Jul 17

Proposed GearHead: Caramel Skill System

Woman with braided hair standing in front of a city at night.

Another advantage of the new coloring routine is that it can handle subtle gradations. No more cel shading! That has nothing to do with the rest of this post, but I wanted to mention it.

I spent much of last weekend thinking about how the GearHead: Caramel skill system is going to work. This is likely to get a bit long-winded, so please bear with me.


These define the untrained abilities of your character. They will not change by much over the course of the game.

Health, Stamina, and Mental

Health measures the distance your character is from death. When health drops to zero, or a character is otherwise incapacitated, they are removed from combat. The party gets to make a Medicine roll to see if the character survives. Permanent injuries (which can be fixed by cybernetics) may result. It is based on Body and the Vitality skill.

Stamina is a measure of physical exhaustion; in game terms it is used to power passive abilities. Passive abilities include secondary defense rolls such as parry, block, and use of anti-missile and ECM systems. So, once your stamina runs out, you become far more vulnerable to attack. It is based on Body, Ego, and the Athletics skill.

Mental is a measure of mental exhaustion. Maybe it needs a better name. It is used to power active abilities, such as repair attempts and certain talents. It is based on Ego, Knowledge, and the Concentration skill.


Skills are trained abilities. These will change quite a bit as you spend experience on them. A skill should be something that your character does. Skills have a rank; if something doesn’t really need a rank, it should be a talent instead of a skill.

A character can have a limited number of key skills. These are skills which the character is particularly good at, and may be improved relatively cheaply. You can learn skills beyond your key skills, but will have to find a skill trainer to teach you at first. Improving non-key skills is more expensive than improving key skills.

There are nine basic skills in the Cavalier Package; every player character starts out with all of these as key skills, and may take a few other key skills in addition. I’ve been referring to the skills outside of the Cavalier Package as Specialist Skills.

Skill usage should be communicated to the player. In previous games, the exact function of many skills was a mystery. This communication can be accomplished in a number of ways: 1) Show when an option is made available because of a skill (maybe even showing options that would be available if the PC had the correct skill), 2) change passive skills to active skills, 3) clearly tell the player when a passive skill accomplishes something.

In general, specialist skills may provide the player with different ways of dealing with a problem. They might get used to short circuit a puzzle or obstacle that can be solved another way. They should not, however, be used to block the player from significant content/rewards.

Also in general, not everyone should need every skill. If your party has one member with a given specialist skill, that should be enough.

The Cavalier Package

These are key skills for all player characters. No matter what other skills you acquire, these can always be improved at the low rate.

Mecha Gunnery, Mecha Fighting, Mecha Piloting

Ranged Combat, Close Combat, Dodge

Vitality, Athletics, Concentration

Specialist Skills

The player will get some of these as key skills- two or three, maybe- and the rest can be acquired as non-key skills (if you can find a trainer). The non-key skills are more expensive to advance than the key skills are.

Repair: The ability to fix mecha and equipment. May allow destroyed mecha to be recovered after battle.

Medicine: The ability to heal people and animals. May allow downed fighters to be recovered after battle.

Biotech: The ability to repair biotechnological constructs and revive downed synths/biomecha. It can be used on dead synths to collect repair parts and other useful items.

Stealth: The ability to hide. This skill will now have to be actively invoked, and will function similar to its use in Dungeon Monkey Eternal.

Science: Used for crafting. You can make robots, improve gear, and so on. It can be used on dead robots to collect repair parts and other useful items.

Computers: This skill takes over Code Breaking and Electronic Warfare. At personal scale it allows doors to be unlocked and computers to be hacked. At mecha scale, it allows the use of active electronic warfare systems. With the right equipment, it may be used to hack certain hostile robots, turning them to the player’s side.

Performance: The ability to play music and get paid for it. This skill may allow Taunt as a talent.

Negotiation: Combines Conversation, Intimidation, and Shopping. Used to activate certain dialogue options. May cause enemy pilots to eject earlier. Improves NPC reaction score. Affects shop prices.

