Quick break: Handle With Care

Just a quick break from the hypothetical game design to get back to pointing out interesting games:

Radiator Vol 1-2: Handle With Care is out, and it's a big step up in atmosphere and storytelling from the first volume, Polaris. I have to say the experience is pretty right on - just going into it cold I felt like I went through all the realizations and emotions that the designer intended, and it was the right length.

Like other games I've been looking at recently, this is more an experience than a game. The game element isn't fantastic actually, but the emotional experience element is amazing, and huge leap from the previous volume. The question a piece like this raises for me is how these sorts of thought-provoking experiential games can be adapted into mainstream game design. The Half-Life series (whose engine this mod of course uses) does a very good job in terms of atmosphere, but there's something about character development and emotional response that this volume of Radiator does so well that intrigues me. I guess though it is sort of like looking at a beautiful painting and asking why music videos can't use something like that - it's a different medium seeking to achieve a different goal. Still, I can't help but imagine a big sprawling game that learns from the experiments being done in the Radiator series.

Maybe though the lesson is more about what games can be than about concrete methods. I think you have to draw a line between games that are created to be fun and diverting and games as a medium for making an impact on the audience and thought-provoking. Not to say that one can't have elements of the other, but Handle With Care is most definitely thought-provoking first and foremost, and (for instance) Portal is most definitely on the fun side of the equation.

One of my favorite games, Planescape:Torment, was both fun and thought-provoking. But even there the emotional response is more abstract and easily glossed over.

What it comes down to, I think, is how the designer approaches character. Handle With Care is really all about the main character, and does a phenomenal job of putting the player into that character's shoes in a novel and engaging way. Even so-called roleplaying games like a Planescape: Torment or Mass Effect distance the player from the character by quite a ways. The effect of Handle With Care is to brazenly confront the player with the heart of roleplaying - becoming that character and really exploring them. In Mass Effect I don't really think about Shephard's character other than to say "I'm going to have him act like a badass" or I'm going to have him be a merciful guy here." But when presented with choosing whether, as James, to repress or dredge up memories of your relationship - and then to witness the consequences - that gives you a connection to the character that is very hard to achieve.

I guess the point that I take away from that also has to do with how your control of a character affects the outcome of the game. I didn't really know what the result of my choices would be in Handle With Care, and so the ending that my choices resulted in really served to illuminate my understanding of the character. In Mass Effect your choices are about action more than emotion - Do I act ruthlessly and kill this person, or mercifully and free them? But Handle With Care presents you with a choice about self - do I lock away my bad memories, or do I set them loose and dwell on them? - and the different endings tell you that that character had all these possibilities inside them all along, and the direction their life takes is about self-reflection, not action. The internal struggle that conventional RPGs gloss over in the "Do I kill them or set them free" roleplaying choice is the heart of gameplay in HWC. Perhaps that is what we can learn from it.

MMO Design: Proper Explanation

I just came across a great succinct series of blog posts by Brenda Brathwaite about game design documents. So I figured as long as I'm spitballing here I might as well make an effort to format it in a clearer fashion. In an attempt to follow her model, here's the game so far:

A Post-Apocalyptic MMORPG

Core Statement: Survive and shape the rebuilding of society in a violent and devastated world.

Feature Set:

Customize your character's skillset

Engage in fast-paced combat with guns, explosives and hand-to-hand combat

Complete quests that tell the many stories of the wasteland

Acquire gear, vehicles and property

Build a network of friends and allies amongst the inhabitants of the wasteland

Influence the character and growth of settlements

Become a powerful leader of your own band of survivors

Protect your vision of the future from other players seeking a different path

Well right off the bat that makes it a lot clearer. The core goal for the overall story feel of the game is somewhere between the player-made world and territory battles of Eve Online and the deep quest and raid narrative storyline of World of Warcraft. I chose the word influence in the feature set because I do want it to be a more behind the scenes type of territory-building, rather than the overt claim-staking of Eve.

MMO Design: Allies and Contacts

Allies, as mentioned in the previous post, are going to be a vital mechanic to the game. Star Wars: The Old Republic is employing NPC sidekicks as a core element to the game, so that a character always has someone by their side to influence and react to the player's decisions. It's a great tool for storytelling, and in this game I want it to also be a great tool for world-building.

Every player will be able to have one Ally traveling with them and fighting by their side. Additional NPCs that the player befriends along the way will become Contacts, and the player will be able to visit them wherever they reside to talk, trade or recruit them as a traveling Ally. Increasing the size of your number of Contacts will be a way to expand your power base in the game, giving you more options of Allies to accompany you, and NPCs to influence. These Contacts you befriend are a major resource as they are used to shape the game world as well as to provide perks and combat ability.

Contacts are acquired through a variety of means - they may be won over through questing, purchased from Slavers, or met in random encounters. They may have personality requirements, so if you have the wrong kind of reputation they may not be available to join your network of Contacts. And of course once a member of one player's network they wouldn't join another's.

Through conversation with your Contacts you can get them to perform a number of tasks - changing their place of residence, enlisting in local law enforcement or crime syndicates, set up a shop, or even run for public office to name a few. Depending on their skills and influence you can use them as a valuable resource through which to control events in the game world. If the gunsmith we rescued from Slavers moves to a town where he joins the local law enforcment and eventually becomes the Sheriff, your past history with him can be leveraged to get special perks, like influencing the laws of the town. Of course that also puts him in a position to be assassinated by bandits, so it has its risks as well. In essence your Contacts are a resource that you build up and can be depleted if they die, but are a powerful way to effect change in the world without direct action. A player with a large Contact list who uses them in effective ways can be a major force in game without ever firing a shot.

MMO Design cont'd: Influence

The buzzword for the mechanic I want to design here is Influence. In a very granular way I want players to see their actions having an effect on the game world. I want all players to be able to make these changes, and to set up a struggle between players with different visions to act as a motivation for PvP and PvE.

