triangle should be different.

Just a thought. More work needed on this, probably.

The scale of the terrain can be a foot, which is probably too small, or can be sixteen feet, or 128 feet, or half a mile or even several miles. An entire triangular tile could be all "grassland" or "forest" but as the party moves into it, the smaller tiles are revealed to be of different, or more specific, composition.

Summary & Conclusions 

I think at this point the next step may be to run some actual virtual campaigns and generate the terrain as the party goes forward. I think I want to write some campaigns as scripts, and do a play by play of the characters' actions and the terrain, and so forth.


A good approach might be to recognize that every triangular space is at the center of three intersecting hexagons each made up of five other triangles. Terrain changes could be based on these three hexagons, eg, one direction leads to forest 🌳 and the two others to grassland. More on this later. 



(1) The tiling level is the power of two.


模倣子 The Macromemetic Appeal of Authoritarianism

 Macromemetic Index 

 Memetics Index 


I've been doing some work with the kids on a special ed bus, trying to get them to behave well, turn up on time for the bus, and so on. It's worked brilliantly, exceeding even my expectations.

I want to apply my findings to the other busses, which have larger, more "heterogeneous" populations, and no aide to help the driver, but there is still a lot of feeling that an authoritarian approach is the best, or the only way. Telling the kids what to do, giving them no choice, and administering certain and severe punishments for noncompliance. Of course, the driver is forced to survey the situation on the bus with his rearview mirror and has to divide his time between driving and keeping the kids under control. The kids can get away with bad behavior while the driver isn't looking, and so on.

This sort of works, but not well. The drivers are constantly complaining of discipline problems on their busses. Things get paticularly bad when a substitute driver takes over. This is all obvious when examined through the lens of macromemetics.

Authoritarianism Feels Good 

Why be an authoritarian when a "democratic" macromemetic approach works so much better? It's more efficient, people are more thoroughly engaged, and they police themselves. In capitalism, workers look for jobs themselves, they train themselves (for the most part), and are often motivated to show up to work and do their jobs without constant supervision.

This is where things get a bit ugly. 

Comparing free-market capitalism to voluntary slavery doesn't have great optics. It also doesn't sound good to say that authority figures simply like to give orders, demand obedience, and administer punishments. Why? It's elementary macromemetics. An authority figure is equipped with memes which are guaranteed resonance, and which are unassailable (free from immunomemetic, or "bullying opportunites"). These are the things that all people, all memetic agents, want the most. To have lots of memes at their disposal, lots of deployment opportunities (1),  and have these deployments be non-assailable (3).

Since there are social memes, many of them, along the lines of "we have to do something" and the demonization of the people who "need to be controlled," there are rewards in authoritarianism beyond just enjoying direct power over those who are subject to the authoritarian power. Authoritarianism is a huge megamemeplex, like many others, and so can have many complex layers and interconnections.

One question for later is what about non-authoritarian societies? Are native cultures less authoritarian than "Western" ones? If so, are non-authoritarian societies more vulnerable to authoritarian ones?

Summary and Conclusions 

Authoritarianism does not have to be ineffective. However, it feels good, and there are memetic rewards to those adopting it and participating in it independent of whether it works. It probably doesn't work very well in general. It's perceived as being some kind of "last resort" or "the one guaranteed way" but in fact none of this may be true. This kinds of ideas, i.e., that authoritarianism works, is the best or only way, may just be immunomemes embedded in authoritarian memamemeplexes, and nothing more.



(1) A deployment opportunity happens when another agent deploys (or enacts) a meme which calls for a deployment, for example, a child misbehaving on a bus. As I've stated, there is a ready supply of misbehavior, or deployment opportunities, for the driver, to deploy things like firing orders to sit down and shut up, or threats of being kicked off the bus. Even though this seems dysfunctional, and it is, it is a very rich environment for well-marked, well-closed, memetic exchanges between well-defined agents. We start to see how authoritarianism leads directly and inexorably to something approximating a memetic chain reaction (2).

(2) A memetic chain reaction, also known as a memetic orgy is where all deployments of any agent creates more deployment opportunities for others and so on such that rate of deployment is sustained at a high level and involves most or all agents involved.