Scouting: Combines Survival and Awareness. This skill passively counteracts Stealth and makes encounters visible on the world map. It helps reveal hidden doors and other concealed things. It may be used on dead animals to collect meat and other useful items.

Dominate Animal: Use it to tame wild animals.

Deleted Skills

Mecha Engineering: Is spending points on Mecha Engineering fun? No, it is not. This skill is nothing but tears and frustration. Instead, I propose that from GHC onward all player characters can modify mecha to their heart’s content without having to learn a new skill.

Cybertech: There were a lot of problems with this skill as well. If a late game character suddenly needed to fix a permanent injury, they were often forced to decide between the injury’s stat penalty and one more skill’s xp penalty.

Weight Lifting, Initiative, Resistance, Spot Weakness: All boost abilities without offering a new play style. This means they are boring.

Investigation: It had no use except in Investigation missions; the thing is, we can still have investigation missions without a dedicated skill, and they’ll be more fun if the player isn’t forced to waste a skill slot.

Pick Pockets: Not nearly as useful as it is in Fallout. I don’t want to write the full burglary simulation that this skill deserves.

Taunt: Was less amusing than I had hoped.

Possible Skills

Meditation: Replaces Mysticism. Using Meditation allows a character to instantly regenerate stamina at the cost of mental. Unlocks martial arts/shaolin monk type talents.

Leadership: In GearHead2, lancemates automatically level up their skills after battle to match the PC’s renown. In future GearHead games, lancemates could automatically level up to match the rank of this skill.

Sniper: I got rid of most of the combat-related skills since they are boring choices. Unfortunately, that means there are few skills for a combat-focused character like a soldier or arena pilot. If I could figure out a way to make an interesting combat skill that is not an automatic must-have, I’d add it.

Areas of Expertise: Many pen and paper RPGs have a ton of knowledge skills. That wouldn’t really work in GearHead, but I thought it might be fun to allow the player to choose several areas of expertise for the PC. In some situations these AOEs would provide additional background information or dialogue options; this wouldn’t normally be useful, but it might be fun. The list of AOEs might include pop music, history, mecha designs, mathematics, recipes, arena champions, architecture, fashion, etc.

Jul 15

Third Edition?

I’m trying out a new sprite recoloring method for GearHead Caramel. In the original version (left), colors are defined by a midtone which gets scaled up or down to a particular value. In the new version (right), the brightest and darkest colors are defined and the color gets scaled within that range. This means that color warmth and saturation can change with value, which should make everything look better. The new method seems to make the colors brighter, but also less flat.
The next thing I have to decide is how characters are going to work in GearHead Caramel. The system in GH1/GH2 is an unholy amalgamation of the Mekton Zeta and Earthdawn rules. Many players have complained about the skill specialization limit, wishing it were possible to take all skills. At the same time, the sheer size of the skill list left many players confused. Never mind the fact that it also left me confused- adding enough content so that every skill would be worth taking was a neverending job. Balancing them was pretty much impossible.

Here are some thoughts about GH3e:

How can I allow each character build to be unique without the players feeling boxed in? If the player could count on finding lancemates with any required skills, that would take off the pressure to make the PC good at everything. I think the Harebrained Schemes Shadowrun games did a good job with this.

What’s the difference between skills and stats? How many skills do I need to have? Initially, I was planning to use a stripped down skill list, with each skill representing a career path. This skill would combine with different stats for different uses. For example, a Thief skill might combine with Speed to allow stealth, and combine with Craft to allow code breaking. After playing through GearHead2, though, I’m not sure this is the way to go. The GH2 skill list is smaller than the GH1 list, and you get fewer specialties. This means that from the midgame onward you spend most of your time hoarding XP and have fewer chances to spend it. This is No Fun.

The Shadowrun games largely avoid this problem by having multiple skill tiers. You need to improve Reflexes so you can improve Ranged Combat so you can improve Shotguns. The progression for a single ability is split across three skills, meaning that at any time you will probably have something that doesn’t require an obscene amount of experience to improve.

