Idea #1 - In one spell-based fighter that I wrote, I had effectively 4 meters for the players, health (self explanatory), attack (used for all attacks), defense (used to raise your defenses) and movement (used for temporary bursts of speed). This worked reasonably well, but, it became really hard to keep track of all of this. To avoid having to constantly look up at status bars, I replaced the standard meters with little runes that floated around the player, but, even then, you spent more time looking at your character than at the battle at large.

I had it so that you could keep on holding down the appropriate button to charge up the attack that you were about to perform, though of course at an additional cost. Defense required a constant maintanence cost to keep your shield up.

In other fighters, there's the typical super-meter, which is similiar to this line of thought, except its a lot simpler to be able to glance at that flashing bar in the corner and now know that you have an additional move available to you.

Soul Calibur 4 has the force meter for doing the "force" moves with the star wars guest characters, that sorta works, but again is hard to keep track of I find. In it, your character curses and is left momentarily open if you try to perform a move that requires it on an empty meter.

Its also important to consider how these meters are going to refill, if at all. Will it be a regeneration over time? Do they recharge as you attack (usual method for super-meters in 2-d fighters)? I had my meters regenerate at a bit of an exponential curve relative to when you last used them (started out slow, but, recovered quickly after a few seconds).



As for the magical elements idea - This might be able to work, though you'll have to be careful to make sure all of the elements are equally valid, or you'll be able to build a character that completely dominates another.



As for timing - Every fighting game relies heavily on timing. A light punch does less damage than a heavy punch but it comes out faster and recovers faster to compensate. This lets you either interrupt "unsafe" attacks or punish the opponent with a quick attack while they are recovering.

For state changes, look at the instances in Soul Calibur. They confer various advantages while being in them, but, take a bit of transition time to go into, and reduce either your movement or attack options to a specialized set.


As for your typical attacks/defenses, you have (forming a system similar to what Edtharan outlined, though with a ton of interacting options):
-Normal attacks, in fast/slow/high/medium/low varieties.
-Blocking, in high/low varieties
-Unblockable moves
-Guard break moves
-Stun moves
-Knockback/knockdown moves
-Moves that simulataneously attack and shift your position in some way
-Usually some sort of parrying
-Jumping/crouching/rolling
-Sidestepping
-Projectiles, typically with an option to deflect said projectiles back
-Environmental situations, Dead or Alive is probably the best example of environmental interaction in a fighting game I've seen (the car level was awesome for this). Things like wall hits, ring outs, multi-tier stages, etc.


Anyone who ends up spamming a move, or couple of moves, will find themselves beaten by their opponent provided they know how to deal with the situation. Whether it be the right timing for a parry, or a quick interrupting poke attack. If there isn't a way of dealing with any move, then that move is overpowered and unbalanced.


Anyhows, feel free to hit me up for ideas, I'm in the process of starting up a new fighting game project of my own here.

like chess I find that games with simple rules have the most strategy while games with more complex rules are a mess of balance problems and one or two broken tactics that always win or just so hard to follow everything is effectively random from the players point of view.

Quote:

Original post by andur As for your typical attacks/defenses, you have (forming a system similar to what Edtharan outlined, though with a ton of interacting options): -Normal attacks, in fast/slow/high/medium/low varieties. -Blocking, in high/low varieties -Unblockable moves -Guard break moves -Stun moves -Knockback/knockdown moves -Moves that simulataneously attack and shift your position in some way -Usually some sort of parrying -Jumping/crouching/rolling -Sidestepping -Projectiles, typically with an option to deflect said projectiles back -Environmental situations, Dead or Alive is probably the best example of environmental interaction in a fighting game I've seen (the car level was awesome for this). Things like wall hits, ring outs, multi-tier stages, etc.

Ok, taking all advice under consideration I'm going to scrap most of my original ideas.

Here's my revised idea. I'm going to make a "sword and shield" character who attacks with his sword and blocks with his shield. Casts, attacks and blocks can be any of the 8 directions (up, up and right, right, right and down, down, down and left, left, left and up) but "downward" attacks / casts / blocks are only applicable in the middle of a jump.

