Time progression in RPGs have always been a dicey matter. In a game where you get the freedom to explore the world map at your own pace, the passage of time has to be handled in a proper way in order to preserve a sense of realism (and to a certain extent, immersion). It’s foolish to have the player wander around the world for half a century of game time when the game’s premise is about a big bad evil wrecking havoc and destruction upon the peaceful kingdoms (aka the premise of 99.9% CRPGs).

Take Oblivion for an example. Demon gates have opened around the world and you, as the protagonist, have to close them all before the world is overrun by demonic minions.

Right…

The problem is you, the heroic protagonist, can skip around and collect flowers for practically your entire game life without even touching one demon gate and the world would still be the same, patiently awaiting for your salvation. In my playthrough that salvation never came – I never did closed any of the gates, but THAT’S OKAY! Because the world will wait for you.

If only real life was like that!

No, Oblivion sucked at handling time. It was immersion breaking, and even screws with its own premise. At least Morrowind’s made more sense. Now let’s take a look at the other side of the spectrum: Fallout. Fallout had a time limit to find the waterchip, but latter patches extended the time to a point where it became insignificant. The reason was that the world was so fun to explore that the developers didn’t wanted to impose a time limit that would certainly restrict exploration.

Come to think of it, no game ever imposes a time limit on the player just so that the player can play the game at his own pace. And in this blog post, I am going to explain why this would not be suitable for Splintered Core.

Why a Time Limit?

While we are building Splintered Core as a one-off game, if it ever sells well we would be able to continue with the second game and eventually (with our fingers crossed) a complete trilogy, which was the original plan. We started out designing the game from a trilogy’s standpoint, with a plot that spans three titles, but structured it in a way that there are NO cliffhanger endings in any titles. Each title has its own story to tell, it’s own theme to explore, and that’s how we’re approaching the plot design.

However, this does impose a design limitation for us. Firstly, we wanted the capability to transfer a completed save game from a previous title to the next title. The reason was because we wanted to attempt to transfer all the player’s decisions and consequences in the first game into the second, and then into the third, creating a highly customized playthrough of the trilogy. The problem here is that what if the player spent 30 years wandering aimlessly in the first game? Would he start off an old man in the second game? Can he then spend another 30 years wandering the fields?

Say for example, the player DID spent 30 years of game time in the first game. This would mean that his greatest investment into the first game will botch his experience in the second game, if we ever implement an aging feature. If we *dont’t* implement an aging feature, you’d have characters at the age of 200 in the third game, which requires suspension of disbelief of epic proportions to maintain immersion.

Catch-22 situation eh?

How is it implemented?

So we decided to impose a game limit. First we calculate the total time to walk from north to south, the longest journey in the game. The distance is about 350 miles, and on average the player can only travel at a speed of 3 miles per hour for a maximum of 15 hours a day. This means that it will take approximately 7 to 8 days to travel that distance.

Therefore, over a span of 5 years, if the player kept the character on the move, he would have been able to traverse the north-south route 234 times. The speed of travel increases significantly if the player ever manages to buy a vehicle, or tag along on mechanized merchant caravans. So we can safely say that 5 years of in-game time would be a good indication of how much time a character would spend in a normal playthrough.

Now we increase that by 2 fold, to a maximum of 10 years, and set that as our time limit. But instead of making it a hard time limit (e.g. game ends when time is up), we instead choose to implement a countdown timer called the resolve meter, which is directly tied to the game’s plot.

Recall that the game’s premise is finding the legendary city of Gauntlet, in hopes of settling down there, away from all the chaos and lawlessness in Oasis. Spending 10 years on this task wouldn’t seem to be out of place. But like all far-fetched quests, taking into account that the protagonist is not even sure that Gauntlet exists at all, it won’t be hard to imagine one’s will of finding that place slipping over time.

So here’s how it would work: your resolve meter counts down slowly, initially giving you a maximum of 5 game years before it runs out and you’re forced to choose to settle down in one of the towns, with the ending explaining your fate in that town according to your reputation there (choices you make in the game will affect the ending in that town). However, if in your travels you piece together information about Gauntlet, like finding Lorekeep at the end of the first arch and securing a way to Severim in the second arch, will recharge your resolve meter, thus effectively giving you extra game time.

The afore mentioned events will completely recharge your resolve meter, while smaller events like finding an person who left the Combine army or finding Gauntlet’s satellite images would recharge the resolve meter at a smaller scale. Therefore, it is not impossible to stretch the first game’s time limit to a total of 15 years.

Conclusion

So this means we’re going to try something that not many CRPG designers have even dared entertain, for all sorts of reasons. The time limit would stay, and to spice it up we decided to allow the game world to change over time, tied to the player’s actions. The key here is to balance freedom of exploration with a time passage system, and this is something we intend to experiment in our first game.

Following the previous post about striking combat, in this post I’ll lay out the framework for grappling combat.

Grappling

There’s some fundamental difference between striking and grappling in real life. Striking requires good judgement and control of distance, where each technique requires a certain minimum distance of travel to be able to impart an effective amount of kinetic energy upon impact. In grappling however, the key success factor is to get in close, disable the opponent’s range of responses, and then maneuver into a dominating position where one can deliver attacks without fear of retribution.

This is what I’ll be trying to model for grappling combat in Splintered Core. There are 3 phases: (1) closing the distance; (2) takedown; and (3) follow-up technique. Maneuver modes and stances determine the techniques used in the first phase, attack stances and

Maneuver Modes

1. Walk
2. Run
3. Charge

Like both ranged combat and striking, there’s not much changes to maneuvering modes. Like striking, walking would default to sneak mode; and both run and charge can have engagement mode turned on at the same time. Keybindings don’t change (they are “Q” for engagement mode toggle and “W” for maneuver mode cycle”).

Stances

1. Offensive
2. Defensive
3. Counter

Like striking, character shifts between different preparation mentality with different stances. If a stance is set to offensive, the grappler will intend to shoot the opponent (shoot as in shoot wrestling) and forcefully perform a takedown; in the counter mindset however, the grappler will try to make the opponent over-commit and then use that opportunity to counter-attack.

