August 11, 2023

Physics Sage Abilities: LLM Edition

The other day, Alexis mentioned he was testing out ChatGPT to suggest sage abilities. I thought it might be interesting to apply this to my investigation of a Physics sage ability and see what came up.

To set the proper context, I pasted a good bit of text from the Authentic Gaming wiki so the model would know what sage abilities and knowledge points were in the context of this specific fantasy role-playing game. Physics has a description already, so I continued with that:

Physics is a sage study in the field of Reality by which the character learns the physical laws associated with the movement of bodies and objects, vibration, sound, optics and light, fire and heat, the movement of water, air and earth, flight, pressure, electricity, magnetism and many similar branches and topics that give the character insight into the way that energy (including magic) ebbs and flows through the environment. This knowledge provides opportunity for the character to understand and ultimately control, to some degree, these forces.

Suggest a list of 3 amateur abilities for Physics

And here we go:

Certainly, here are three amateur sage abilities for the Physics study within the Reality field, representing a basic level of understanding and manipulation of physical laws:

Elemental Insight: The character gains a fundamental understanding of the elemental forces—earth, air, fire, and water. They can identify the presence and effects of these elements in the environment, such as detecting slight vibrations in the ground, predicting changes in weather patterns, or recognizing the potential for combustion in certain materials.

Basic Optics: The character learns to manipulate light and understand its behavior. They can create simple illusions, manipulate shadows, and refract light to obscure or reveal objects. Additionally, they can evaluate and enhance the reflective and refractive properties of surfaces to create basic optical effects.

Kinetic Control: This ability allows the character to manipulate basic kinetic energy. They can exert force on objects to move them from a distance, give a boost of speed to projectiles, or dampen the energy of a moving object to reduce its impact. This manipulation is limited to small-scale effects and relatively simple movements.

These amateur abilities provide characters with a foundational understanding of the physical world and limited control over its basic forces. As characters progress and gain more knowledge points, they can unlock more advanced and intricate abilities related to the Physics study.

Not bad at all. So far, it tracks reasonably with my own Amateur level. Kinetic Control seems a bit more advanced, though.

Repeating this process for Authority:

Energetic Resonance: The character can tap into the underlying energetic patterns of objects and elements. By understanding their resonant frequencies, they can induce vibrations, causing objects to shatter, create harmonic tones, or even manipulate the structural integrity of materials through controlled resonance.

Thermodynamic Manipulation: This ability grants the character control over temperature and heat distribution. They can create localized zones of extreme cold or heat, extract heat from one area to warm another, or generate intense thermal energy bursts. Their mastery over thermodynamics allows them to shape temperature-based phenomena in their surroundings.

Field Mastery - Electromagnetism: The character gains a deep comprehension of electromagnetic fields. They can generate and manipulate electromagnetic forces, creating barriers that repel or attract metallic objects, distorting light paths, and even generating bursts of energy akin to electrical or magnetic pulses.

I would rename the last one to simply "Electromagnetism."

For Expert:

Spatial Distortion: The character has reached an advanced level of understanding of space-time manipulation. They can create localized distortions in space, altering distances, warping trajectories, and even generating temporary pockets of altered reality that affect the behavior of objects and energy within their vicinity.

Waveform Mastery: This ability grants the character control over various types of waves, including sound waves, light waves, and more. They can manipulate sound to create illusions, control the intensity and frequency of light to blind or illuminate, and generate complex patterns of waves to manipulate and interfere with energy flows.

Chronokinetic Command: The character gains mastery over the manipulation of time on a small scale. They can momentarily accelerate or slow down the passage of time for objects or themselves, allowing them to react faster, age or heal slightly faster, and exert a degree of control over the temporal aspects of their surroundings.

These are certainly reasonable. Waveform Mastery might overlap with Electromagnetism, so more specific rules might be useful to distinguish these, perhaps in terms of their actual in-game effects.

And finally, the Sage:

Dimensional Weaving: The sage's understanding of physics transcends conventional boundaries. They can manipulate the fabric of space-time to create temporary rifts between dimensions, allowing them to manipulate energy, matter, and even consciousness in ways that defy conventional physical laws.

Unified Energy Manipulation: This ability enables the sage to perceive and manipulate the underlying universal energy that connects all things. They can harness this energy to achieve remarkable feats, such as transmuting matter, temporarily altering the laws of physics in localized areas, or even tapping into cosmic forces to generate cataclysmic effects.

Quantum Entanglement: The sage can exploit the phenomenon of quantum entanglement to instantaneously affect distant objects or energies. They can manipulate the entangled states of particles to transmit information, energy, or even physical effects across space and time without conventional limitations.

I think this has a great deal of potential. There's certainly editing to be done here, and actual concrete examples assigned. For example, we might say that Chronokinetics can allow the character to double their AP for one turn per day or allow 20% more healing overnight. So it's not as if we can just copy/paste a study, but a significant portion of the initial thinking can be offloaded to these models.

July 6, 2023

Trade VIII: Pricing Abilities

DM's Escritoire says:

Been a while, but Alexis' recent post generated an interesting comment from you regarding using his Sage system to effect manufacturing and prices. Care to share the direction you're taking. I recently watched his (ten year old) youtube and caught something I missed in the last third regarding references for "service providers." I may try to revisit this approach unless the Sage abilities shows more promise.

I would care to share the direction I'm taking, thank you for asking.

Alexis' approach is to make the yearly income of a person with a certain skillset equal to the amount of time needed to train them to that level. I took a different approach, but I'm hoping his solution can inspire me to close some of my gaps.