Before I move on, another thing I should mention about the Shadowrun CRPGs is that they dropped all non-combat skills. This largely solves the skill-specific content problem because whatever skills the player chooses, they should be applicable to just about every mission.

One way to keep different characters different might be to use key skills, as in Fallout. At character creation you choose the PC’s specialties. These skills may be improved cheaply. Other skills can be learned, but those will be more expensive.

It may be possible to move many of the noncombat skills to binary proficiencies- something the PC either knows or doesn’t. Most of these proficiencies would have a negligible effect upon the game, but allow different dialogue options or provide additional background information.

After reading the Rogue Trader rulebooks, I would be interested in expanding the character creation lifepath. The player chooses the PC’s life story, each step altering the character’s abilities. This seems like a good way to set the PC’s initial key skills, if I choose to use key skills.

In GHC, Stamina will be used to power passive abilities (mostly the extra defense rolls like shield block and antimissile) while Mental will be used to power active abilities. I am planning to add talents that act like the spirit commands in Super Robot Wars, allowing the player to boost abilities at the cost of MP.

What do you think?

Jul 12

Buru Buru Loading

GearHead Caramel can now load a complete SAN-X9 BuruBuru mecha from a text file. I can’t necessarily say everything is functioning yet, but it’s all there. The GHC version is 52.5 tons and is worth $117,280, which fits between the GH1 values (45.0 tons, $106,287) and GH2 values (53.0 tons, $194,368).

Here is the complete text of the mecha description file:

    name = "Buru Buru"
    desig = "SAN-X9"
    imagename = "btr_buruburu.png"
            size = 5
                    size = 4
                    size = 2
                    name = "Intercept Laser"
                    reach = 2
                    damage = 1
                    accuracy = 2
                    penetration = 0
            size = 5
                    size = 4
                    size = 605
                            size = 2
                    name = "Collar Mount"
                            size = 4
                                    name = "Swarm Missiles"
                                    reach = 6
                                    damage = 1
                                    accuracy = 1
                                    penetration = 1
                                    quantity = 20
            name = "Right Arm"
            size = 5
                    size = 4
                    name = "Right Hand"
                            name = "Shaka Cannon"
                            reach = 5
                            damage = 3
                            accuracy = 0
                            penetration = 3
                            ammo_type = Shells_150mm
            name = "Left Arm"
            size = 5
                    size = 4
                    name = "Left Hand"
            name = "Right Leg"
            size = 5
                    size = 4
                    size = 4
            name = "Left Leg"
            size = 5
                    size = 4
                    size = 4

It’s a bit more verbose than the GH1/GH2 mecha description language, but it gets the job done.

Jul 07

Gear Loader Working

I’d like to report another milestone in the development of GearHead in Python: that item on the right is the first bona fide Gear to be loaded from disk and displayed onscreen. It’s a human scale class 5 armor plate. As in previous GearHead games, I’ve written a parser to load mecha and item designs from human readable text files.

Why create my own parser when there are tons of serializers already available for Python? Two reasons. First, so I can maintain a format close to the GearHead1 and GearHead2 mecha description files. Using json or whatnot to design mecha would require too much semantic acrobatics. Second, because it was easy. It’s fully functional and only around 80 lines of code.

My next goal is to get an entire mecha file working.


Jul 02

Random Maps from Dungeon Monkey Eternal

I’ve been ripping out the skeleton from Dungeon Monkey Eternal and genericizing it for GearHead Caramel. Python has made this process far easier than it would have been in (non-Object) Pascal. The random map generator presented a special challenge because of how much game-specific stuff it relies on, such as the terrain list and wandering monster mechanics. To get around this problem I added an Architect class that translates generic map generation requests, such as “draw a floor here” or “add a door there”, into game-specific operations. Plus, because it’s object oriented, I can always subclass the generic Room or SceneGenerator classes to do anything I want. You can check out my progress at GitHub.