The upcoming MMO Fallen Earth has a set of factions based around the idea of what the future holds for the post-apocalyptic world it is set in - one faction wants to restore civilization, another wants to preserve the anarchy that currently exists, another thinks the devastation could be a boon to nature, etc. We will take some similar broad themes and build them into gameplay and story development, but hopefully in a less overt and more granular way.

Law

The first example of this mechanism is the rule of Law. I used an earlier example of a town in which firearms were restricted, and NPC police cracked down hard on gunfights in the streets. So how did the town get that way? Civilized areas in the game will have a Law rating, based directly off the number of NPC police inhabiting the area and their morale. As in the classic western - a town with a nervous and cowardly sherriff will be much more lawless than one with a courageous team of lawmen walking the streets. Lawbreaking NPCs like Slavers and Bandits will be fewer and farther between in areas with strong Law scores, and those areas will have rules and regulations regarding combat.

So how do players impact the level of Law in an area? Some quests will be law enforcement or crime quests, and their completion will change the attitudes of NPCs in the area - putting police on higher alert and instilling a greater sense of lawlessness in the population. More crimes committed in an area may inspire certain NPCs to concoct criminal schemes of their own, opening up further crime quests. Likewise cracking down on crime in an area will embolden the police forces and encourage stricter regulation.

Quests are a great way to shift the balance. But I don't want too many repeatable quests as that is tough on immersion. Rather, I'd like quests to open doors for the player to influence their surroundings in other ways. In many ways Law is simply the absence of Crime. So a player who has allied themselves with the anarchic forces of Crime through questing may simply commit criminal acts in an area. Crimes going unpunished will decrease the Law score of an area. Other players may enforce the law once they've made good with local lawmen NPCs by apprehending or killing player and NPC criminals. Since it is easier to be casually criminal than to interrupt a crime, players abiding by the law in an area count towards the general atmosphere of Lawfulness.

Finally, NPC allies (to be fleshed out later) may be able to be influenced to take up residence in an area and contribute positively or negatively to that area's Law score. If the player in an earlier example has rescued a gunsmith from Slavers and that gunsmith has become the player's sidekick, the player may be able to encourage him to lend his skills to a town's Sherriff. That gunsmith NPC will then leave the player and take up residence in the town, contributing to the quality of the NPC police force and potentially even becoming a member of it. Likewise the player may be able to convince the gunsmith that 'if you can't beat em, join em' and have him join a Bandit gang in the area, increasing the crime levels locally and adding a member to their gang. NPC allies will be a valuable and interesting way to impact the game world, because players essentially will be able to install them in locations or groups to act as a liason and to change the face of the world. Those allies become a type of resource that you can earn and spend to make changes.

In the same fashion as Law/Crime, broad conditions like Economy, Technology, Wilderness, Culture and Politics can be affected. Towns may grow to be bustling economic hubs, used by players as a refuge and place to trade (much the way Jita in Eve Online became a de facto trade hub without it being built into the original game that way). Political structures may develop, trade between areas and even war between towns in which players may take sides (or may have instigated in the first place!) All through players gaining and exerting influence on NPCs.

So in general the Influence system comes down to: 1) Developing a relationship with NPCs through questing, 2) Exerting that influence to push the environment towards the player's vision.

Groups of players would have even more power in this sense, because they could jointly direct their efforts to build up a trade hub, start a war between two towns, or eradicate the goody-two-shoes lawmen of an area in a bandit raid.

Because attacking NPCs may be a part of this system, something has to account for whether they can actually be killed or not. Since players of course will never truly be destroyed, if NPCs can be destroyed permanently then you might have a depopulation problem. But at the same time if you can't ever truly kill someone, it becomes a bit silly and the value of killing them is diminished - I do want permanent changes to be an effect. So our hypothetical town sherriff should probably be killable, permanently - but we need a way to balance that with the fact that "if something is possible, it will be done" and figure out what it means when a player decides they want to assassinate a public figure in a town. Of course we do also have a mechanic by which players can install public figures in a town - so it might simply be a battle between building and killing. Plus if the right behaviors are built into the NPCs, it could be next to impossible to kill them - if they are protected by bodyguards, or there is a strong local police force to eliminate any lone assassins. In general the struggle of what would otherwise be called griefers to kill off NPCs should really just be part of the game, since it's going to happen anyway. So every NPC will be in theory killable - but the tactical difficulties of doing that are part of the game. If players form a vast army to invade and try to wipe out an NPC town - more power to them! But if somehow a town forms out of the post-apocalyptic wilderness to become a major hub of civilization and law - well, that's pretty cool too. Groups of players dedicated to the preservation of law and order may naturally spring up, and they would add to the town by rescuing allies and installing them, making the town grow. In this way we could really have a living breathing world that the players themselves are vital to the development of.

MMO Design: Power Base Endgame System

Going off the last thoughts, the next element I want to flesh out for this hypothetical MMO is the endgame. That is, once the player has mastered the basics and fully built their character, what is it that they do with them?

I've already decided that I don't want the endgame to be about continued self-improvement through increasing combat power. In fact, I want combat power to be relatively static to emphasize player tactical choices over most other considerations, so I don't want a fight to hinge on which player has the magical assault rifle +5. Instead, I want players to use their combat power to acquire other kinds of power - political and economic for instance. So by completing quests players will acquire broadened capabilities, rather than heightened capabilities. A good example would be if in WoW, instead of doing instances and quests to get gear and experience, you did them to acquire flight paths, mounts, bank slots and the ability to create or join a guild. All of these things enhance your gaming experience but aren't directly related to your combat capabilities. I'll call these collected perks a player's "Power Base".