Weapon Positions

1. High
2. Normal
3. Low

Again, like striking, this sets the character to a specific attack vector. For example, in an offensive stance and a high weapon position, when the character engages an enemy he will attempt to takedown the opponent from a standing position. If however the weapon position is set to normal, the character will instead perform a shoot towards the opponent’s waist. Finally, if the weapon position is set to low, a tackle to the legs would be executed instead.

Attack Stances

1. Close Range
2. Normal Range
3. Shooting Range

Like striking, attack stances are basically engagement range. This is to determine from what range would a character execute the grappling techniques. The general rule-of-thumb is that the closer the distance, the higher chances of successfully grappling the opponent but at the same time the higher the chances that the opponent can resist (or even reverse) the takedown.

Attack Mode

1. Joints
2. Limbs
3. Circulatory System

Again, like striking, attack mode mainly determines the attack type. For example, on a high weapon position, close range, and with the attack mode: “joints”, the grappler would close the distance and attempt to perform a neck snap maneuver. On the same position and range but with attack mode at “limbs”, the grappler would instead execute a arm bar or arm lock maneuver. Each of the attack mode will provide different status ailments if executed successfully: joints mode will cause sustained pain and temporary loss of mobility; limbs mode will cause paralysis or disability; while circulatory system will cause blackouts and even death.

Primitive Weapons

Primitive weapons, which comes in two categories – light and heavy, defines a character’s skills in using melee weapons like batons, sledgehammers, knives, axes, maces and so on. To break it down:

Light Weapons

Knives, Batons, Switchblades, Knuckles, Maces, Machetes and Hammers. They are generally fast but do not have a long reach.

Heavy Weapons

Shovels, Staffs, Spears, Sledgehammers, Axes, and Flails. They generally have a longer reach and more damage but are slow to wield.

The only difference in combat model for primitive weapons is the Attack Stances and Attack Modes:

Attack Stances

1. Short Range
2. Medium Range
3. Extended Range

This is pretty much self explanatory. The shorter the range, the easier it is to hit an enemy, but the lesser the damage.

Attack Modes:

1. Flurry Attack
2. Normal Attack
3. Power Attack

Mainly this is the speed of the attack used, or the level of commitment. The faster it is, the more attacks per round at the cost of damage, while the slower it is the more damage it deals.

Summary

I’ve pretty much covered all possible aspects of the combat model, and I’m quite happy at this model. While things might change in the future (as always), I don’t think it’ll stray too far apart from this.

One concern is that it might be too complex. Which is true. It IS a little too much to manage all these parameters in the heat of combat, especially in a tactical-heavy environment like what I envisioned. So in my next post I’ll talk about streamlining the game mechanics into pre-packaged techniques.

Until next time!

Continuing from the previous post about ranged combat, I’ll be splitting melee combat into two distinct sections: Unarmed (Hand-to-hand) and Melee Weapons (Primitive Weapons).

To recap from the STEN Character System, there are two distinct schools in melee: striking and grappling, both of which attempts to accomplish vastly different goals. But before we begin, I should take some time to explain the reason of expanded melee combat in a modern setting with firearms.

Conventional wisdom would question the effectiveness of melee combat in a setting where firearms are lethal. How do you play a martial artist when a bullet or two could seriously cripple your character? The answer lies in the general economics of the setting.

Good quality firearms are rare, and match ammunitions are even rarer. I’ve made it a point to have make decent military grade firearms extremely rare (the current estimate is about 100 of them of various make scattered in the game world, all owned by important NPCs). This way, players have to rely on sub-par firearms in ranged combat, if they choose to specialize in it. They are equally deadly, but leaves much to be desired in terms of accuracy and reliability. Weapon jams will not be an uncommon thing in combat.

This is not done to open avenues for martial artists to shine. Rather, this design decision was made during the world creation process, where we decided to go for a low-economy post-apocalyptic world where high-tech stuff are extremely precious just because they are rare.

To add to the woes of a ranged character, every gun has a minimum range. If an opponent gets too close, you suffer massive penalties to your accuracy. This is where a melee character can shine: melee characters need to focus on stealth and speed, while ranged characters need to focus on reaction and perception.

Now having said that, let’s rip apart the mechanics of Hand-to-hand Combat.

Striking

For those not familiar with martial art jargons, the term striking arts refer to techniques that have you on your feet, at a comfortable distance, using your elbow, forehead, fists, knees, legs, etc to land hits on your opponent. The goal is to cause enough damage through striking impact.

Taking a cue from our unified combat model discussed earlier, we have 5 major categories of actions: Maneuver Modes, Stances, Weapon Positions, Attack Modes, and Attack Stances. The same applies to striking too.

Maneuver Modes

1. Walk
2. Run
3. Charge

There’s some difference between maneuver modes between ranged and melee mode. In melee, walk always triggers sneak mode. In addition, the engagement option can be turned on for charge. The same key binding of “W” and “Q” applies here too.

Stances

1. Offensive
2. Defensive
3. Counter

Stances are completely different from ranged combat. Here, offensive stance prepares a character to launch attacks (increase in accuracy and speed at the cost of defense and evasion), defensive stance prepares a character to parry or block attacks (increase in defense and evasion at the cost of accuracy and speed), while counter stance prepares a character to launch a counter attack following a successful parry (increase in evasion and speed at the cost of accuracy and defense).

Like ranged combat, the key SHIFT works the same here too, shifting from offensive to defensive to counter.

Weapon Position

1. High
2. Normal
3. Low

It should be called preparation vector rather than weapon position, because the effect depends on the stance. If a high position is set on an offensive stance, the character will launch a high attack (kick to the head, uppercut, etc) which has high damage but slow in execution. High position on a defensive stance however, allows the character to focus his guard on his head, increasing his chances of stopping attacks launched towards him in that direction/vector. Finally, high position on a counter stance works the same as a high defense, except that a high attack is thrown right after parrying.

Like ranged combat, CTRL + mouse move cycles through the three stances.

