I'm currently helping to design a pc game (Mac/Linux/Win32) that is based off of *multiplayer* swordplay. I'm sure this has been visited beforeI'm just now realizing that I don't know of a (real) precedent on how to develop just a game. Consoles have various implementations - the best I know of is "Kengo:Master of Busido", and Busido Blade. Of course, these excellent games are taylored for the playstation gamepad - something not known for translating to PC controls easily. The most dazzling display I know of is Knights of the Old Republic - however it's combat is D&D style, with some really heavy NWN flavoring. Jedi Academy/Outcast doesn't count, it's just button pressing to me. What I'm looking for is a skill based melee combat system - I could even go as far as to call it a low-fidelity simulation. This could be a blessing in disguse, I can set the standard myself for once, instead of following precedents set by previous games. This game is unique in many ways to everything else, and is likely to make a few waves. I know this has been discussed before, but not in the detail I require. I'm posting here to grab a little feedback from those of you who can "imagine" what it would be like to play while I prototype the game. (It's not time to talk alpha testing quite yet, it's a little early for that). Bear in mind that this is a multiplayer *cooperative* PC game - mouse and keyboard are all that can be reasonably assumed. Also note that I'm NOT looking for a PC version of Dynasty Warriors. For a clue as to the context of this post, look here: http://www.gamedev.net/community/forums/topic.asp?topic_id=280968 Here goes: -------------------------------------------------------------- The major variables that affect combat in the game(for sure) are: battle fatigue - the more you fight, the more exaustion you build up, taking hard blows erodes your exaustion also. Meant as a general timer that governs how much continious fighting you can take and remain effective. knockouts - taking an exceptional blow or falling due to extreme fatigue can cause you to fall unconscious. Other players can rustle you awake, fatigue reduces (slowly) while out cold - when it's full again, the player (slowly) wakes up but will be wary for a short while. disarms - are a major part of gameplay, the players swords are quite destructive, and are absolute nightmares when placed in the hands of the enemy. The odds of getting disarmed climb with fatigue/damage, each disarm has a level of severity also: how far the sword falls away, and if the player was "cleanly disarmed" (weapon wrenched out of his/her hands directly.) grapples - players can lock swords with other enemies. I'd like to avoid a random dice roll to determine the winner, ju'uken (paper-rock-scissors) isn't quite enough either. It needs to be fast and easily resolved - but hard to cheat. Also counts for when players are grabbed by oversized enemies - maybe the player can squirm out before he/she earns a KO. parry - a big thing. Players should be able to react with more than simple blocks if this is to be an efficent fighting engine. flanking - the player's #1 weakness. To prevent being overrun the players will learn to fight together - usually back to back. attack stances - A way to quick select fighting styles for different situations/strategies. Players will often fight back-to-back, or in formations. Players need to be able to quickly select (custom?) move sets: jabs, sweeps, charges, etc. - similar to Kengo. (Idea: Allowing players to compose their very own movement sets/combos would rock...) horses - players and NPCs can be mounted. Scenes will be based on horse chases, so I've gotta keep this in mind. weapons - for now, there is only one type of sword, as the simulation improves, things such as blade length will be taken into consideration. Most weapons also have "special abilities" that can be activated, and passive effects on stats. (I know I need to consider both pikes and bows.) dual weapons I've yet to consider akimbo weapons due to unresolved balance issues. some characters will have more than 2 arms, or a tail - but won't be player characters. Generally, players will most certainly have a 2 bladed weapon maximum. accessories shuriken, kunai/tanto, may eventually be allowed/incorporated - it depends on how it affects balance. NPCs will likely get to use them. -------------------------------------------------------------- I'm a little bent for overall control ideas. It's gotta be fast/effortless - but not mindless. I'm wondering about using: "Freeflowing" systems: * A system that links your sword directly to your mouse. Joysticks would work also, trackballs would present a certain advantage (or so I think). * A guided system that uses the mouse to establish relative movement of the sword. Click and drag to swing. Macro Systems: * A mini-gesture system (or button combinations) that allows the player to select moves from a premade library of moves ala mortal kombat. (Where is the fun in that? Leads to button pressing) * A specialized gesture system that allows the player to create moves in an editior, or from a vast moves list. This would still require an implementation of the freeflowing system in a (in game?) move editior. The free-flow systems would require adding ragdolls/mesh collisions to the engine with either Tokamak or ODE, possibly novadex (EEEWWWW!), motion capture, several bottles of vodka, and gallons of orange juice. I'll have to prototype it using OGRE and see how it works. Any suggestions?

Not sure if this would fit into the type of game you're doing, but how about the inability to use the sword when in very tight or small areas? Light in real life, how are you going to swing a sword when the walls are touching your shoulders?