The character in the above screenshot was drawn by my son, and presumably won’t be appearing in the final game.

Jun 16

GearHead: Caramel Isometric Map Handler

A short video demonstrating the isometric map routines I’ve written for GearHead: Caramel. The big advantage of this new system is that it allows sub-tile movement of characters and animations; instead of jumping from one tile to the next, things can move smoothly across the map.

Apr 04

New Idea: The Story So Far

After yesterday’s post about GearHead-2, I started thinking about how the game reminds the player of the story so far. Instead of listing the history items as above, I wonder if it would be possible to generate a single paragraph describing the current state of the story? Something like:

The space colonies of the Free Commerce States have been attacked by the Red Mask Raider pirates. Led by Yecemk, they destroyed the Silver Knight garrison in Athera Spinner. Your victory against Yecemk earned her respect. Now, you plan to seek training in Wagner Spinner.

I don’t think I’ll be adding this feature to GH1 or GH2, but it’s something I plan to keep in mind for GHC.

Apr 03

More GearHead2 Thoughts

Sorry for the silence- it’s been a busy couple of months. I’ve been playing through GearHead-2 again. Tonight, my character Hawk finally earned the respect of his criminal space trucker rival.

One of the things that makes GearHead-2 very different from GearHead-1 is the scale. In GH1, you will travel between a small number of cities, most of which consist of a single scene. You will probably get to know a number of NPCs, especially if you join a faction and see the same few faces every time there’s an invasion. In GH2, on the other hand, you get to travel through a large number of cities, each of which contains many sub-scenes. You probably won’t remember any of the NPCs because you’ll be interacting with so many of them that they all become a blur. Also because of this scale, there will usually be plenty of missions available, even if your character hasn’t mastered Conversation.

In those cases where GearHead-2 allows the PC to form a relationship with an NPC (mainly the lancemates and the core story enemy), it works much better than GH1. Unfortunately, the other NPCs melt into a faceless blob.

The last thing I want to do before making a new release is to add graphics for spaceships and fortresses. As several players have noticed, these used to be rendered as meshes in the OpenGL version and I forgot to give them sprites after switching back to SDL. Hopefully I can get that done this week.

Jan 24

Random Story Generation Part 2: GearHead-1’s Core Story

To briefly recap part one, one way to create a random story generator is to write a whole lot of story fragments, arrange a subset of the fragments into a list, and count on your player’s sense of closure to assemble the list into a coherent story. In this part I’m going to look at how GearHead-1 did this.

It isn’t really correct to describe the GearHead-1 Plots as story fragments; instead, each one is a complete (though usually quite short) story. For now I’m going to skip talking about the basic Plots (which comprise most missions and random events) and instead just concentrate on the core story plots.

The core story is the series of Plots leading from the player’s introduction in Hogye to the eventual battle with Typhon. The choice of plots is based on three variables: the identity of the PC’s enemy, the mystery that confounds the PC, and the bad thing that happened to the PC. The possible states are:

PC Enemy

  • Unknown (Or, no enemy yet)
  • Character
  • Faction

PC Mystery

  • Resolved
  • Unknown (Or, no mystery yet)
  • Family Secret
  • PC has Amnesia
  • PC is not Human
  • Searching for Item

PC Misfortune

  • Resolved
  • Unknown (Or, no misfortune yet)
  • PC’s Family Died
  • PC is Incriminated
  • PC is Seeking Revenge
  • PC is Former Member of Enemy Faction
  • PC has a Lost Love

The PC’s starting scenario sets the initial values for these variables. For instance, if the player starts as a defector from Luna, the enemy is set to Faction, the mystery is set to Unknown, and the misfortune is set to Former Member of Enemy Faction. These three variables are collectively referred to as the context.