Attack Stance

1. Clinching Range
2. Short Range
3. Long Range

Like weapon position, it should be called attack range instead, because this determines what kind of attack (or defense against) the character will throw. Attacks in the clinching range involve headbutts, knees, elbows, etc; attacks in the short range involve punches, low kicks, etc; while attacks in the long range involve roundhouse kicks, side kicks, flying punches, and so on. The shorter the range the faster the execution, higher the accuracy, but lesser damage and knock out capability.

Like ranged combat, the button “A” cycles through the three modes.

Attack Mode

1. Body
2. Arms
3. Legs

Unlike ranged combat, attack mode determines the exact body part used in the striking, defending or counter technique. Body attacks include headbutts, shoulder bashes, and so on. Arms include elbows, fists, palms and backhands, while legs include knees, heel strikes, kicks and front kicks.

Like ranged combat too, the button “S” cycles through the three modes.

Summary

So as you can see, the UI remains the same for striking. It’s just the definitions that have changed. In my next post I will talk about Grappling.

A while ago I briefly dipped into ranged combat and over time the concept kind of matured after long hours of discussions and debates. Right now our combat system is more unified, so I’ll take some time to talk about how ranged combat works in Splintered Core.

Tic Timeline

In the past, melee and ranged pretty much operated differently, making it difficult to memorize the details of each without going back to long winded design documents. So we decided to turn it into something more intuitive and easy to understand.

Firstly, we have ditched turn-based combat for tic-based combat. In tic-based combat, a round is defined in a number of tics. Let’s say for now 1 tic = 100 milliseconds. This means that there are 10 tics in a second, and assuming a round is 10 seconds long, you’d have 100 tics per round.

Each action costs a number of tics. Readying a weapon for instance takes 70 tics, which is equivalent to 0.7 seconds. Moving from one tile to another in run mode takes 50 tics per tile, translating to 0.5 seconds to traverse a tile, etc.

Tics are represented as a segmented bar on the top of the window, looking a little like this:

Splintered Core UI prototype

Notice the white dots along the bar. Those are committed actions for the currently selected character. The first dot might be a command to pull out his rifle; the second dot is where he readies the weapon; the third dot is where he runs towards a cover; the fourth dot is where he goes into a kneeling position; the fifth is where he readies a claymore; the sixth is where he plants it on the ground; the seventh is where he sneaks to another cover and the eight is where he readies the rifle again.

Each dot corresponds to an action, and as every action costs a certain amount of tics, the distance between the dots on the timeline varies too. How does the player set this sequence of events up? Well first he clicks on a character to select it; click on a rifle to ready it; press “W” twice to cycle from the walk action to the run action and then to the charge action, then right-click on the map to note the destination; press “SHIFT” and move the mouse down to select the kneeling position; click on a claymore to ready the weapon; right-click on the ground to plant the weapon; right-click on the map to move to a new location; and finally click on the rifle again to ready it.

This is call a commit chain, which is basically the planned tactical movement for that particular character. If a commit chain successfully completes and fulfills a certain criteria (e.g. successfully score hits on the enemy), the character’s rush meter fills up, reducing the tic-cost for his actions. Note that the player can click on the dots and drag them to the right to lengthen the duration of the action, which is similar to the action of having a character wait in between actions. This can be used to coordinate maneuvers between party members.

Once the player completes the commitment, he presses the “GO” button and the game is unpaused. A time marker begins to move from left to right, and all characters in the world performs the actions according to the timeline. Once the marker reaches the end, a new round begins.

The commit chain can be broken if the character is intercepted successfully. This usually happens when a character runs into another’s field-of-fire or attack range, and the other character rolls an intercept check successfully (whether the character is hit or not doesn’t matter – as long as the intercept check is successful, the commit chain is broken). This is called a stumble, which causes the rest of the planned action to be discarded.

When a stumble happens for any player-controlled character, the time marker is stopped, and the player is given the chance for course correction. Note that during this intercept period, only the intercepted character can be chosen and ordered around.

Let’s say the character suffers a stumble when planting the claymore. So happens that when he reached the first destination, and halfway through the act of planting the claymore an enemy steps into view and snapshoots at him, scoring no hits but successfully breaking the chain of action. All further actions are dropped and the game is paused, while control returns to the player, with the intercepted character selected. Now the player can click on the rifle to ready it, and then right-click on the enemy to shoot back.

Note that the player can press the SPACEBAR key anytime when the game is unpaused to pause it, and from there the player can change the commitments of each party character at the cost of breaking the command chain and abandoning previously earned rush points for that round.

As as you can see, the combat model in Splintered Core is pretty much a paused real-time system, with much emphasis placed on streamlining the controls to make it easier to plan tactical maneuvers.

Maneuver Modes

There are 3 kinds of movements that a character can perform, cycled using the “W” key. They are:

1. Walk
2. Run
3. Charge

On each mode, there’s an engagement option to fire at targets of opportunities along the way. To toggle this, press the “Q” button. If engagement is turned on, the character will open fire along the way if an enemy presents itself as a target of opportunity. This is disabled in the charge mode however.

The difference between the three options is speed, and resistance against stumbles. Walking is the slowest and the most dangerous, as you get virtually zero defenses against intercepts. The advantage however is that in engagement mode you don’t get penalized in accuracy as much as the other maneuver modes.

Run is the middle-ground, where you suffer a moderate amount of accuracy penalty to reach your destination faster and gain a little resistance against intercept, while charge disables engagement mode but gets you to your destination in the shortest time, and has the highest resistance against intercepts.

Stances

In ranged combat, stances are extremely important. There are 3 stances available:

1. Standing
2. Kneeling
3. Prone

The higher the stance the faster one can maneuver as well as the farther one can see, at the cost of a larger target profile. Note that when in prone, the charge maneuver mode is disabled. Pressing and holding down the SHIFT key while moving the mouse up and down changes the stance.