Expanding your Power Base

Grand Theft Auto is a good example of a game where expanding your character's power isn't about increased combat ability. Instead as you progress through the game you gain access to more safe houses, more friends, more money and more cars. So let's brainstorm some ways that characters in this MMO can acquire expanded power.

Vehicles

Like mounts in WoW or cars in GTA or the car in Fallout 2 - acquiring a vehicle is a sweet perk.

Storage

I will probably want a relatively realistic inventory system, since what gear you choose to carry around is important - so having a place to store your extra stuff is a very valuable thing to unlock.

Allies

As the game incorporates a living breathing NPC world, the characters will be able to manage their relationships with NPCs and factions to garner allies who will provide them with services and combat backup.

Networking/Communication/Information

I'd like to see chat channels and the ability to create a guild or form a party or raid be something that characters acquire through questing. More critical communication abilities would be something you get early on, but having some depth to the ways you can communicate with other players would allow some players to specialize in creating information networks.

Zone Access/Safety

Another type of power would be the ability to unlock access to new helpful zones or safe areas. A safe haven town in a hostile area or a shortcut through otherwise impassable terrain would be examples of this.

Fame/Titles

Imagine if NPCs talked about you based on your deeds? Other players hear your name before they ever even meet you.

Affecting the Game World

But by far, the most powerful ability a player can have in an MMO is the ability to effect change in the game world - to alter the course of story events. This kind of power is relatively unheard of in MMOs simply because there are so many players that it is easier to give them personalized experiences than experiences that project change into the public space. However, since this is my dream MMO - that's exactly what we're going to work out. The game's story design will be guided by the influence of players. The day to day 'work' of a player won't be just about earning gold, grinding experience or finding loot, but about making bold changes to the game world, molding it to their desires. All the above features will be part of a character's personal power base, but the height of their exertion of power will be the struggle to control the course of history in the game world.

MMO Design: The System So Far

Time for a quick recap on what's been worked out so far:

Combat

Characters making attacks on other characters are making a roll against the target's defenses. The Attacker's FPS aim, character skill and environmental conditions (like cover, visibility and weapon characteristics) compete with the Defender's character agility, armor and awareness (AAA) of the attacker. Should the Attacker's roll win, they score a Kill Shot on the Defender, potentially killing or incapacitating them. If the Defender is unhittable due to armor and character agility, each successive attack that fails due to armor or agility grants the attacker a Concentration Point bonus on their future attacks. Thus, heroic defenses will eventually succumb to repeated attack if the attacker is not killed first. This also means that in group on group combat, there is no advantage to focus fire. In fact the most heavily armed and armored combatants should be trying to engage with the most heavily armed and armored enemies, since they pose the greatest threat. Weaker combatants should seek out weaker enemies as they have the greatest chance to eliminate them.

This results in a more chaotic and realistic combat system than the traditional Health bars and tank/healer model.

Skill Categories

Characters are built by training skills in a number of disciplines. Weapon and Support skill trees are the two main categories that affect combat. Non-Combat skill trees may also exist. Weapon skill trees include things like Pistols, Rifles, Blades, Martial Arts and Heavy Weapons. Support skill trees include First Aid, Armor, SpecOps (Stealth), Survival, Engineering and Vehicles. A Skill Category is basically a collection of related skills that a character may choose to train some or all of.

Learning and Experience

Training is the act of selecting what skills you want your character to be able to use. Practice is the act of using a particular skill enough that the character gains a special bonus based on the tree that skill is in. Depending on what skills a character uses most they become Practiced in one Weapon category and up to two Support categories. Being Practiced in a skill category is generally indicative of a playstyle, and may grant a special ability as long as the character stays in Practice. For instance, a Practiced Pistol user may gain the ability to perform a Quickdraw and a Practiced First Aid user may gain the ability to revive a player from the brink of death.

In this way players are free to select they want through a classless skill system, and their playstyle will unlock abilities that complement their chosen path.

The next question is whether there is any further character growth beyond the selection of Trained skills and Practicing. A game like WoW utilizes a combination of leveling to a cap and then acquiring better and better gear in order to progress. The plus side of WoW's leveling system is that it is very addictive and satisfying, however one might ask "why not just start at the level cap?" In a lot of ways the whole point of a WoW-style leveling system is to maintain a continual goal in the game: self-improvement. Initially self-improvement is very straightforward, as you slay monsters and complete quests to earn experience and level up. Upon reaching the cap the self-improvement game changes slightly and it becomes about gear.

Eve Online takes a slightly different approach but ends up with a similar scheme: Skill training is continuous and improves performance in a given ship, which is further improved through equipment. However it is quite easy to reach a point at which you are as good with a particular ship as you can possibly be, and it is just a question of whether that ship is the right ship for the job. Around that point your self-improvement becomes more about versatility than potential. Eventually self-improvement becomes about projected power through political, economic and military power that isn't necessarily based on personal skill-based performance.

In many ways it is that later endgame that is the most interesting. Eve Online has the interesting characteristic that in an isolated PvP encounter skill points and equipment matter, but they are not critical - a less skilled character (or group) in a less expensive ship (or fleet) is quite capable of killing a more highly skilled and equipped player or group, with the proper tactics.

I lean more towards the Eve Online model in that I would prefer a character's worth to be based not on the skills and equipment they have but on the fruits of playstyle and tactics. So rather than focusing on self-improvement through a leveling system or gear, I'd like to build a different model into this game. A feature that I was intrigued by in The Agency was the ability to collect Operatives, NPCs that become part of your own personal spy network and expand your capabilities by allowing you to call them up to tail a suspect, hack a computer, etc. I love that concept and would like to build something similar to be the endgame self-improvement for this MMO - so rather than becoming stronger and stronger as a combat force, characters will cap out in raw capability relatively early, and turn their attentions to utilizing that combat capability to achieve and acquire power in a more abstract sense. Also, Grand Theft Auto IV has a similar concept, where by making friends with various characters you can call them up to get helicopter rides, buy weapons without having to visit a store, get some thugs to back you up, etc. These types of perks are the way to go.