3 elements Fire Earth and Water with Absorb, Shield and Attack

Any "spell" has three levels and the spell level increases offensive spells in dammage and cast time and defensive spells in cast time and duration.

Shield is a static spell which leaves an artifact on the screen which can be moved away from or through and decreases effectiveness of ALL spells passing through it regardless of the caster

Water Shield
Decreases Water attack damage 5%
Decreases Earth attack damage 25%
Decreases Fire attack damage 50%

Earth Shield
Decreases Earth attack damage 5%
Decreases Fire attack damage 25%
Decreases Water attack damage 50%

Fire Shield
Decreases Fire attack damage 5%
Decreases Water attack damage 25%
Decreases Earth attack damage 50%

Absorb is a defensive spell that remains on the caster and replaces any previous absorb spell. Absorb is a longer lasting defensive spell cast with high cast time and duration.

Fire Absorb
Decreases Fire attack damage 75%
Increases Water attack damage 75%

Water Absorb
Decreases Water attack damage 75%
Increases Earth attack damage 75%

Earth Absorb
Decreases Earth attack damage 75%
Increases Fire attack damage 75%

At any time you can only have 2 elements "Equipped" a primary and a secondary, and secondary specialization has 25% reduction in effectiveness.

I will look at these more in depth to ensure that no primary and secondary specialization is disproportionately weak against another one. (If all the modifiers are the same then each of the six specializations should have the same number and types of weaknesses and strengths) Changing specializations is still evident to the other player and takes approximately twice as long as the worst stun state. A specialization is "Secret" but will become obvious as the other player uses spells.

Quote:

Anyhows, feel free to hit me up for ideas, I'm in the process of starting up a new fighting game project of my own here.

I may do that!

Quote:

3 elements Fire Earth and Water with Absorb, Shield and Attack

From my initial look at it, the core is a fairly simple Scissors/Paper/Rock type system.

Water beats Fire
Fire beats Earth
Earth beats Water

You have then added the Shield, Absorb, Attack System on top of that (and to implement it).

Another layer you might want to add to increase the number of options available to the player is that of Attack Rates vs Damage Reduction.

How this works is you have an ability /spell/stat that reduces the damage of a hit by a set amount.

For example: If you had an attack that did 5 damage each hit, and the target had a Damage Reduction of 3, then the target would take only 2 damage. However, if the attacker had an attack that did 10 damage then the target would take 7 damage.

This makes interesting choices because an attack that did lots of fast, but individually weak attacks are more effected by this Damage Reduction than ones that are slow but powerful.

For example: If a character had two attacks:

1) 5 damage each hit and hit 4 times a second.
2) 15 damage each hit but only hits 1 each second.

Which is the better attack to use?

Well if you have an enemy that has no damage reduction, then the first attack would be best as it does more damage overall. But, if the enemy had a damage reduction of 2, then the first attack would do 12 damage in total ( (5-2) * 4 = 12 ), but the second attack would do 13 damage every second.

So if the enemy has a damage reduction of 2, then the second attack is slightly better. If the damage reduction was higher, then it would make the second attack significantly more powerful as compared to the first.

The rule of thumb here is that slow attacks do less total damage against a target without damage reduction than the quick attacks. If the slow attacks was as good (or better) than the quick attack against a target without damage reduction, then they would be far better as there would be no situation where they would be less effective than the quick attacks.

You could keep the scissors/paper/rock base and have a damage reduction based on the elements.

You might have this:
Water
DR vs Fire attacks: 2
DR vs Water attacks: 1
DR vs Earth attacks: 0

Earth
DR vs Water attacks: 2
DR vs Earth attacks: 1
DR vs Fire attacks: 0

Fire
DR vs Earth attacks: 2
DR vs Fire attacks: 1
DR vs Water attacks: 0

Now if player 1 uses a Water defence and Player 2 attacks with a quick but weak fire attack, player 1 will take very little damage. But if player 2 used a slow but powerful fire attack, then player 1 would take more damage. Or if player 2 attacked with a quick Earth attack, player 1 would take a lot of damage.