Weapon Position

Weapon position refers to how the weapon is used. There are 3 positions available for most ranged weapons:

1. Free
2. Hip
3. Shoulder

Having the weapon at hip level allows a fast target acquisition rate, at the cost of lousy accuracy. This is good when you want to fire on the move with the intention of suppressing the enemy. For instance, you have a M240 equipped, and want to pin down an enemy behind a cover so that you can move safely from one location to another. Firing from the shoulder gives a larger accuracy boost, at the cost of slower target acquisition in the sense that your character takes a slightly longer time to shift fire from one target to another. Free position on the other hand, offers the fastest target acquisition at the cost of horrific accuracy. One thing to note about the free position is that you can shoot around corners or cover using this position.

Pressing and holding CTRL while moving the mouse up and down changes the weapon position.

Attack Mode

Attack mode determines the kind of attack that you will perform with a weapon, and comes in three flavors for ranged weapons:

1. Snap Shot
2. Aimed Shot
3. Suppressive Fire

Snap shots are less accurate, but has the shortest target acquisition rate. Aimed shot take longer to acquire a target, but has a significant accuracy bonus. Suppressive fire has zero target acquisition rate because you will be firing at a location regardless of the existence of targets or not, but suffers from having the worst accuracy. Pressing “A” cycles through the modes.

In each mode, if the character is carrying an automatic weapon, he can cycle through three fire modes:

1. Single-shot
2. Short Burst
3. Long Burst

Single-shots does not incur an accuracy penalty, but provides a low volume of fire too. Long bursts offers a high volume of fire (more bullets flying down the range) at the cost of a high accuracy penalty. Short bursts is the in-between. Pressing “S” cycles through the fire modes.

Summary

So in summary, this is how the toolbar might look like:

Toolbar UI layout

I think this would be a great framework to work from. The nice thing about this model is that you can apply the same things for melee combat as well, which I will cover in detail in a future post. In melee you too have maneuver modes, stances, weapon positions (high, medium and low), attack stances (fast, normal and cautious), and attack types (parry, riposte, all-out). This is what I meant by a unified combat model.

Phew, this is a long post. I hope I haven’t been too long winded lately!

Character creation is half the fun in playing an RPG. At least, that is for me. Before we begin, let me do a quick rundown of some of my favorite character creations.

Darklands

Darklands Character Generation Screen

I have not played this game, but watching the let’s play videos on youtube I fell in love with the character generation sequence. Here’s how it works: First you choose a background, which modifies your base stats, and then distribute your stat points. After that, you choose a profession, which further modifies your stats and skill points, and then you get to distribute those points into your skills. This is interesting because it wouldn’t be hard to imagine players spending hours just min/maxing the values. I’m not sure how the background affects the game however. It’d be nice if there are specific plot branches that caters to different backgrounds or professions.

Darksun: Shattered Lands

Darksun Character Generation

While the character creation sequence is pretty typical, where you choose a race, sex, and a few other attributes, it stood out at that time due to the wide variety of interesting races and unique classes to choose from. Gladiators? Psionics? Elemental Clerics? Half-giants? Thri-Kreens?

The system itself is not outstanding, but the theme itself was.

Fallout 1 & 2

Fallout character generation

The first two Fallouts had outstanding character creation mainly due to the SPECIAL system. Since your seven attributes do not improve naturally through the course of the game, you had to make hard decisions even before the game begins. It’s pretty hard to power-game the attributes. Well, that’s until Fallout 2 was released, and then Fallout 3 had to take it to the “next level” where anybody can become a master of every fucking thing.

But back to character creation. After setting your attributes, you get to choose traits. Most traits provided gave both benefits and penalties, which is a brilliant touch. Then you get to tag skills, which doubles their rate of improvement.

So many things to do, so many different kinds of characters to choose from!

Splintered Core

As we can see, character creation could be fun in various ways. Darklands was fun because the backgrounds and classes that you chose were not just thematic flavors but had a real impact to your actual stats. Darksun was fun because of the fresh races and classes introduced. And finally, Fallout was amazing because there was so much to do, and none of them confusing at all.

Time to take cues from past masterpieces. Taking into account that our plot revolves around one single character (like Fallout and Baldur’s Gate) and that said character has a surprising background which is gradually revealed to the player (like Baldur’s Gate and Planescape: Torment) through the plot, we couldn’t really use Darkland’s backgrounds and professions. And since our stat system has been streamlined to only feature stats where attributes and skills are grouped together in one single term, we had to think out of the box to inject more fun into the character creation process.

This is what we came up with:

First you enter the character’s bios. Name, sex, body, and portrait. The body selection are not just aesthetic options; they have an impact in the game world. Choose to be a fat chick and you won’t expect to charm those who likes skinny girls.

Then you are given a bunch of stat points to distribute amongst the 30 general and specialized stats. Here, you can tag a few of the stats to reduce their stat point cost, which effectively means it can be improved at a much faster rate compared to the rest.

Finally, you get to choose traits, which are divided into two categories: positive and negatives. For every positive trait you choose, you will need to choose a negative trait too to balance it out. Also, traits directly modify your stats. Positive traits will raise certain stats while negative traits will reduce them.

How about that? A combination of Darklands and Fallout.

The eventual goal in the game is to create towns with realistic economies. This means each town produces and consumes commodities, exports the surplus and imports the deficiencies. While the basic parameters are all streamlined and simplified, I believe the implementation of such an economic model would greatly contribute to the immersion and believability factor of the game’s environments.

Commodities & Consumption

First off, let’s talk about consumption and commodities. There are 10 commodities in the game, 5 of which are raw commodities and 5 are manufactured commodities. The difference between the two is that manufactured commodities are tied to two parameters while raw commodities are tied to only one parameter:

1. Medicine (Manufactured)
2. Munitions (Manufactured)
3. Clothes (Manufactured)
4. Alcohol (Manufactured)
5. Parts (Manufactured)
6. Food (Raw)
7. Energy (Raw)
8. Fabrics (Raw)
9. Materials (Raw)
10. Metals (Raw)

Depending on the needs of the respective town, the consumption rate for the commodities would fluctuate. Here are the factors governing consumption for each commodity:

Medicine
Population Health & Wealth

Medicine consumption depends on the population’s health and wealth. It is consumed at a small amount daily, and the consumption rate increases with the population’s general wealth. Health levels that dip below average would result in an exponential growth in medicine consumption rate.

