Combat Systems: (TL/DR at bottom)
I'm doing some groundwork for myself for this right now, figured I'd talk it out, out loud. I'm going to assume a discrete Cartesian coordinate system and opponents of a single size wielding weapons
Factors of Combat:
This is what really matters or could matter. Few systems implement more than a few of these. All of them can and should be modified for character, weapon and move
Speed
How fast an action is.
Range
given a square and a facing, what squares are targetable
Accuracy
How likely is the attack to hit. This is degenerated into a simple probability, but is usually calculated from an attack value and a defense value.
Attack
Ability to Hit
Defense
Ability to not get hit
Damage
Obvious, but may depend on variables such as critical hits, hit location, armor and accuracy of the blow
Damage Reduction
Ability to absorb/reduce damage
Position
Different stances are good for different things. At the end of some moves, you are out of position, or in a different position. Two good fighters may not be able to hit each other without getting the opponent out of position first.
Damage over Time
This is usually associated with MMORPGs, but is simply the damage capacity of a strategy with Weapon Speed vs Damage vs Accuracy all factored out. It allows you to compare a slow, inaccurate axe with a quick dagger, with an accurate sword.
Number of Attacks
Similar to Speed
Wind Up
Time it takes to start a move. Usually used in fighting games (Street Fighter). Useful for AAO interurpts
Wind Down
Time it takes to finish a move Usually used in fighting games (Street Fighter) useful for aao counters
I'm sure there are more things that can be tracked, and I may add more later.
Example: D&D
The D&D combat system is useful for VGs, both because of what it does, and what it doesn't do. Hitting in D&D is an all or nothing affair. Each attack either hits, dealing a full measure of damage, or misses. Damage Reduction is a rarity, as Armor serves the purpose of defense and not DR.
Most important is the fact that D&D is a linear system. Two opponents of equal skill have a 50% chance of hitting each other. Increased skill gives better odds in 5% increments maxing out at 95% chance of hit or miss. That means that there are 20 gradiations of skill possible.
Blah blah blah, it's boring and coarse. The advantage is that most fights will degenerate into contests of DoT which are quickly resolved. (Rather than hours of misses with only a few hits required to win).
Computers have advantages over PnP systems, because they can do the math instantaniously and they can generate the random numbers extremely quickly. That means that you don't need to worry as much about overcomplicating your system with multiple sub-systems. That is, you can have a complex positioning system feeding modifiers into a complex to hit system feeding into a complex b-p damage system, and not have to worry about the player keeping up with it all.
What you have to worry about is the user inputs. Say that it takes a keypress to attack your oponent. The complexity of what happens when you attack doesn't matter, so long it's all resolved under the buttonpress. If the keypress is unlikely to change anything, you have a problem. This really comes into play with high defense values. If all parties have a 5% hit probability, things will only change once every 10 keypresses (you attack, they attack). This is boring and should be avoided.
Additionally, complexity can only modify probability between 0% and 100% It's a fixed range. If you create to fine of a scale of differentiation, individual tweaks will no longer be worth making. (Base chance to hit is 20%. +.01% to hit spell isn't worth casting. Ever)
Statistics
One of the problems I have with D&D is that the end points aren't scaled. Personally, I'm looking for a nice bellcurve shape on all my skillsets. If I've got 100 points in attack, the difference between 100 defense and 110 defense should be bigger than the difference between 110 and 120, and bigger even than the difference between 200 and 1000. I also don't want to worry about max stats, and I want equal stats to be 50%. This means that a simple X - Y vs 1d20 isn't going to cut it. I'll probably use a logarithmic scale.
Weapon Balance:
Sowelu had the master post on balance earlier: Encourage players to do what you want.
At the end of the day though, you want your players to have a few roughly equivilent models. While a normal accuracy/speed/damage normalization works on simple systems, you don't always want to reduce a weapon to it's lowest common denominator.
There's three methods. Obviously, the first is equivilency. Every weapon is basically the same as every other. Offsets in speed are exactly offset in damage and accuracy. All bonuses apply equally well to all weapons. (Increasing weapon speed by 20% does 20% more damage for swords, axes, and daggers alike). The second system is Generalist where the stats are pretty much orthogonal, and the quality is largely determined by volume (+20% speed to a dagger doesn't help much, but it more than doubles an axes effectiveness). Finally Specialist systems encourage players to specialize in a trait. (Adding extra accuracy points to the axe doesn't help, because it's so far below the average opponents defense that it's lost in the log, while the sword uses it to get a significant advantage.)
I'd like to see a system that encourages both Generalist and Specialist behaviours.
Armor
I want armor to serve as DR, not defense. Additionally, I want armor to be good for everybody, but better for lower defense characters. (I.E. Plate will protect the expert swordfighter, but it doesn't improve him as much as the newb) In order for this to be true, I'll likely use raw DR and percent combined, along with a defense penalty. Since the newbs defense won't help as much as the DR, that'll work there, and the expert will face a tradeoff in useful defense for guaranteed DR.
Combat System:
Here's what I've got so far: (Mind you, I'm planning a tactics style game, not flat-up rogue)
Each Character will have an (derived) Attack, Damage, Defense, DR, and Number of Attacks. (Range can be abstracted for now, and I don't plan on using speed). Characters will also have counter, interdict, balance and position pools.
Combat Turn (Move Turn happens first, as I'm separating Combat from Movement as a design decision)
Regain Balance
Reset Balance Pool
Attack : Choose an Attack move from the available list (randomly, or choose, or just one).
Interdict? : If the opponent has points in interdict, check interdict chance and if so, resolve that attack before continuing. It is possible to stop attacks completely with a successful interdict.
Hit? : Determine if the attack hits by comparing Attack/Defense (Math not finalized here)
Damage : Based on the accuracy, deal the weapon damage. (Kind of a sliding critical hit scale, where you deal 50% at the bottom, 100% at 50pts above the min to-hit and increasing from there. This serves as a scaled in damage reduction (A good defense takes less damage). DR also applied by armor.
If you have a 75% hit chance and are rolling d100, you deal 50% damage at a roll of 75, 100% damage at a roll of 25, and every other roll is extrapolated from there. (Thus, if you only have a 25% chance of hitting, you will only ever do 75% of max damage). Critical hit ranges may apply (0-10 = double normal damage?)
Combo
Some attacks enable lesser sub attacks. For example a feat that enables a free punch after a punch attack.
Counter? :
Check for counter attacks as for interdicts
Next Attack
It is possible for characters to have multiple attacks per round
Position
Refill Position pool
Position and Balance
Attacks deal position and balance damage, as well as incur position and balance costs. (these may be collapsed into 1 pool). At the end of your turn your position pool refills somewhat. At the begining of your turn, your balance pool fills somewhat. Thus, it is important to manage these two pools.
For low-level characters, they should refill all the time, and very few attacks should cost either. Advanced strategies involve using them both. Batting your opponents blade out of position before lunging, finding that lunge overextends you, and your opponent kicking your feet out from under you should be modeled that way. The purpose is to create a secondary subsystem for controling significant numbers of combos and extra attacks. (since a 3 attacks each consisting of 3 hit combos is 9 attacks)
Ok, I've rambled enough for one time frame.
TL/DR : Talking about actual combat systems, brainstorming, and what I'm thinking of doing. It's complicated.
Author
Topic: The Roguelike Development Megathread (Read 245783 times)