Some possibilities include acquiring vehicles, creating player-owned structures in the game world, gaining followers and sidekicks and improving their capabilities, vanity items and of course garnering fame and fortune. The basic formula should probably be that success in the game translates into minor gameplay perks and major coolness perks. I would prefer to avoid a situation where success unlocks further gameplay (as in a WoW raiding progression) as I think there are plenty of MMOs that follow that PvE model. I'd rather see few obstacles to reaching the pinnacle of character development and more interesting and optional sideroads for characters to explore once they've mastered the basic gameplay.

Still, it is important for characters to have goals and feel like they have achieved things, so I guess I shouldn't neglect that. Achievement point systems are an interesting way of giving that feeling of accomplishment without impacting gameplay. If WoW instances didn't grant gear or experience but instead granted achievements for completing their storylines, would people still do them? I think that might actually encourage better storytelling in the instances - since you'd be going in there to explore and have fun rather than to grind out a level or hope that a piece of gear drops. Likewise I don't want to have prized gear dropping from monsters or being acquired by running an instance - instead I'd rather see characters defeating enemies to accomplish story goals.

In other words - the Slaver Boss you are sent to take out isn't going to drop a special pistol. But eliminating his base might allow a local gunsmith you freed to begin producing weapons again. The next post should focus on brainstorming more specifically the types of endgame self-improvement that will drive players to play, explore and achieve.

Designing MMO Combat cont'd

So in the last post we came up with a neat alternative to the traditional hit points and tank-healer-dps model for MMORPG combat. Next we'll continue to flesh out the idea by taking a quick look at playstyle and classes for a hypothetical modern setting (or post-apocalyptic) MMO. Once again we'll come at this from the angle of trying to reinvent the elements that annoy me about current-gen MMO mechanics.

In Star Wars Galaxies one thing that bothered me was that Pistoleer, Rifleman, Commando, etc were entire classes built around using a single type of weapon in a single combat style. This is okay in a raid situation where every specialist fills a role: Tank, range dps, melee dps, debuffer, etc. But inevitably you come up against the problem of balancing a specialist class across an entire game, including PvP. You either have to give the Rifle specialist some kind of strange close-up combat specials, or AoE attacks, or you have to acknowledge that the player that chose to specialize in Rifles is just out of luck when encountering sub-optimal combat situations like melee. Likewise you either have the Pistoleer and the rocket-spewing Commando do the same damage, or you relegate the Pistoleer to being a sub-optimal DPSer and try to make it up to them with crowd control abilities or some such.

You see this argument a lot in regards to hybrid classes in WoW: Performance in a raid setting is only a slice of the overall game, but raiding is about maximization of specialties, so it is cold comfort to a class able to do many things fairly well (healing, tanking or dealing damage) when they feel that their raid performance is unfavorable compared to a class that has no options to switch over to healing or tanking or has more limited crowd control abilities.

In essence, I feel that class systems bring a lot of problems with them, and so I'd like to avoid them. Because our combat system does not include the traditional tank-healer-dps model, I think that we are free to reimagine how players decide "What is my job in this world?" without having the answer packaged as a pre-fab class. By virtue of that we're not as beholden to balance concerns, creating content and playstyles explicitly balanced in terms of fun and potential for every class. We are also free to focus exclusively (for the moment) on combat and not worry too much about non-combat skills that are often weighted into a class' overall balance.

Don't Bring a Knife to a Gunfight

The first principle I think is important in terms of combat balance is that different styles of weapons do not need to be equal in all situations. Because we won't have a Blademaster class and a Sniper class, we don't need to be concerned about balancing knife combat with rifle combat. Instead a player will likely carry multiple weapons and choose what skills to learn or focus on in a variety of weapons, switching between them as the situation warrants. We can also devise unique strengths and weaknesses for various combat styles that will add another layer of tactical thinking to the game.

For instance, say we have a legal system in the game in which NPC police don't allow gunfights in town, and anyone brandishing a weapon is arrested. Perhaps gunfire would attract the local law enforcement. In a town like that, being skilled with a switchblade would come in pretty handy. By contrast, out in the wild wastes the ability to pick off an opponent from half a mile away with a sniper rifle is a clear advantage. A pistol may be able to be drawn more quickly than an assault rifle slung across your back, and surely one wouldn't be advised to use explosives or rocket launchers in close quarters or indoors. So our players will likely follow the Mass Effect model of carrying multiple types of weapons and choosing the best one for the situation.

With that understanding, let's step away from weaponry and consider some combat styles that we might want to allow players to learn or specialize in through skill training. Here's a quick brainstorm of some combat playstyles we could create skills for:

First Aid - although we aren't going to have a 'healer' role, the Kill Shot system could be further adapted to allow for non-lethal kill shots to extremities resulting in the need for splints, stopping blood loss, administering of painkillers or futuristic healing gels, etc. An ability to do some quick first aid after a firefight to get people back on their feet would be a nice set of subskills. It would be fun to have this be a fleshed out set of skills to pick some or all from, so a character's first aid training could range from the very basic to full out high tech battlefield medicine.

Armor - It's somewhat traditional to require training to use various types of armor. In our world we might even have advanced mechanized armor that requires specialization to use - something like Fallout's Power Armor. Relying on Armor in our setting might indicate a more head-on combat approach.

SpecOps - Stealth skills would of course have a role and mesh well with a variety of weapon choices, from camoflauging an anti-tank weapon to eliminating an unsuspecting guard from behind with a knife. Of course defeating security measures would be a great sub-specialization here.

Vehicles - From simple driving that all characters will probably have skill at up to operating more complex vehicles, a variety of training and specialization could be included for all sorts of rides.