If you combine this with the percentage system. Then different defence types have different effects.

A slow attack works better against a damage reduction type defence, but because it has a lower total damage against a target without damage reduction, it is not so good against a percentage type defence.

Quick attacks, on the other hand, are poor against damage reduction type defences as each hit is reduced and because there are more hits, the actually damage is lower. But, if the opponent is defend by a percentage type defence, then as the number of hits is no longer important and the total damage is, it means that these attacks become the preferred attack method.

As you already have 2 types of defences, this kind of system would be easy to implement in your game. Also, you have stated that there will be 3 levels of power for spells. By making the more powerful attack take longer, and the weaker attack take less time, you can create a system that takes advantage of Damage reduction and percentage based defences.

I like the idea of keeping the specializations somewhat secret, at least until used. But, that is going to be kinda hard to do if your game isn't online, or the other guy will be able to see you pick your specialization off the bat.

Oh and add to my list of move things: throws, the universal answer to excessive blockers! Can't believe I overlooked that one.

Being able to place shields around is a neat idea. In my aforementioned spellcasting fighter, all of the defensive abilities worked by generated fields of force around them to deflect, attract, or fling away incoming spells. And some of the offensive spells could do the same (the tornado was particularly cool). I want to add in something like the classic d&d Wall of Iron spell to my latest fighter and have it not only be able to be a physical barrier, but also to be able to crush people...

You can do other cool things to charged up spells instead of just doing a straight damage increase. I had things like: increased # of projectiles, increased projectile speed, bonus knockback power, increased mass (important for penetrating those shields), in addition to just plain damage addition.



The one ackward thing that I've always found in fighting games with any sort of projectiles, is the rather contrained aim that you generally have with them. This is more of an issue in a 3d fighter than a 2d (where you only really have to aim forward, up or down). I'm thinking some sort of point-and-click with the mouse might be the best approach, though that unfortunately makes hotseat play rather more difficult than just plugging in another controller. Though perhaps a quick thumbstick direction input during the casting could aim it (but I do prefer the dpad).

Another consideration to take in mind, is, what happens when 2 spells pass through each other/collide? Though this might be best handled on a case-by-case basis.

Edtharan brings up some more good points about numbers and balancing fast/slow attacks.

Quote:

Original post by andur I like the idea of keeping the specializations somewhat secret, at least until used. But, that is going to be kinda hard to do if your game isn't online, or the other guy will be able to see you pick your specialization off the bat.

With only three elements I may go with a single specialization, or abandon the idea altogether :(

Quote:

Oh and add to my list of move things: throws, the universal answer to excessive blockers! Can't believe I overlooked that one.

Judo techniques!

I'm getting all excited about a game that doesn't exist yet. LOL. I spent over 100 hours making a simple clone of the old NES Zelda II. (Granted much of that time was learning an engine and 'sprite programming' in general.

Quote:

You can do other cool things to charged up spells instead of just doing a straight damage increase. I had things like: increased # of projectiles, increased projectile speed, bonus knockback power, increased mass (important for penetrating those shields), in addition to just plain damage addition.

My "initial" idea was to have a power meter with a function something like pow = abs ( x * sin x) and then "optimal" spell casting occurred at one of the three peaks. This was back when I was considering having "form" in every technique including stab, and using mouse gestures to control Power VS Strength. As someone pointed out, interesting fighting systems are not necessarily fun.

Quote:

The one ackward thing that I've always found in fighting games with any sort of projectiles, is the rather contrained aim that you generally have with them. This is more of an issue in a 3d fighter than a 2d (where you only really have to aim forward, up or down). I'm thinking some sort of point-and-click with the mouse might be the best approach, though that unfortunately makes hotseat play rather more difficult than just plugging in another controller. Though perhaps a quick thumbstick direction input during the casting could aim it (but I do prefer the dpad).