That depends wholly on if I get my ODE physics with polymesh collisions. I indend on doing ratherprecise detection with the swords - walls would be trivial. I imagine that players would simply use Parting/stabbing attacks when their sides are to the wall, I can't forsee *too* many effective 360deg moves anyway.

I would suggest a Hybrid Mouse-Macro method.

The player uses specific mouse movements and curves to activate specific moves. Similiar systems have been used in Black & White and Arx Fatalis initially for convience and later as a Magic system. If you want a free flowing game that requires less 'mortal combat' button mashing, this would probably be the one to go with. Though the full potential of Mouse Macroing through movements hasn't been explored that much in game development.

An example would be holding down the attack(mouse) button and moving the mouse down alot, and up alittle to activate a macro where the players sword stabs his opponent, and moving the mouse up alot, and down alittle for the player to perform an overhead swing. Moving the mouse up and to the left in an upward curve to perform a raised block with his sword, and moving the mouse down and to the left in a downward curve to do a parry (assuming the player is holding the sword in his left hand).

The mouse itself wouldn't control the moves directly, but by completing specific patterns, the Macro (or combination of macros) would be activated, much like how a Rune being drawn in Arx Fatalis doesn't do anything until your finished drawing it.

*edit* there are different ways of doing it, Freeform may work equally well, depending on implementation.

[Edited by - Gyrthok on November 14, 2004 12:17:35 PM]

Quote: horses - players and NPCs can be mounted. Scenes will be based on horse chases, so I've gotta keep this in mind.

By having mounted combat you increased the difficulty level by at least 2, if not more. I would think that dealing with the regular non-mounted combat should be dealt with before turning to the mounted case.

If the player has only one sword, then he should have some choices about what to do with the other hand: 1) he can use a shield 2) he can wield a knife 4) he can punch or grab people 5) he can activate switches, open doors, pick up items, etc. 6) he can control his horse 7) he can swing his sword with two hands, adding power but changing his range of motion somewhat. Obviously you probably won't want to implement all of those, but you should consider what to do with the player's other hand, since if you force it to remain on his sword he will ask things like, "How did I just pick up that thing on the ground? With my feet?" and, "Why can't I let go of my sword with one hand to extend its reach a bit?"

If you ever implement horses, you should have different qualities of horse. Some would be farm horses: slow, easily spooked, and more likely to fall than other types of horses. Others would be war horses: huge, quick, fearless, and living weapons. Probably something in between would be appropriate as well.

You'd definitely have to add some detail to it to get what you want, but I've always felt that Thief 1 & 2 had some nice swordplay.

You couldn't swing the sword if you were in tight places and you and your enemy could hop back to dodge an attack sort of like fencing. I always felt that the swordplay in Thief was pretty immersive.

I had the same idea but I couldn't explaine it better.
the joystick could be better for "immersion" but the mouse also works(its mandatory because you cant count on that everyone owns a joystick).
You've inspired, better take note and retake the design of my military hack & slash game(nice ring to it).

PS. My vision was a single player main mode with epic battles, but multiplayer doesnt sound bad at all.

Two handed weapons are appealing, in traditional martial arts the off hand is (usually) used for defense/parry. I'm just turned off by the additional motion capture necessary to make it sing. There is enough for now, this will make a good enhancement.

To that end, the (empty) off-hand can be used to present some pretty killer moves.

I've given some thought to the wall thing too. I'd like to make it where weapons can get "stuck" in trees and wooden walls. Should present just a touch of comedy the first time it happens. (And a bit of strategy when being attacked - having an enemy embed his sword in the wall would be kinda cool, especially if you lured him into it.


Simple strafing isn't quite enough either. Real martial artists MOVE. Ducking, handsprings, dodges and leaps. I'm quickly stretching what I think I can do with a mouse and keyboard.

It's looking like joysticks/trackballs will have an advantage no matter what. Using gestures make using a touchpad feasible too.

Thanks for the input so far - this is coming together quite nicely.

To my knowledge, there was a game called Rune that was more or less what you're describing. Been a long while since I've seen it.

Jedi Academy's lightsaber fighting was a little crazy, but Outkast's was pretty good. I remember long tricky duels that I got into in that game that were a lot of fun and weren't random button pressing. The trick is that the learning curve for such games is pretty damned steep. It took me months to get good enough at Jedi Knight to hold up in a real duel. The same with Bushido Blade...in fact more so, as one hit meant you were done. Though of all the one-on-one games I remember, my duels with my friend in Bushido Blade were my favorites. Those fights would go on forever with us parrying and dodging everything we threw at one another.

Those games are beautiful, but their learning curves are considerable. If you can find a way to make it so the control is as intuitive as say Tekken but involves weapons, I'd be very interested to see how it turns out.