My trade system still operates largely as described several years ago. My upgrades have largely been relegated to a better processing system: I no longer calculate each stage of good in turn. Instead, I overhauled the recipe system so that each good is calculated recursively based on the price of its ingredients. This also makes it trivial to add new items to the system, as no recalculation needs to be done for complex manufactured goods.

Raw resources are still assigned based on rarity calculations: I attempt to make these as accurate to the real world as possible, but I also keep a file to override those results in case I want to force a particular resource to occur somewhere.

Industries/labor was a different matter. I'd started out giving each industry a "support value" based on MDME: how many people are required to support the resource of an industry. Many people pointed out issues with MDME's support values, but I do think the core idea is sound. But that leaves us back at square one: what number for SV do we use? My answer so far, unfortunately, is do what I feel right.

{
  "item": "potter",
  "cpUnit": 10,
  "tech": 6,
  "SV": 300
}

Occasionally, a price will not make sense, and I'll reevaluate the SV. But this is the base number I use to determine whither and which. Let's take a population of 1000. We randomize the order of industries and begin with potter as above. $\left\lfloor\frac{1000}{300}\right\rfloor=3$. So there will be $1-3$ potter references here. Let's roll a $2$ (the actual roll percentage is also calculated from the SV and population). This means that $2\cdot300=600$ of the population is tied up in "supporting" the potter, leaving 400 for the next industry we consider. These base references are then spread across the trade network just like anything else.

This process results in a unique fingerprint of industries for each city.

It is no complex matter to extend this to sage abilities.

{
  "item": "hunting",
  "cpUnit": 2,
  "tech": 4,
  "SV": 50
}

Our fiat values are again cpUnit (the unit here being /day) and SV. Hunting ought to be a more common skill than most, so we begin with these numbers and see how it turns out. Because each ability is assigned independently, we can define all Amateur Logistics abilities to have the same base numbers, and the final output for a given market will still be unique. Authority/Expert/Sage abilities can be likewise scaled up appropriately from the Amateur values.

This lets the party acquire a hireling with a bespoke set of abilities and gives a total monthly price based on the specific set of skills they desire.

Future considerations:

  • Better understanding of how the various fiat numbers are derived, something more internally consistent like the price of training/apprenticeship
  • Replace some industries with appropriate skills, e.g., Potter can be replaced by Pottering. As mentioned above, I think this is probably a better approach overall, since it more directly ties the  market network into the sage system

March 10, 2023

Combat Simulation IV

I have made a great deal of process on the combat simulator, to the point where I was able to use it to run a simple combat for my party. That was a useful exercise, as it revealed a couple of places where I needed to clean it up. Two main uses for this tool have emerged more clearly in my mind: running a combat under various conditions to get a better idea of difficulty, and automating the choices of NPCs during a manual combat. Currently, it's not super easy to input PC actions without manually writing JSON - which can be done of course, but is not very scalable.

But it works! I'm trying to avoid premature optimization. I also haven't added some of the more complex effects yet, such as spells or difficult terrain conditions.

Among other benefits, the tool adds up all gained experience (and will work for hirelings and henchmen later on), and provides a useful double-check on the final disposition of health. I encourage players to keep track of their own health, but things slip through the cracks, and more so online.

This fight was rather simple and even blase; an orcish ambush (red) on the party (blue), who was inspecting the aftermath of yet another previous (and successful ambush). With so few combatants and evenly matched in numbers, there is little opportunity for much strategy once melee is met.

True, the display leaves a lot to be desired here. But I don't intend to spend a great deal of time on that, preferring instead to focus on usability and features.

January 6, 2023

Combat Simulation III

Let's continue with movement selection.

We're still only considering melee. I'll address ranged movement next, and then move on to the actual attack.

The following considerations are taken into account:

  1. Distance to target. A closer target is better, or considered a higher threat. For now, I'll consider all enemies to be of equal threat all things considered. I have a planned upgrade to allow creatures of higher intelligence to weight threats based on the enemy's HD
  2. Required AP. Again, closer is better, in this case because the character can move there with AP to spare and possibly attack
  3. Flanking. Could possibly make this available only to higher intelligence creatures also

This generates the following maps. I'm still using the constrained movement but I'm thinking through strategies for a freer (and therefore more efficient) system. The darker the hex, the higher the movement score. Notice in the second image that the preferred move would flank Blue.

Blue guy (right)

Red guy (top right)

This is probably fine for now. There's always more things that come up, but I'm in a move-fast-break-things mood so we'll fix those as they become problems.

January 2, 2023

Combat Simulation II

Previously, I discussed basic, unhindered movement. But movement is rarely unhindered, and so we'll need to add constraints and obstacles.

The trickiest part of this was figuring out how to represent a battle map in memory. I want to use an inefficient format like JSON because its human readable and we can check it into source control easily. As I was going over old combats, I realized that this tool could be very helpful to review both your own historical fights but also those of others, like a quarterback going over old game film.

For now, I'll try to use the simplest possible case. Map hexes are flat-topped and have an equal movement cost. To represent buildings or other structures, one of the sides of the hex is denoted "blocked": that is, no movement may take place through that side. If all 6 sides are blocked, the entire hex is impassable: perhaps a large tree. If two or more sides are blocked, the hex is considered a partial hex and movement/combat buffs may apply.

The following diagram shows two groups of combatants: blue and red. The thick black lines show the outline of a small keep, with several broken walls providing gaps.

Let us say that the blue combatant entering the keep at the bottom wishes to move elsewhere. He'll have to spend 1 AP disengaging from combat, leaving him with 1 or 2 AP for movement. His possible moves are then as follows:

For our purposes here, we'll ignore certain moves that would end in an opponent's hex, such as charging or overbearing.

Later, we can make the map more complex by introducing travel costs for each hex, but this will be fine for now.