Munitions
Town Culture & Security Level

Munitions like weapons and ammunition are first and foremost governed by the town’s culture. Security level plays a significant role too, and when it dips below average munitions consumption would grow exponentially.

Clothes
Population Count & Wealth

Clothes (including armor) are consumed daily, and the rate of consumption depends on population count and wealth. As with medicine, when population wealth goes above the average, consumption rate increases too at an exponential rate.

Alcohol
Town Culture & Population Count

Alcohol consumption is similar to food, albeit at a lower level. Town culture affects it too. For instance, Spike, which is controlled by the Saberions, have zero need for alcohol due to their religious belief against its consumption.

Parts
Town Technological Level & Population Count

Parts refer to spare parts for all kinds of things, including electrical devices, vehicles and machinery. The higher the town’s technological level, the higher the consumption of parts.

Food
Population Count

Food consumption increases at a linear rate with population count, although the rate can be affected by various factors such as weather, culture, events, and so on.

Energy
Town Technology Level

Energy can refer to fossil fuels or Hydroplasma, both of which are needed to power machinery, vehicles and devices.

Fabrics
Population Count

Fabrics are used for a variety of general purposes, and is used to make ropes, tapestries, clothing, backpacks, and so forth. They are also used to make tourniquets. It increases at an exponential rate, but is less affected by external factors.

Materials
Population Count

Materials are used to build and repair houses, warehouses, factories, containers and so forth. Their consumption rate grows at an exponential rate too.

Metals
Town Technology Level

Metal consumption increases at an exponential rate to the town’s technological level. It’s used in machineries, munitions, and in most other things in moderate amounts.

More to come in part 2!

I’ll have to clarify first – I’m a sucker for RPGs with turn-based combat. I started my gaming life back in the days of 80386 and DOS, and my first game was QBASIC Gorilla which was about the only game that brought my entire family together. Not that I had a broken family, but it was one of those games you played in the past not alone but with a bunch of people, and end up having fun despite the simplistic gameplay and basic graphics.

But that’s not the point of this entry. I went on from Gorilla to Gods (a platformer), to other DOS games too numerous to recount (One must fall, Traffic Department 2192, X-Wing series, Aces over Europe, Wolf3D, Prince of Persia, Another World, Dune 2, Doom, Doom 2, Quake, Duke Nukem 3D, etc), and then eventually found my true love in RPGs (although Starcraft, Quake 3 Arena and Half Life are the usual exceptions) starting with Darksun: Shattered Lands.

There are only 2 RPGs that stood out in my memory: Darksun and Fallout 2, for its combat system. While I played games like Might & Magic, Wizardry and so forth, for some reasons the combat system in Darksun and Fallout 2 actually made it FUN to get into fights.

Darksun starts out by throwing you into the arena and have you fight for your life. After every fight, if you so wished, you could continue fighting, and the game will throw in random monster parties for your entertainment (although in the game you’re supposed to be fighting for theirs). Years after playing that game the second game that gave me that impression was Fallout.

Why?

It’s kind of hard to sum it into words. In Darksun you controlled a party, and in the game you consistently get the chance to fight difficult monsters. I could still remember stumbling into Vrocks in an abandoned temple, and got my ass handed to me for not having +3 magical weapons. That was when I was winning every single fight, including those against Mountain Stalkers.

Games nowadays often turn you immortal in the late stages of the game, where your high level plus powerful loot allows you to steamroll over most enemies save for the toughest boss. Final Fantasy games are often guilty of these.

I believe that games should challenge the player. There is no challenge in fighting war of attritions against monsters which are hard to kill just by the merit of having shitloads of HP. It is far better to present players with battles containing genuine opportunities for actual tactical play. The final battle in Darksun is but one of the best examples in this case: You had to fight through 3 waves of enemies, and no matter how strong your characters were it was still a damned difficult fight. Had you waded into it unprepared (e.g. no allies, no genie, poor spell choices, etc), you would have been slaughtered by the second wave.

The final battle of Darksun enforced one thing – resource conservation. The tactical focus in that scenario was simple – how do you ration your ammunition so that you could pull through all three battles, each of which contains enough adversaries to warrant an all-out approach on your side?

Moving on to Fallout 2: I hated the trash mobs (GOD I FUCKING HATE THE TEMPLE OF TRIALS!), but at the same time the combat was fun. Yes, it was easily abused, but that somehow made it fun. Why? Heck, I don’t know, maybe I just enjoyed seeing torsos blown to bits by high-velocity 2mm EC shells? But honestly, there were two fights that really stood out in the game: clearing the raider hideout and starting a city-wide gunfight in New Reno.

The former was interesting because at that point I only had Sulik and Cassidy in my team, and only moderately powerful equipments (I was carrying an assault rifle, Cassidy has a combat shotgun, and Sulik had… I don’t remember him doing anything in that battle though. :S ). To make things worse, in my first attempt I approached from the heavily mined entrance, with nobody having enough lockpick skill to open that door. Not to mention the mines.

Suffice to say that I got my ass handed to me within minutes. The second time I tried the ladder, and fighting from close quarters with lots of cover I managed to clear the hideout (not without a few reloads though).

That was it: plan it badly, and you get your ass handed to you. Do it the right way, and you are rewarded with not only a completed quest but a whole variety of superior gear and ammunition as loot.

The New Reno fight on the other hand, while having little tactical significance, was memorable for the sheer chaos that ensued. Junkies were running like Olympian marathoners and some even got into fights with bouncers; at one point a hooker was torn to shreds for being at the wrong place at the wrong time. It was also a case where you tackle on the enemy thinking the odds are well balanced, only to realize that some of those people in the streets were jumping in the fight too. It was unpredictable and if not for the long waiting time between turns, it would have been my favorite activity in the game.

Splintered Core’s Combat

The examples mentioned above highlight a common factor in entertaining combat. They require the player to plan ahead, to treat that battle with a level of respect not present in modern RPGs where it is entirely possible to steamroll through every encounter, even bosses. This has nothing to do with the actual combat mechanic. As a contrast, the turn-based fight in the Oil Rig against evil Optimus Prime (forgive me for forgetting his name. Was it Frank? Damn. Too lazy to Google) paled in comparison to the fight against the demi-lich (Kangaxx or something) in BG2, and the latter was in real-time.