(3) Non-assailability means that when an agent deploys a meme, there are no opportunities for other agents to deploy immunomemes (4). In other words, non-assailability means that a memetic deployment does not generate immunomemetic deployment opportunities for other agents.

(4) An immunomeme suppresses the effect of a deployed meme. This takes the form of not creating deployment opportunities based on the original meme. This has to do with the "state" of the system not changing, but rather being dragged back to the state it was at before the deployment of the immunomemetically attacked meme.


模倣子 Macromemetic Dungeons & Dragons


I just got done taking part in a D&D campaign. My character, Maedac Twigfollow, died. We also did a one-shot heist campaign with another character.

I thought about how rather than rolling up characters at the start, it would be more interesting to attempt challenge after challenge, and each success somehow adding to whichever "abilities" are operative. For example, shooting an arrow at a given monster. Since we like d20s, each chance of success has a minimum of 5%. This could be accomplished by the DM rolling a hidden d20, and then the player rolling their own d20, and rather than some kind of "minimum roll" or such to "hit," it could be if the two die rolls happen to match. More on this anon. This could be called a "challenge roll."

The point is that with each success, the DM could track it, and a series of successes would translate into a higher and higher bonus value that could be later attached to that character's attempts at future challenges that relate to the given "ability."

The other thing is the macromemetic concepts of states, state transitions, and immunomemes. Other players, and the environment, could have immunomememes at their disposal. What is the function of immunomemes in this context? States of the system are kind of obvious: where the characters are in the environment, which treasures and monsters are present, and what all of them may do--which memes they may deploy. The progress of the players through the adventure is in a way predetermined, i.e., they start in the beginning of the adventure, or some point in it, and progress towards some final conclusion. In the simplest case, we have a Point A and a Point B, and a stream of transition-rich memetic pathways between them. A more complex, more realistic picture is perhaps one start, a Point A, branching out into multiple other points, all connected by a web of possible memetic deployments. Imagine "tunnels" or branches of richly connected deployments. The density and concentration of these "trunk lines" is enforced by layers of increasingly intense immunomemes that prevent the adventurers, or rather increasingly impeding them, from persevering on to regions of memespace that do not lead to well-defined areas of the adventure.

DMs do this sort of thing all the time anyway. There may be a way to make it more deterministic, using macromemetic modeling, and this might make a fun example of how macromemetic works in the context of a dice-driven adventure.

Rolling Dice to Create the Future 

Any task or challenge has at least a 5% chance of success, whether it's hitting an armor class, wounding or killing a monster, picking a lock, lifting a heavy object, jumping a gap, moving silently, deflecting an attack, casting a spell, etc.

One way to do this is for the DM to roll a d20 die, and have the player roll one as well. Instead of a natural 20 being a guaranteed hit or needing something above a certain value, say, sixteen or greater, it would be whether the player rolled the same value as the DM's hidden roll, a Natural Perfect Match. For a 10% chance, or 15%, say, and the DM rolled a 10, then if the player rolled a 9, 10, or 11, possibly with a ten being "better."

How this "creates the future" is that if the player succeeds in a series of challenges, the DM can keep track and build up a collection of "bonuses" for each character (11). This is an Emergent Ability Bonus. So rather than saying something like a character has a 15 intelligence, and therefore a +2 bonus on all challenges related to intelligence, such as casting spells or deciphering things, repeated successes against such tasks results in such a bonus.

For instance, a character succeeds at a number of tasks to do with intelligence, that is, they roll the same roll as the DM (or within the range) and the DM notes that character as now having a plus-one intelligence. This would translate into, for example, if the DM rolls a 15 for a challenge, the player could not only roll a 15 to win the challenge, but also a 14 or a 16, using the plus-one intelligence to expand the range for success.

I hope it's obvious that the dice rolling system as I've described it so far is quite straightforward and simple to apply.