Survival - Depending on whether it is appropriate to the setting, a whole set of skills could be gathered around the principle of survival in the wild - tracking, scouting, orienteering, moving quickly across terrain, recognizing natural dangers and using camoflauge.

Engineering - Remote control of drones, use of mines and tripwires, battlefield repair of weapons and vehicles, etc.

By allowing players to mix and match skills from these broad categories and decide on a number of weapons to train in, a number of different playstyles emerge that are often similar to pre-packaged classes but at the same time can also be highly flexible to allow creating the kind of character each individual wants to play without locking them into a class. We will build the skill system to account for the fact we expect all characters to be pretty good with a number of different weapons and to have the ability to flesh out their particular combat style by training in some of the above areas, mixing and matching as they see fit. Because I don't want people to be too locked into specialization, I'd like to structure skills to allow full use of at least 6 things, be they weapon styles or specialization categories like the above brainstorms.

Right now it is actually looking fairly similar to the Star Wars Galaxies skill system in that regard - in which it was basically possible to max out 3 skill trees. However, I was never a fan of the way those skill trees worked, so let's try something a little different with how characters become able to do the things they do.

I'd like to examine skill learning in two ways: Training and Practice. Training is the initial teachings needed to begin practicing a skill. Without training you can't use a skill. Practice is the experience you get as you use a skill you're trained in. The problem with Practice is that it inevitably leads to a grind mentality. Leveling up through the Pistoleer class in Star Wars Galaxies consisted of simply repeatedly killing things with a pistol. Incredibly boring. By contrast, in Eve Online there is no grinding you can do to increase your skills - you simply train them (although it does take time) and you are able to perform certain tasks. I'd like to find a nice middle ground between these that rewards practicing of skills but doesn't require or reward a grind. Here's what I'm thinking:

Training

All characters get a certain number of training points, which they can use to unlock the skills of their choosing. These unlocked skills represent the floor of their capabilities. No matter what activities they perform, they always have access to those skills. Some training may open up the ability to acquire more specialized training, but the mechanic is the equivalent of spending a talent point in WoW.

Practice

Upon unlocking that training, one's ability to perform a particular skill is improved by Practice. Think of Practice as an overcharge booster on a skill. So if you are Trained in Pistols and Rifles, but you spend a few minutes in combat using your Rifle and keep your Pistol holstered, you have more Rifle Practice than Pistol Practice and consequently your Rifle shots get a boost to accuracy and damage. With enough Practice, some skills may perform much better than without Practice, or allow you to perform special maneuvers - perhaps a Practiced Rifle user has increased accuracy while moving or can make called shots to specific body parts. A Practiced Pistol user might have a quickdraw ability and a Practiced First Aid user might be able to save characters that who have suffered from a Kill Shot and would otherwise be dead. Practice caps out at a certain point after which you cease to increase in benefit but stay "in practice". Trained skills that you neglect to use remain at their floor in terms of effectiveness - it's just like riding a bike.

So that overcharge you get from Practice sounds pretty nice - the question is how to prevent that from being a grind. On the surface of it you'd figure players would want to always have that Practice overcharge so they'd grind it out for all their skills. We can address this by making Practice a finite resource - for instance only 1 weapon skill category can benefit from Practice at a time. Using another weapon category will begin to drain the Practice from the first. Switching back and forth between several weapons can be very tactically useful, but none of those weapons will benefit from the intense practice and focus that using a single one will. In combat, Practice functions like being "on a roll" with a particular ability. If you're pounding your way through an underground base with your shotgun, kicking down doors and blasting away you'll find hit that hot streak via Practice, and become able to pull off cooler moves as a result.

Since Weapons are likely to be used a lot more often than support skill categories like First Aid or SpecOps, and it would be pretty annoying to have to go stealthing around or repeatedly picking locks in order to outpace your weapon practice, we'll probably consider support skills as a separate animal from Weapon skills, and allow more than just one Practiced category amongst them.

So something like:

1 Practiced Weapon category max

2 Practiced Support categories max

This cues us to design the categories of Weapon and Support skills with that mechanic in mind. We want each category to represent a distinct playstyle - what is the player actually doing with those skills. For instance if you are the only medic in your group and you are patching up your comrades after each encounter, you're likely to appreciate the First Aid Practice bonus. If you are charging headlong into each encounter relying on your heavy armor to protect you, you gain Armor Practice everytime your Armor saves you, reinforcing that bull rush playstyle. By contrast if you are being sneaky and stealthing up behind your opponents and slitting their throats, you'll end up with SpecOps and Blade Practice that reinforces that playstyle.

More later.

Creating a 'realistic' combat system

I was just reading some tidbits about the upcoming MMORPG Fallen Earth, and in digging around to find out what the gameplay would be like I started thinking about the kind of combat system I would like to see in a post-apocalyptic MMO.

There are three major issues that I think a really great combat system needs to overcome, aside from the basic "is it fun?" concern.

First, the current crop of combat systems invariably dampen the RPG part of the MMORPG. I mean this in the sense that combat always seems to lean in the classic tank-healer-dps direction replete with health bars, crowd control and dumb monsters that don't realize they should be killing the one in the dress. Especially in an MMO with realistic physical depictions of the characters (as opposed to cartoony ones) it is rather strange to see characters run up to one another and trade blows until one of them falls over. This was a real immersion-breaker in Star Wars Galaxies, as the melee classes just ran up to people and stood there doing jump kicks in their faces. It was pretty dumb. On a similar note, the idea that you must shoot/stab/bludgeon a character repeatedly in order to diminish a pool of health points has always been a realism issue since basic pen and paper D&D. This flaw is exacerbated in modern day settings like Call of Cthulhu where it is possible for characters to have so many hit points that a gun is actually incapable of killing them in one shot. There must be a better way.