While mechanics have a great influence on the kind of tactics that are available in the game, and also governs the level of control possible over the flow of events, the most important thing is to actually provide enough tactical options to the player.

So what kind of options should we provide then? First of all, there must be multiple ways for the player to wrestle control of the battlefield. Having multiple status ailment spells to disable/convert/distract/debuff enemies are good examples; having multiple versions of Fireball of different temperatures are not.

In modern firefights, position and subsequently, zone control are perhaps the most important success factors. Flanking is made that much more important in a battlefield when we understand the typical human’s incapability of dealing with simultaneous threats from multiple angles. And by that, we arrive at the issue of area denials.

Ideally, you want to pin down your enemies, keep them in an exposed area and spray them with fire until they either give up or die. If they move, reposition so that you can force them back into the “killzone” by either overwhelming firepower or area denial devices like grenades and mines. In short, you will want to maximize your own people’s mobility while denying the enemy’s.

To achieve this, we need to have a suppression system in the game. Automatic weapons can be set to suppress a specific zone, firing and discouraging any movement across it. Breaching charges and destructible environment should be explored to provide players (and enemies) with more tactical options. And finally, damage should be lethal and NOT tied to experience levels. It is pretty pointless to have an option to overwatch an area only to have enemies moonwalk through them and back just because they’re wearing some Modern Warfare 2 Juggernaut armor (read my lips: I fuckin’ hate those abominations).

Here’s the kind of battle story that I would love to be able to simulate in Splintered Core:

You are about to launch an assault at an enemy camp with your party, which consist of one heavy weapons guy backed by two riflemen with automatic weapons, a grenadier, two melee specialists and a medic. Creeping up on the enemy, you send the medic (who doubles as a scout too) forward to gain line-of-sight at the enemies, and after observing them for a while (NPCs have schedules and timetables – they don’t stand around doing nothing), you realize that the best time to launch the assault is in the evening where most, if not all of the enemies would be gathered in the courtyard for the daily flag lowering ceremony.

You evaluate the map, and send your heavy weapons guy paired with the grenadier to the west of the camp, and the other two riflemen to the main entrance. The medic follows the riflemen, while the two melee specialists are moved to the east of the camp, ready to sneak in and do some serious damage.

After ensuring that everybody has gotten into positions, you sneak the two melee artists into the camp, and hide them in two buildings that you expect the enemy to pass through in their attempts to counter-flank your positions.

The idea is this: your heavy weapons guy open fire at the courtyard, which would send the enemies scrambling. Once they recover they will return fire, and attempt to flank your position, which they can only do so by sending people out through the main entrance. If they do so, they will be caught by your riflemen and your medic who doubles as a grenadier. After sustaining a couple of injuries they will get the idea that their only option is to go east, which if they do they would risk spreading themselves too thin, allowing you to move your heavy weapons team forward INTO the base to utilize the base’s defenses against the enemy. Your melee artists would help in the advance, by hitting the base defenses from both ends. At this point your riflemen can reposition themselves to deal with the enemies exiting from the east of the camp, and if everything works properly you would be able to funnel them BACK into the camp where your heavy weapons team would be waiting to cut ‘em down.

If all works out well, you would have successfully pinned the western side of the camp down with suppressive fire, denied the front entrance with the two riflemen (which the enemy would naturally assume that you have ANOTHER heavy weapons setup on that front) and subsequently forcing them to split their forces. That would then significantly improve the survival odds of your melee artists, which you then spring to mow down the base defenders, which contributes to your eventual success in isolating the second enemy group attempting to flank your riflemen.

Of course, to achieve this would need a fair bit of realistic enemy AI, but still the argument here is to provide the actual game mechanics to make this possible.

That’s it for now – I’ve written more than I should, and this is possibly the longest post in my blog so far. I’m archiving this post to remind myself of the kind of combat I want to deliver to my players.

The goal of the NPC AI framework is to provide a general guideline to designing believable behavior of the NPC that reflects the consequences of choices the player makes through the game. Let’s take an example and break it down, as usual.

Tycho chats up with the player in the bar, knowing that he is a stranger in town and could do some errands befitting a stranger. He talks to him about a job with nice pay, after finding out that the player is a person who is willing to overstep the boundaries of law for personal gain.

Tycho recognition check: does not recognize player

Tycho moral compass check: player is chaotic neutral (for example)

He then “suggests” the player “remove” a particular NPC named Nancy, who he claims to be disrupting his business. Once Nancy is gone, the player can come back and collect the money. The condition is, the player must keep this a secret, and make the death look like an accident.

Tycho quest marker: quest initiated

The player then proceeds to Nancy’s and observes the NPC’s schedules for a while. He notices that Nancy often climbs down an abandoned well to pray in the bottom of the well, and schemes that a slight knock will cause the walls of the well tumbling down, crushing Nancy underneath.

The player casually walks up to the well and pushes the rocks in, and returns to Tycho for payment. Tycho tells the player to wait a few days to see if the job was done according to requirements.

The next morning, and NPC notices the well being broken and investigates. The NPC discovers Nancy’s body, and reports it to the Sheriff. The Sheriff comes and take a look, and then realizes that it is not an accident but a murder. He alerts the entire town to the murder and begin questioning suspects.

Townsperson recognition check: does not recognize well

Townsperson script trigger upon foreign data: investigate data

Townsperson discovery check: well is destroyed

Townsperson destruction check: ignore well (low priority)

Townsperson discovery check: Nancy’s dead body

Townsperson destruction check: report death (high priority)

Sheriff reaction check: death investigation (high priority)

Sheriff discovery check: well destruction is not natural

Sheriff reaction check: alert for murder in town (high priority)

Sheriff detective check: check NPC schedules

Sheriff detective check: no NPCs passed by

Sheriff memory check: player (stranger) is in town

Sheriff script trigger: murder dialog script with player

The sheriff proceeds to investigate all NPCs that happen to pass by that area, and finds nothing. Then he remembers that the player is in town, who is a shady stranger and proceeds to question the player.