Then there's stuff like we have in D&D such as "damage" inflicted upon a successful hit. My hopefully not too oversimplistic idea is that rather than "doing damage" one would think in terms of "killed" or "somewhat hurt" or "seriously injured" in an attack would be part of the attack roll itself. For example, matching the DM's secret roll might immediately kill a monster, while a one-point-away might result in serious wounding, possible impeded future action, and two points away might result in "slight wounding". The DM could keep track of this and relay the outward appearance of the results to the adventure party.

By the same token, magical items, such as weapons, could be "clarified" (their magical scores added to as more and more challenges using the weapon succeed). So they wouldn't have to be "nailed down" at the time they are found.

I had the idea that everything about a character, even to the point of race and gender, be up in the air at the start of the campaign, to be decided in the course of play. This makes a lot of sense, since the goal is to get started up right away, and so leave the maximum latitude for imagination along the way. One idea leaps to mind. If a character fails a challenge roll, and the DM knows that the challenge would have a better chance if the player were female as opposed to male, or an elf as opposed to some other race, or the weapon they were using were magical (or more magical) as opposed to non-magical, and if these properties were as yet undecided, the DM could determine them then and there. That is, for example, if the challenge failed, but only by one (the DM rolled 17 and the player 16 or 18, or was off by one with bonuses applied, or suchlike), the DM could say the challenge succeeded, and also secretly determine the gender, race, or item magic sub rosa, to be used in play going forward, and ultimately revealed to the player at the end of the campaign, or if it comes up unavoidably in play later (like race or gender might have to be revealed). In other words, many if not most things would exist in an uncollapsed probability wave, only to be revealed (if ever) by successes in challenges that involve that "ability" or "bonus."

In sum, challenges are addressed with simple pairs of die rolls. The abilities of characters and items (such as magic weapons) are increased as more and more challenges are successfully met. It could be that two successes in a row leads to a plus-one, and three in a row a plus-two, and so forth. Care should be taken that the system not become exponential, e.g., that getting one level of bonus makes it easier to get to the next level, and so on. This could perhaps be avoided by only counting a bonus increase for a direct hit, i.e., an exact match with the DM's secret roll.

The Immunomemetic Environment 

It's immunomemes that keep a system on track, rather than the "regular" memes, that is, the memes that "define the character" of the individual adventure. Immunomemes are more workaday, dull, and may well be quite the same from adventure to adventure (though I'm not prepared to state that definitively at this point).

Answers like "I don't know" or "nobody's every tried it" or "all that've tried it have died/ never returned" from NPCs could be immunomemes, designed into the system. Bad things (or feelings of dread) among other characters, like getting hit by ricochets, attacked by monsters who are provoked inadvisably by foolhardy fellow players. For instance, if the DM wants to discourage a certain course of action, or "immunize" against it, other characters in the party could get attacked or have had things happen to them when another character takes an unadvised action.

The DM could pass secret notes to characters such as "You experience a sudden overwhelming sense of dread."

In other words, as with any macromemetically engineered system, one designs immunomemes against behaviors that one does not wish to see. In D&D, this would be things like wandering too far off the path of the adventure. In short, make a campaign as normal, maps, treasures, monsters, NPCs, but also create a collection of immunomemes for each step of the way to steer the campaign along.

What About Monsters? 

Monsters, and also treasure and other such features, may be generated on the fly as well, I think. One thing is that luck may be treated as a character attribute. There may be a number of others. For example, negotiating ability or persuasiveness can be paramount. There may be gender and race bonuses for this, e.g., if negotiating with or trying to persuade a NPC or monster of the same race and/or opposite gender (especially if there is a period of fertility, if we want to start down that particular rabbit hole). Should persuasiveness be treated as a separate attribute, or as a feature of charisma or wisdom or intelligence, or a combination, a "chord" or a "constellation" of abilities that combine (7)? I think luck with be an important, necessary, non-D&D attribute, but I also think that avoiding creating abilities that don't directly translate to D&D is an important goal.

What's a play-by-play of how a character might try to accomplish something, a task, an attack, or even collect part of a hoard? I feel I have to speak to this at least a bit with some clarifying examples.

The number and type of monsters who attack a party, how they fight, and if they win, what kind of treasure do they find, if any? And, how do luck rolls translate into more luckiness? Or other emergent abilities?