Second, trying to build a combat system capable of handling both classic tank-healer-dps AND PvP is rife with peril. Warcraft continually moves in the direction of making PvE and PvP combat almost two entirely different games, each with their own set of mechanics and rules (here I'm talking about things like resilience, and the talk of implementing special global rules for arenas, etc). A lot of this comes down to the tank-healer-dps and aggro systems, and is something that pen and paper combat systems don't necessarily have a problem with.

Lastly, fantasy games like World of Warcraft are so dominant that even supposedly realistic settings (like post-apocalytic Fallen Earth) feel the need to incorporate some kind of combat balance between different styles. Star Wars Galaxies fell prey to this again as each profession needed to be able to create similar classes of effects for the sake of balance, because combat was balanced as a separate entity. A rocket launcher user had to be balanced with a pistol user or a swordsman, which is pretty silly since a rocket launcher is a bit more devastating. In a fantasy genre I have no problem believing that a Mage's frost bolt does X amount of damage in order to be balanced with a Warrior's sword - since a frost bolt is a completely made up attack. But now games feel that because you have a Gunslinger class that uses pistols and a Commando class that uses an Gatling gun, in order to have balance they have to end up doing about the same damage with about the same effects.

So I've been thinking about ways to address those peeves with a modern-esque combat system devised from the ground up that generally works on its own but is also meant to be built into a larger rules set in order to balance that last point about firepower. In theory this could be used for a pen and paper type system as well as an online MMO.

First Principles: Stats + Behavior + Conditions = Success

One cornerstone of this system is that it will be based on both character statistics, the behavior (or 'player skill') of the player and environmental conditions within the game, which may be static or altered by stats or behavior. So let's start with "How do you kill something?"

Kill Shots

To avoid having out and out health bars, we have to say that virtually any attack has a chance of 'killing' the target. That chance could be very good or very bad, but it's still a chance. A mechanic like this is similar to the d20 rule that a 1 is always a miss and a 20 is always a hit, regardless of modifiers. But in order to press towards realism, we are saying a 20 is always a kill shot. So in essence, a combat ends when one of the parties involved scores a kill shot on the other.

Variables

That being the case we need a way to determine whether a given shot is a kill shot. I don't want it to be entirely random, and so I'm introducing a mechanic that creates a variable chance for a kill shot depending on a number of factors. The attacker's variables are pretty straightforward: character skill for the attack attempted and player skill in aiming the attack. The target's variables also will be a factor though: The target character's agility and armor will of course dimish the chance of a kill shot, as well as whether the player is moving around at the time of the attack. So far, it's pretty much the same as making a d20 attack roll, with a little player skill added in. Conditions that may also modify chance of a kill shot can include cover and lighting.

Additional Defensive variable: Awareness

Something I want to add in though as a defensive modifier is what I'll call Awareness. If the target is 'aware' of the attacker, additional defensive modifiers will apply. In the game this will really come down to whether the target is looking at the attacker. So the triumverate for an opponent with a good defense will be Agility (both character stat and motion), Armor and Awareness - AAA. Likewise the easiest target will be a stationary, unarmored and oblivious one.

Additional Offensive variable: Concentration

But I don't want a target with great AAA to be unkillable - that's not realistic either, and it's a potential strength of the old hit point system - the ability to wear down a target. But traditional HPs have too many flaws in terms of realism, and I'm determined to keep the kill shot mechanic in place. So instead I'm introducing Concentration Points (CPs). Concentration Points add positive modifiers to attack rolls, increasing the chance of a kill shot. If you make an attack on a target that meets certain conditions but not a kill shot, you gain a Concentration Point on the target, increasing your chances of a kill shot on the next attack. In this way we have something akin to hit points in that a battle inevitably moves towards a kill shot, but we don't have to pretend that it took 100 bullets to the head for the target to die - those bullets whizzed past their head instead, but still took a karmic toll.

What is a CP-earning shot? It should probably be defined as a shot that would have been a kill shot if raw defensive stats were not applied. That way any defensive actions taken by the target (such as maintaining Awareness, using Cover and staying mobile) are still valuable because they can deny CPs to the attacker. If the only thing that saved the target were stats (including armor) then a CP is granted. That way CP accumulation gradually eliminates the benefits of raw defensive stats, but defensive actions by the target is key to maintaining a tactical advantage.

Another side effect of Concentration Points is that they are personal, rather than public. So Concentration Points one attacker earns on a target do not apply for another attacker - they have their own variables and CPs that modifiy their attacks. CPs may decay after a certain amount of time, or may not decay until combat is over - still not sure about that one.

This means that focus fire is not more effective than distributed fire. That's kind of awesome.

So let's take a break for a moment and imagine a 1 on 1 combat using these rules. Two opponents facing off would be trying to maximize their chance of getting a kill shot on the other, while minimizing their chance of receiving a kill shot. Whoever gets the kill shot first is obviously the winner. Initially it may actually be very difficult for either fighter to score a kill shot - due to high AAA defensive modifiers or low offensive modifiers. However as the combat progresses and CPs are racked up through well-aimed but non-kill shot attacks, the chances of a kill shot begin to increase. Eventually one fighter scores a kill shot and wins the battle.

In a 3 on 3 battle, there is not a distinct advantage for one team to focus fire on a single opponent (this IS an advantage under a traditional hit point system) because all it takes to kill the opponent is a kill shot, and the CPs accrued on the dead target will cease to be of use when they move on to try to take down the next target. It would be better for each teammate to focus on a separate enemy in order to eliminate them as quickly as possible.

This alone is a feature of the system that I think is pretty great. It would result in a real break from the tank-healer-dps model and allow for a much more realistic playing out of group on group battles. Healing and crowd control-type abilities could still be useful and fun - smoke grenades for instance to create cover - but combat does not hinge on the classic triumverate, instead we've created something new!