Is this a nice way to create emergent behavior? Take for an instance the player kills an NPC. Nobody knows of the murder, right? After a few days, other NPCs tagged as the dead NPC’s friend’s will notice his disappearance and will investigate, eventually triggering the sheriff’s investigation script.

If the player leaves town before the news got out, he might be spared from the questioning, but the information bit would have been flagged and next time when he returns the sheriff will seek him out and have a world with him.

Would this be a good way to inject life into the game world? Will this be too much to handle?

This post will detail how the combat system works.

Combat in Splintered Core will utilize a tic-based, phase-based timekeeping system, in which actions happen in real time based on time units called “tics”. One tic is equivalent to 100 milliseconds (which will be adjusted during testing time for balance purposes), and every action has a tic-cost. For example, moving 1 tile would cost 10 tics (1 second), while moving 1 tile diagonally would cost 15 tics (1.5 seconds).

Tic Cost Reduction Bonuses

Tic cost is reduced by 1% for every rank the player has in his speed stat, which is further reduced by 5% for every familiarity level for applicable actions. This means an action can have a maximum of 25% reduction on tic cost.

Rush Meter

Every character in combat has a rush meter that begins with 0 and fills up when it reaches 100. Rush meter further decreases tic-costs to a maximum of 20% (every rush point gives a 0.2% reduction), and special techniques incurs a certain rush point cost.

Rush points are gained when a character performs a successful combat roll. Passive combat rolls (where the character is the target the roll) gains lesser rush points than aggressive combat rolls (where the character is the initiator of the roll). Multiple consequent successes triggers multipliers that exponentially increase the rush point accumulation rate, until the chain of successes is broken by a failed roll. Note that multipliers can only be gained through the commitment mechanism (see below).

Getting hit in combat reduces the character’s rush meter, in addition to the natural decay of the meter per tic.

Commitment

When combat begins, the player enters the planning phase, in which the player can decide on the level of commitment. What this means is that during the planning phase, players can choose their party members and issue chains of commands.

For example, the player can tell the character to move 5 tiles to the east, kneel down, set up an overwatch zone and hold it for 50 tics. Assuming that each tile movement costs 10 tics; kneeling down costs 15 tics; and setting up an overwatch costs 20 tics; the total tic cost for the chain of command is 135 tics.

Now when the player presses the execute button, the character will carry out the chain of command. The player has the option of breaking the chain of command prematurely by pressing the enter planning phase button, which will cancel out accumulated rush multipliers.

Rush multipliers can only be generated through unbroken commitments. Using the example above:

Successful Commitment

The player runs 5 tiles to the east, getting shot at twice along the way but succeeds in both fumble checks. Upon reaching the destination, he kneels down, again getting shot at but succeeds in the fumble check. He sets up the overwatch under enemy fire, and through sheer luck survives all fumble checks, and begins returning fire for 50 tics, succeeding in all his combat rolls.

In total, the player should now have more than 10 consecutive successes, netting him a nice x3 multiplier to his rush meter. Therefore, instead of gaining (for example) 1 rush point per tic, he now gets 3 rush points per tic.

Broken Commitment

The player runs 5 tiles to the east, getting shot at twice along the way. He succeeds in the first fumble check but fails the second one. The player’s further actions are immediately cancelled – along with any multipliers to his rush meter – and the game returns to the planning phase.

I hope this explains the combat mechanics well.

Having lined out how the basic mechanics work in the previous post, in this post I’m going to attempt to define the framework for melee combat.

Melee combat is inherently different from ranged combat. In Splintered Core, it is the goal to balance out melee and ranged mechanics so that melee specialist builds have a fighting chance against ranged specialists. In the most extreme case, a ranged master build would have around 50% chance of winning an encounter with a melee master build.

Before we begin, let’s examine a few ways unarmed/melee combat can gain the advantage over ranged builds. To do that, we first examine the drawbacks of ranged builds. All firearms have a minimum range, and usually cost more to fire a shot. This is due to the act of target acquisition; aiming and firing being abstracted into one single action, which is far slower than unarmed combat. Which brings us to a balance design decision: while which build (ranged or melee) is faster is certainly a highly debatable issue, I will be making a balance decision here to ensure that melee builds are always faster than a ranged build of the same level.

Ranged vs Melee

This is partly due to the stat requirements of a melee build: for ranged builds, the senses stat tree is important; while for melee builds the atheletics stat tree would have priority. Development along these two branches would ensure that range builds will become great in determining enemy positions (e.g. a ranger-like build in traditional RPG conventions), while a melee build would instead of a powerhouse in direct physical confrontations (e.g. a warrior-like build in traditional RPG conventions).

The modus operandi of two builds would also be vastly different. Melee builds tend to have lots of special skills in negating reaction checks while maneuvering in the battlefield (mostly agility checks), and I would expect melee builds to invest extra stat points into either stealth or maneuvering, which is far more important to their survival on the battlefield.

The path to a target is always the most perilous of a melee build’s journey. Once the melee character enters a ranged build’s deadzone, tactical advantage would greatly favor the former as ranged characters suffer massive penalties when attempting point-blank shots. Sidearm masters would fare better in this case, but they would not have the necessary knowledge or experience to deal with disarming or grappling moves that melee characters would attempt once within melee distance.

Match Up

My intention in this design is simple. When a ranged build is matched up against a melee build, the tactical options for the former would be to maintain distance while attempting to successfully defeat the latter’s speed advantage. Shots from a firearm would deal lethal damage to melee builds who usually shun heavy armor in order to move faster or without encumbrance. Also, in this situation ranged builds would fair better if they changed to close quarter combat weapons like pistols, carbines, shotguns and submachine guns, sacrificing stopping power (damage and penetration) for volume of fire in close ranges (fire rate and AP cost).

Melee builds on the other hand, will spend most of their time plotting the approach. The key to a successful melee build is to know how to utilize the terrain to approach under cover, and undetected. Their lesser reliance on equipment protection allow them better chances against reaction and detection checks, and once they successfully make their way to the target they will have a wide variety of reaction-resistant movement techniques that would allow them to close the remaining distance without triggering deadly reaction fires.