The DM could have each player make a luck roll, rolling a secret die for each player. If players already have an emergent luck score, that can be added. If any perfect matches occur, that player gets a tally on their emergent luck ability bonus (11), and possibly an increase on the bonus (8). If a party is about to be attacked by a group of monsters, all of them, say, make a luck roll (the DM does not tell them it's a "luck roll" but just tells them to each roll). Suppose, for instance, the worst outcomes are one or more monsters appearing for each character in the party. A natural perfect match on a luck roll might result in no monster appearing for that character, and a modified hit might result in a wimpy monster opponent. Failing the luck roll might result in a full-strength monster for the given character, or multiple opponents. All players rolling poorly could lead to a larger group of monsters, and stronger ones, i.e., bad luck all 'round. Strong rolling could result in far fewer monsters, or even none, or even non-hostile or able-to-be-persuaded monsters. The degree to which each player fails, or the party as a while fails, could translate into how badly things go.

Players could attack, and again have strength or intelligence or dexterity bonuses apply depending upon what they try to do. A lightly armored player might charge into the fray of more well-armored and less mobile opponents, slide along the flagstones between their legs and hamstring them with a dagger from below. Their light (but more maneuverable) armor needn't be a hindrance, or automatically mean they'll get clobbered by the enemies, so long as they can succeed their multiple dexterity and other throws, and then we also have things like speed that come into play. A player says, "I'm going to run from point A to point B before they can attack," and then has to succeed on a speed roll. This could be like the luck roll, but would actually translate into D&D ability scores later, e.g., one's "speed" or how much turf one can cover in a "round" (9).

Translating to D&D 

Perhaps a good way to think about what D&D abilities (1) look like in terms of the "memetically emergent" ability bonuses (MEABs) I've depicted above might be to look at an "average" score as 10 to 11 (2) and how each achieved bonus (6) or emergent bonus point translates into two ability scores above those. Or rather, if one has a +1 emergent strength bonus, then one must necessarily have a D&D strength ability score of twelve or thirteen...or more. 

If we leave off the possibility of emerged scores going down, we still have to deal with the lower end scores, i.e., if you fail enough times, does your ability drop below 11? More on that later, perhaps. 

When we think of a character going from meme-D&D to regular D&D, then things get quantum physical. The box is opened and the cat is either alive or dead, or at least it's feline ability scores are now set in stone, they are translated from "emerged ability bonuses so far" to actual ability scores. In other words, the possibility that a character will, after playing some potentially large number of further campaigns, replete with challenges against all abilities, have their ability scores continue to increase in response to the chance of, say, succeeding in three Intelligence challenges in a row, or four strength challenges in a row (natural rolls only, thank you very much), and thereby have effective Intelligence and Strength scores of 16 and 18, respectively. Once the given campaign ends (unless the DM keeps the emerged ability bonuses through future campaigns) and/or the character leaves meme-D&D-Land to play some real D&D.

Conclusions & Summary 

I need to hash out the idea of D&D ability scores below ten to eleven as a result of (consistent) poor performance. I've come up with how to handle the post-campaign scores (8), but this would not translate into "negative bonuses" during the mDnD (10) campaign. This may be fine, however, especially if you're surrounded by characters with bonuses, and thus have more success than you by comparison.

I've put out the idea of "luck" as an emergent ability bonus, which could relate to things like success in battle, luck in having weaker, fewer, or even no monsters attack, lucky guesses, better and more relevant treasure, and so forth. While in D&D the charts and chances are already stacked, and according to one's pre-determined ability scores and so forth, in mDnD they are both in flux and not determined, often only within some envelope.

Speed is an ability that would emerge and clarify in the course of play, and would then finally translate into an ability in D&D.

I still have a lot of imagination and filling-in-the-blanks on the part of the DM. I'm thinking in terms of the kind of monsters and other challenges that appear, and the kinds of treasure and such one runs into. This may be okay, but D&D lays all this stuff out pretty thoroughly in the form of many, many, many volumes of books and other materials.

I need some notational systems, like for how to tally against abilities (11). Also how to apply multiple ability bonuses against certain challenges.