In another example, say a heavily armored, shotgun-wielding Fighter is confronted with a machine gun nest. The Gunner in the nest benefits from cover that will make it hard for the armored fighter to score CPs or a kill shot, so the armored Fighter decides to charge the nest! The Gunner fires away at the charging Fighter, and although the armor repels bullets and the Fighter dodges and weaves evading gunfire, the Gunner is racking up CPs. Will the Fighter make it over the nest embankment to get a clear shot at the Gunner before the Gunner scores a kill shot? His armor and agility just need to hold out long enough to make it there! So thanks to the CP and kill shot mechanic, raw stats can buy you time in a fight but they can't make you invincible.

The unresolved issue is that of balance - how do we deal with Rocket Launchers and Pistols and Bare Fists? My solution to that will come in the next post, but the quick answer: Don't bring a knife to a gunfight.

MMO Teamwork Solutions

A brief interlude to discuss a gaming dilemma and outline some possible solutions.

The Situation:

An MMO guild partaking in endgame raiding has a bad week. Formerly defeated bosses wipe the raid despite almost all the same participants receiving the same instruction. Leadership percieves a lack of discipline and general paying attention-ness. Experienced raiders in particular are guilty of this and appear to be coasting, while inexperienced raiders are generally just lacking in knowledge. Raid Leaders report several instances of players who should know better making lazy mistakes and ignoring instruction.

Analysis:

It's frustrating for sure. The previous tier of content was downright easy and the guild blew through it. The 10-person version of this same content is much easier as well. So on the occassions when the guild confronts this top-tier 25-person content, they are slacking because everything before was easy. Particularly frustating is not the failings of newer or inexperienced players, but the disregard by the most experienced players of the mechanics of the fight and the instructions of the raid leader.

Solution:

I like to implement systems to create solutions. So the first thing I thought of when I heard this was that some additional structure was needed to shape things up. Like in any endeavor, skilled and experienced people 'coasting' indicates that they aren't being challenged. Even though the team as a whole is failing, the highly skilled individuals are not percieving that as personal failure. This is in part because they are over-exposing themselves to successful 10-person raiding, and while there is some control that Raid Leaders have over that, trying to starve them of raiding success in the 10-person version probably won't solve their engagement issues in the moment.

Instead, targeting the unchallenged players with additional responsibilities in the moment should drive them to focus instead of coasting. My suggestion is that Raid Leaders do the following:

On fights with specific mechanics that all players have to understand, charge a problem experienced player with explaining the mechanic to the raid.

In general, make problem experienced players responsible for mentoring less-experienced players. A great situation is when you have one or two new players filling a similar role to a more experienced problem player, tell the experienced player that you won't be slowing down to explain things and that they are responsible for making sure the less experienced player knows everything they need to know to succeed.

These practices are generally useful in situations where a single player is not living up to their potential. Apathy is one of the worst enemies of any group endeavor and yelling and chastising will not necessarily create the desired motivation, as especially in the digital world it is very easy for someone to just close a chat window or click over to another program to avoid a virtual berating. Especially in a gaming situation where negative consequences are very difficult to enforce (no one's job is on the line for instance) being the boss means having to use even more carefully considered methods of enforcing discipline and motivating.

There is a lot of distance between the feudal notion of "Do what I say or die" and the gaming notion of "Do what I say because...it's fun". Although the recent downturn has probably instilled a little fear into the young worker about job security, there is still a very strong sense of "If I don't like it, I'll just go somewhere else" in us. Increasingly rare is the idea that as workers we should have to pay our dues and put up with whatever our bosses throw at us because our survival and success depends on it. In one sense that can be seen as apathy or entitlement. But it's also social evolution - the greater social and economic freedom that we've won as a society is changing the basic rights we demand in any social or economic contract.

Older folks might look at that situation and forsee the end of society as everyone becomes to apathetic to work, or the failure of the machine because nobody wants to be a cog. But for the first time we have a very interesting playground (in the form of online social gaming) to experiment with these sorts of social situations and learn what works and what doesn't. I think a generation of managers raised on these sorts of gaming situations may come at real-world management problems from a different angle that is better suited to the changes that previous generations can only lament.

When is a game not a game?

In the past few days I have been playing a few of what you might call "indie" games. Lot of Dwarf Fortress (which deserves its own post frankly) and also Polaris, a Half-Life 2 mod that is partly a game but mostly a story - anyway I won't spoil it just check it out.

What the designer of Polaris is working to achieve is very interesting, and if you read up on it you'll see references to other 'games' that are more works of interactive fiction than straight-up gaming experiences. And that is what interests me most - that point at which what we generally classify as a "game" (because we play it on a computer or gaming system and it uses many of the mechanics of a game) but that really isn't a game at all.

In Polaris there are 3 different endings possible based on what you do, but none of them are really wins necessarily. How you react to the stargazing date that forms the plot of the scene determines the outcome, but what you walk away with after playing it is more a sense that you experienced a compelling story than that you 'beat a game'.

How often have I heard someone say "Oh yeah, I beat that" when discussing a game? But this brand of game we're talking about here isn't about beating it - it's about experiencing it, and in that sense they are much closer to a short story or novel, or television show or film than to a game.