And once a melee build closes to melee distance, the range build would have practically lost the battle already, unless he has some backup melee skills of his own.

Ranged builds, when fighting against melee opponents, would play a very defensive game while the latter would almost always be on the offensive.

Mechanics Test

It’s better for us to use a real-life example of how a fight goes, and then translate this into mechanics and combat rolls.

A is a master in striking arts, capable of delivering precise blows to vulnerable spots through his opponent’s guard with the “Serpent Strike” technique, and is a master of the “Three-inch Force” technique that allows him the capability to deliver full-powered blows from awkward and unpredictable angles with ease.

B is a grappler whose modus operandi involves taking down his opponent and then attempting to manipulate his opponent into submission holds.

A and B meets in the field. The bell rings, and the two combatants circle each other, looking for an opening. B is patient – he knows that a defensive starting strategy will give him a vast advantage against A, whose style he is familiar with. In fact, he had beaten several so-called grandmasters of said style before, and he knows the style inside out. He prepares himself for kicks that A’s style is so famous for.

A on the other hand isn’t that fortunate. He does not have much experience with grapplers, with rudimentary training in counter-grappling skills, but he knows he has one advantage: he had been training so hard in his style that he had reached the point where he can deliver a powerful blow from any angle, even from limited spaces.

Knowing that B would not take the first move, and that starting with a kick would be too predictable, A decides to switch it up and charges B. His tactic is to feint a charge and then at the last moment, sidestep and deliver a light, harassing cross to B’s jaw before quick stepping backwards to put distance between the two of them. He knows that close distance combat is B’s home ground, and thus keeping a distance would be his main strategy in this fight.

He executes his plan, charging in and then sidestepping. B couldn’t react fast enough and eats a light punch to his jaw, which dazes him for a split second. He tries to reorient himself and tackle A, but it is too late: A had moved out of his grappling distance. Knowing that A is much faster than he is, B changes his strategy. He takes the initiative instead and pressed forward, pulling up his guard to weather A’s flurry of blows to get close enough.

Bad idea. A does a spin kick and drives his foot between B’s guarding arms, landing a solid hit on his chest and knocking the breath out of him. B stumbles backward, his guard slackens, and A follows up with a vicious combo known as the “Hidden Tiger Pounce” where he closes the distance and lands a right hook; followed by a right uppercut, and then a one-two punch before ending it with a reverse leg sweep.

B successfully defends against the right hook but that leaves his guard open, and A slams the uppercut through his guard and flips his head backwards with the force of the blow. The one-two catches the rebounding head solidly, and B drops like a fly. The leg sweep was not needed at all.

Before B hits the ground, B realizes his fatal mistake. He had underestimated the striking precision of A’s techniques and the amount of force he could generate from virtually any angle. His guards were designed to block general blows, but to A they were laughable. Had it been any other striker, B would have gained the advantage by aggressively pressing on the attack with a strong defensive posture, but A had the ability to penetrate any guard with his style’s “Serpent Strike” and “Three-inch Force” techniques.

Fortunately, A deliberately avoided dealing B fatal blows, and B is merely knocked out.

I apologize for my horrible writing. I hope it does the job though, which is to give you a mental image of how the fight went. Now, to translate this into mechanics:

A’s stats:

Unarmed, Striking = 8, Mastery Level = 3 (Master, Three-inch Force)
Unarmed, Counter = 5, Mastery Level = 0
Dexterity, Agility = 8, Mastery Level = 3 (Master, Flash Step)

Equipment:

• Hands = Light Gloves (-1 Penetration, -2 Damage)
• Body = MMA pants (0 Penetration, 0 Damage)

Three-inch Force
+2 penetration and damage. Bonus applies even when clinching or grappling.

Flash Step
3 free movements immune to reactionary counters.

Hidden Tiger Pounce (Learned Technique)
Delivers a hook, followed by an uppercut, followed by a jab and a cross, and then ending with a reverse leg sweep. All actions are performed at 50% AP cost. If first hit connects, next hit is delivered with a +2 bonus mod; every hit that connects after the first gets a +2 accumulative bonus. However, if any move is dodged, the opponent gains the same amount of bonus to his counter-strike.

Striking Familiarity Level 3
Gives a +3 bonus to combat rolls involving striking maneuvers

B’s stats:

Unarmed, Grappling = 8, Mastery Level = 2 (Expert, Advanced Takedown)
Unarmed, Counter = 6, Mastery Level = 1 (Advanced, Flow Manipulation)
Dexterity, Agility = 5, Mastery Level = 0
Dexterity, Reaction = 6, Mastery Level = 0

Equipment:

• Hands = Light Gloves (-1 Penetration, -2 Damage)
• Body = MMA pants (0 Penetration, 0 Damage)

Advanced Takedown
Re-roll a takedown if unsuccessful on the first try with a -1 mod

Flow Manipulation (Passive)
+1 chance to divert damage from one body part to another

Grappling Familiarity Level 2
Gives a +2 bonus to combat rolls involving grappling maneuvers

Anti-Striker Familiarity Level 2
Gives a +2 bonus vs strikers.

The combat begins with both combatants aware of each other’s presence, so no surprise environmental factors are in play here. Also, since both of them are fighting in a ring, no external factor comes in either.

Thus it’s pure skill vs skill. One thing though… B has vast experience in fighting A’s style, and thus gains +1 familiarity level vs A.

Both roll initiative, but B wins the initiative. Both combatants are 2 tiles away from each other: within distance to kicks and extended punches, but beyond distance for grapples.

B prepares a counter vs kicks, giving him a +2 advantage towards kicking strikes.

A uses the flash step ability, which gives him 3 free moves. He takes 1 step towards B to close towards grappling/punching distance and throws a fast cross, targetted at B’s head.

A’s agility = 8, flash step = +2

Martial Technique Schools

Wildstrike

Wrestling

Archonite Hand-to-Hand Offense Neutralization System (ARCHON system)

Combine Grappling

Janissary’s Silakhas