Another issue is how to annotate everything. The DM will, as things are laid out so far, be doing a lot of the record-keeping. How much world-building is needed in terms of maps and NPCs and so on? If mDnD succeeds in avoiding the front-loading inherent in D&D campaigns, it may pay for it on the back end. The things that happen must be documented as they happen, or at least as the probability wave collapses. There needs to be some kind of documentation system to contain these things as they happen, presumably.

Again, a major point of this is to get started with little firmly decided and have the experience of the party determine what their properties and abilities are. I think it's off to a pretty good start. The point is the strong assumption that nobody knows how things are going to turn out, what is to be expected, or even who the members of the party really are, and that all this emerges in the experience of doing it.

Who would want to play this system? Seasoned D&D players might object or have problems with how things work, in terms of how they know nothing from the start, and they are effectively being told many things about who they are and what they are like by the DM and by the vicissitudes of the adventure itself.



Challenge Roll: the DM rolls a hidden d20 and the player attempting the challenge also rolls. If the values match exactly, the player succeeds and also gets an emergent ability bonus tally. If the player has salient bonuses which may be added to bring the two rolls into alignment, the challenge succeeds, but more weakly, and there is no EAB tally, indeed, the tally for any and all abilities are reset.

Emergent Ability Bonus: see "Challenge Roll," and "Memetically Emergent Ability Bonus." The bonus, associated with an ability, e.g., strength, added to a roll, by the DM, to make a failed natural roll still line up. For example, if the DM's challenge roll is eleven, and the character has a strength bonus of +2, if the player rolls anything between nine and thirteen, the challenge succeeds (with an eleven still being a powerful natural hit).

Emergent Ability Bonus Tally: whenever a character succeeds a challenge related to a given ability (with a natural perfect match), the DM tallies against that ability. The tallies continue to add up until the player fails a challenge, or fails to roll a natural perfect match, in which case the tally resets to zero. The highest continuous string of successes in a tally is the player's Emergent Ability Bonus for that ability, and the maximum is four, since that would translate to an ability score of 18.

Environmental Roll: see "Challenge Roll." The d20 roll made by the DM to which the player must roll in order to determine a success or failure outcome.

Optimal Memetic Resonance: see "Natural Perfect Match." Also referred to as "Hitting an OMR" (read "oh-mer". As opposed to "hitting a BoMMR" (read "bomer") or a "Bonus-Modified Memetic Resonance." (12)

mDnD: The name of the game. "Memetic Dungeon, non-Deterministic." How do we like this?

Memetically Emergent Ability Bonus: see also "Emergent Ability Bonus." Also known as "MEAB" or "meab" (say "meeb"). The bonus is memetic because it represents the players ability to deploy memes (take actions) with greater and greater success, and is awarded by the "environment" based upon whether past memetic deployments were successful. In other words, you successfully deployed that meme (more than once in a row), ergo you must be good at deploying those memes.

MEAT: "Memetically Emergent Ability (Bonus) Tally." See also "Memetically Emergent Ability Bonus (MEAB)" and "Natural Perfect Match." This is when the DM tallies a natural, unmodified success against an environmental challenge (11).

Natural Perfect Match: when a player rolls a d20 in answer to a DM's challenge and gets the same rolled value. This results in maximal success in the challenge, and also a tally against the given ability related to the challenge. A more catchy term might be wanted for this! Like MEAB and MEAT. This represents, memetically speaking, a resonance between the deployment opportunity made by the DM by making a challenge roll, and response by the player. The challenge may also come from the player, e.g., "I'm going to try this," to which the DM makes an environmental roll.



(1) Strength, Dexterity, Intelligence, Wisdom, Charisma, Constitution

(2) One rolls three Vegas dice, d6 x 3, to get an ability score, which thus ranges from 3 to 18 (3). The "average" score range, rather like seven in craps, 50% ten or below, 50% eleven or higher.