A looong time ago I played this game that I wish for the life of me I could remember the name of. It basically had you in the shoes of a soldier in a future sci-fi world where you were supposedly part of a interplanetary or interdimensional research and resource mining team that was operating in a hostile environment. Most of the game you were walking around inside this underground bunker and never saw the outside. When the soldiers went outside they wore virtual reality helmets that depicted what they were seeing as atari-like graphics - so they never really saw the outside. The environment was supposed to be toxic and would result in painful death if the suit was breached so they were equipped with euthenasia systems in case of a breach. Meaning none of the soldiers who saw the outside lived to tell about it. The gameplay basically consisted of walking around and watching scenes play out as you talk to other characters, and you start to get the idea that something fishy is going on. The only actual game to speak of was a final scene in which you are sent out to fight off an assault by the strange bug-creatures that are the 'hostiles'. I believe there was no way to win that battle though, because you are quickly overwhelmed and mortally wounded. Your character hears the euthenasia system activating and in his last moments removes his helmet to see his enemy for the first time. Instead of a bug-like alien creature it is a tall beautiful woman with angel-like wings and a flaming sword. At the time I thought that it was the dumbest game ever. In retrospect I'm kind of amazed that so much of the story stuck with me. Now I'm wondering: was it inter-dimensional or interplanetary? Was the idea that the military were invading heaven? Or just another planet with winged women? Weird.

But those sorts of games have got me thinking about the next evolution in fiction. Consider the leap between stage plays and film. One can write a story for each medium, but when you write for film you have to consider the camera and more importantly the use of editing to manipulate time. A play can have successive scenes that take place at different points in time, but that's rare and unwieldy. A film does that all the time. What would Shakespeare have made of Pulp Fiction?

I think that the next leap in storytelling is going to be to Pulp Fiction what Pulp Fiction was to Shakespeare. Polaris hints at it - to fully experience it you play it three times and get three different stories, based on what you choose to do. Writers of the next generation of stories will be writing with interactivity and the choices of the audience in mind.

Bioware's writing system is along these lines: they want choices for their players and their writers create dialogue with choices in mind. Although in the end the stories of games still play out no matter which choices you make, the interactivity draws the player in in a way that isn't possible with passive entertainment.

There have already been some incredible stories told through games. I'll never forget my experience of playing Planescape: Torment. But currently those experiences are catalogued under the heading "games" and so don't really enter the mainstream.

Imagine however, a time when there is a Law and Order of interactive fiction. One that everyone knows and plays weekly, and talks about around the water cooler. An experience that isn't passive, where you participate in the solving of a crime and the telling of a dramatic story each week as a shared experience. Imagine when each summer studios release their slate of scenarios for the public with the star power of Hollywood and the appeal of blockbusters but they aren't movies - they're 'games'. And people like them not because they are gamers, but because they allow them to experience a story more fully than simply watching it unfold on a screen.

There are a lot of possibilities once you go down that route. And I find it very interesting that the games that really point the way on this are not the big hits of today - not Mario or Call of Duty but strange little games like Polaris and forgotten pieces from 15 years ago.

Quake Live - More Browser Access to Online Gaming

Following a similar "all you need is a browser" scheme as OnLive, Quake Live is the next iteration of the 10 year old Quake franchise.

Pretty much following the same principle, except that it's free (paid for by advertising), Quake Live allows you to log in to a Quake online match and play against other people with only an internet browser.  A clever skill-matching system is built onto the concept along with the same sorts of peripherals that OnLive is offering - community, messaging, etc.

This trend says to me that our computers are likely to become merely portals to far more massive computing power existing in an offsite location, and that game developers will be selling access and advertising more than boxes in the near future.

p.s. It seems Christoph Hartmann, founder and president of 2K Games agrees!

Making Your Console Disappear

Say goodbye to your console.  At least that's what OnLive would have you do.

Riffing off of the Steam model and taking it a step further, OnLive (set for a "Winter 2009" release) is an online gaming service accessible via PC, Mac or "microconsole" - a piece of hardware you connect to your HDTV - and a broadband hookup that purports to provide access to any game title instantly.  Having partnered with almost all of the biggest game studios (Activision/Blizzard is a notable exception so far) OnLive aims to be a less hardware dependent way of delivering gaming content by essentially "streaming" the games to your computer or television.

Obviously the big advantage here is that there are no trips to a brick and mortar store to buy games, but OnLive looks to be building in a few other handy elements.

A Demo feature allows you try out a game before you buy it.

An Observe feature lets you watch other people play (another way of demoing a game).

A recording feature lets you record your gameplay for display later (they call them "Brag Clips").

There are some other community elements, but the big allure here is the instant nature of the whole thing.  Because the games are server-side, you don't even need to download anything to play them - your controller is just communicating with a big powerful centrally located gaming server via the internet so all you really have on your end is a display.

It occurs to me that this model is also highly resistant to cheating - since all the calculations are happening server-side.

If it really works as advertised I don't see how this couldn't be a huge success.  The ability to easily demo games before buying them is great on its own and since you aren't physically buying (or even downloading) the games they HAVE to be priced lower than in-store and potentially even lower than Steam-esque discounts.

Overall OnLive is obviously something to keep watching.  They've clearly already gained the trust of the biggest studios and if they can get it to work as smoothly as advertised they'll be putting some serious competition on the console market (the Wii being an exception because the hardware is generally part of the gameplay).

Can they still call it an RTS?

No, in fact they'll call it a "meta-time strategy game".

Achron, an RTS (if you must) game being produced by Hazardous Software (founded by the awesomely-named Christopher Hazard), fulfills every nerd's dream of being able to muck around with time travel as a weapon of war.  It was unveiled just a couple weeks ago at the Game Developer's Conference in San Francisco, and it's the kind of game that immediately sparks conversations beginning with "But what if you ..."

Built onto a simple (so far) RTS combat engine is the ability to move forward and backward in the timeline of the overall battle to change the past in order to outwit your opponent.  Didn't like the way an attack turned out?  Go back in time to before you issued the attack orders and change them!

The stroke of genius about what's been revealed so far is how well the time travel plays into your strategic thinking.  Like any good sci-fi concept, they took the "what if" of how time travel could be used militarily and appear to be incorporating it into the traditional RTS thinking of terrain, unit and resource management.

This game will be one to watch as it develops, not only because it already seems like the gameplay will be mind-bendingly intriguing, but also because as a proof-of-concept it shows how time-travel may be able to work in a multiplayer setting.