Only one core story Plot exists at a time. When a new core story Plot is needed, the master list is checked for Plots that match the current context. One of the matches is chosen at random. To continue the example above, from the Lunar defector beginning there are five possible first episodes: the PC could be pursued by a bounty hunter (which requires that the enemy be a faction), a price could be placed on the PC’s head (which has the same requirement), the PC could be sent to rescue a kidnapped professor (which requires that the enemy be a faction and the mystery be unknown), the player could be sent to intercept a secret plan (which requires that the mystery be unknown and the PC an ex-member of the enemy faction), or the PC could be accused of being a member of the enemy faction (which requires that the PC be an ex-member of the enemy faction).

Completion of a core story Plot will result in at least one of the context variables changing. For instance, completing the “Accused!” plot will change the misfortune variable to Resolved. The next Plot will be generated from the new context. This story generation method is similar to a Markov Chain, though I’ve been told that it isn’t exactly one.

There are several problems with the GearHead-1 core story generator. First off, there’s nothing to prevent the story from repeating itself. It can get trapped in a loop, and Plots which don’t alter the context can appear several times in a row. Second, it may be possible to arrive at a dead end- a context which has no available plots. Third, the context variables don’t hold a lot of information- having Aegis Overlord as your enemy is identical to having a local bandit gang or a supposedly law-abiding corporation as your enemy. Next time I’ll talk about how GearHead-2 solved some of these problems… and made one of them much worse.

Jan 16

Random Story Generation Part 1: Nothing More Practical Than A Good Theory

A visual depiction of how a Markov Chain story generator works.

I’ve decided to document my experiments in random story generation, in the hope that these notes will be helpful to other developers and hobbyists. This first post will define some of the terms I’ll be using and introduce some of the theory.

My first idea for a random story generator came during a literary criticism class I took during university. We were learning about reader response criticism, which is a model that focuses on the interaction between reader and text. In RRC it doesn’t make sense to say that a text has an inherent meaning; instead, meaning is something that is actively constructed by the reader through the process of reading. The author of the text, and whatever they intended the text to mean, doesn’t enter into it. If we don’t need to worry about authorial intent, why bother with an author at all?

The next piece of the puzzle came from Scott McCloud’s explanation of closure in Understanding Comics. A comic consists of a sequence of separate images; the reader uses closure to combine these images into a coherent story. I realized that the same would apply to a series of short narrative arcs in a computer game. Even without intentional connections between the arcs, the player’s sense of closure would interpret them as a meaningful story.

Finally, Vladimir Propp’s narrative functions provided a way to arrange these arcs so that the reader/player would be likely to interpret them as a coherent story. Propp analyzed Russian folk tales and discovered that all of their plots could be constructed from a finite list of story events which always appear in the same order.

From all this I got the idea to create a big list of story fragments, then use some kind of algorithm to arrange them into reasonably intelligible sequences. Turns out that’s pretty much how all procedural story generators work, even today.

I named the smallest narrative chunk a Plot, which I will capitalize to differentiate from the regular use of the word. A Plot modifies the game world for as long as it is active; it might alter an NPC’s dialogue options, or add an encounter to the world map. In GearHead-1 the Plots are quite coarse, defining an entire mission or a chapter of the core story. In GearHead-2 the Plots are much smaller; a single combat mission might consist of four linked Plots- one for the NPC offering the mission, a separate Plot defining the mission’s combat encounter, a Plot that is activated if the PC wins the mission, and another that is activated if the PC loses. The Plots in Dungeon Monkey Eternal are even more finely grained.

Using lots of small Plots is much better than using few large Plots. First off, if a quest is composed of multiple Plots, it means that the player won’t know how it ends just because they’ve seen this beginning before. Second, it allows reuse of code, which is a huge advantage for both bug control and refactoring. Instead of every mission having to define its own combat encounter, they can just call for a standard combat encounter Plot. Third, it greatly increases the number of combinations your system can generate, and in general you want your random plot generator to be able to generate a really really big number of combinations.

There are numerous ways to arrange Plots, including Markov Chains and context free grammars. The important bit is to have some way to describe the context of a Plot, so that Plots which belong together get placed together. Next time I’ll describe how the GearHead-1 core story context system works. Let me know in the comments if you have any questions.

Older posts «