(3) There are bonuses associated with higher scores, which is how much over ten divided by two, rounding down. Thus a strength score of 12 would have a plus-one bonus. Given this, one could imagine "working backwards," so to speak, to generate a D&D ability score from performance over time (4). In the case of lower-than-average scores, we consider cases where no bonuses have been earned, i.e., the emergent ability bonus is still zero. Probabilistically, that translates into a score of nine to eleven or lower. This may be rolled when a character leaves the campaign (8).

(4) Generating "ability score" from "performance over time" could look something like this: given a challenge, if the player succeeds (5) twice in a row, an emerged bonus of +1 is awarded, and if later thrice in a row, +2, and so forth (8).

(5) Success in a challenge looks like the DM rolling a secret d20, and the player rolling a d20, and if they match, then the player has succeeded in the challenge, with an unmodified, or natural (no bonuses), roll. This can either be hitting a monster, avoiding a fall, making a jump, perceiving something, casting a spell, getting lucky enough to get a magic item in a hoard that's useful to you, getting enough money to pay the rent or that hitman out to get you, or what-have-you.

(6) As a memetic engineer, I need to come up with some good names (10) for these things, i.e., the bonus points you get from protracted success (see Glossary).

(7) There are any number of ways to combine emerged ability bonuses. The maximum bonus is +4, since that would be the bonus for an eighteen ability score in D&D. Taking the hypotenuse of an n-dimensional cube where each side is one of the salient ability scores (including "luck" and other non-D&D factors). The formula for this would be, for example, for a challenge involving strength, dexterity, and luck, and for which the character has +1 strength, +2 dexterity, and +4 luck, or strength squared equals one, dexterity squared is four, and luck squared is sixteen, or nineteen, the square root of which is 4.4. Do we round up or down? The highest all emerged bonuses can be is four (let's say) since as we mentioned that translates to an ability score of 18. If we had three such bonuses, it's 16 + 16 + 16 = 48 with square root of approximately 7. Six such bonuses should produce square root 96  or close to ten. How meaningful this is is open to question, as is how often such a high-dimensional stat would ever be used. In short, add up the squares of all of the abilities in question and take the square root of the total.

(8) Any time a player rolls a natural perfect match to a DM's roll, it's tallied (11), and if there is already a perfect match tallied, a bonus is in order (4). For instance, two matches in a row translates into a plus-one total bonus. If three in a row happens, that gives plus-two, and so on up to plus-four (the maximum). I still need to work out how negative bonuses (if any) are going to work. I guess if at the "end of the campaign" (collapse of the probability wave) and there are no bonuses on an ability, that means that the ability is at eleven or lower, or three to eleven. Since three is strangely low for an ability, we can roll a d6 and add 5, or six to eleven, another way is to roll d20, divide by three, rounding down, and add 5. And while a bonus of +4 translates into an eighteen ability score, with the rounding factor, a +3 bonus, for example, translates into either a 16 or 17. This would also have to be determined by the DM, for example, by coin flip or rolling d20, so a +2 emergent ability bonus would give a D&D ability of 14 on a roll of one to ten, and 15 on a roll of eleven to twenty.

(9) The concept of a "round" and "speed" might be rather loosey-goosey in mDnD (10). Again, succeeding multiple times means that one has a minimum speed value of that value, i.e., if a player proposes to cover a 30' distance, and succeeds at the roll, their speed ability then is set at 30' and can go up with further multiple feats of speed.

(10) mDnD or "memetic Dungeon non-Deterministic"

(11) Tallying against abilities. I was thinking something like, when a challenge succeeds, tally a kind of "O" and then when another challenge succeeds right after, put "O1" to show that the ability is now plus-one, but if one fails next, put an X next to the O to indicate the count starts over, "OX" and then when another success happens, "OXO" to start over, and then if another success happens, "OXOO" and then another, "OXOOO" and if it fails then, put a two to indicate the bonus is now 2, "OXOOO2" but if it continues to succeed then put another O, "OXOOOO"

(12) Success in a memetic environment is determined by deploying a meme that resonates with the environment, usually other people (agents). In mDnD, the DM sort of represents the environment, i.e., whether the player's actions "succeed." In a sense the "whole world" is the memetic audience. In this sense, having bonuses, meabs, translates to "power" in the memetic sense of the word.