I'm deliberately avoiding 2D or even isometrics until I know I can make a fun side scrolling game. So often I've seen gameplay sacrificed for graphics. (I'm sure that wasn't the intention, but that is what it felt like)

Quote:

Another consideration to take in mind, is, what happens when 2 spells pass through each other/collide? Though this might be best handled on a case-by-case basis.

Hmm . . . I don't think I'd have static spells (barriers) AND spell projectile collisions. HOWEVER, I may consider replacing barriers with collisions :)

Quote:

Edtharan brings up some more good points about numbers and balancing fast/slow attacks.

You both bring up good points and I appreciate your feedback.

There's nothing in your list I don't have an intuitive idea of how to implement. However, I could benefit from some more concrete/technical direction. Do you know where I could find a guide of how to implement your suggestions (or some of them) with helpful hints on how not to "screw it up"? (I think the "Not Screw Up" would come in the form a lessons learned, caveats or even FAQ) I remember when I went looking for the complete idiots guide to making fighting games, I came up empty handed. It has been awhile though.

I am not looking forward to latency and dropped packet issues when I implement two player over network mode! I pinged my friend in Australia and got a 221 ms ping time, and I was thinking that would probably translate into a mandatory 150 ms delay between pushing a button and the command executing probably even higher for comfort. I'm not sure if that is unplayable or not. If I have less than 1 volley between delays of actions then I will need a retroactive pause if the packet doesn't return. ugh :( LOL

I think my first network game will be pong ;)

Oh I never really explained my game and that led to some confusion. Sorry about that.

It is going to have a PvP in the general "Street Fighter" setup, and I will have slightly more elaborate boards available as well. I plan to use this heavily during development of my AI. (My AI for the Zelda II clone was fairly easy to write, to my surprise - but I never made it defend properly against casts.) Here I will make NPCs as well.

The main "game" will be an RPG in the Zelda II setup (side scrolling battle areas, and top view for everything else - unlike Zelda II the towns will also be top view) While I'll have all the tools in place to make the RPG itself two player, cooperative fighting in a side scroller would be difficult, I think.

My experience tells me that if I can build a standalone real time state machine engine and then have it interface with the communication and pause the game if necessary then it will greatly simplify development and testing. No matter how cut it this is going to be an adventure!

Quote:

Do you know where I could find a guide of how to implement your suggestions (or some of them) with helpful hints on how not to "screw it up"?

There are many articles on the Gamasutra website: http://www.gamasutra.com/

Here is a good article at gamasutra on Rock/Paper/Scissors (with a lean toward beat-em-up gmaes): http://www.gamasutra.com/view/feature/1733/rock_paper_scissors__a_method_for_.php

A good book is "Game Architecture and Design" by Andrew Rollings and Dave Morris : http://www.amazon.com/Game-Architecture-Design-Practices-Programming/dp/1576104257
Especially chapters 3 and 5. They take you through the theory and a bit of the maths as well.

Quote:

Judo techniques!

In the book Game Architecture and Design, the example they use to demonstrate the concept is one of a martial art fighting game.

They use the system:
Leg Sweep beats Forward Kick
Forward Kick beats Stomp
Stomp beats Leg Sweep

They also go into how you can change the ratios between them by introducing a cost to use them.

The Witcher tried to implement defensive moves in their bar-fight mini-game. It boiled down to:
1.Click Attack
2.Hold Defend button until enemy attacks
3.Repeat
Did you enjoy the mini-game?

Age of Conan has Active Shields, but I haven't used them, so I don't have an opinion on them. Did you, and if so what do you think of them?

I like the idea of defensive moves. Martial arts have them. But I'm not so sure about making a player choose between dealing damage and damage reduction, because the second option is generally just postponing an inevitable death. That may have to do with using hitpoints.

Moreover, can you really judge which attack, and execute the appropriate response in a fighting game? Combat button clicking is at a rate of around 4~8/sec, but Eye-Hand response to a simple go/nogo signal is about 1~2/sec. So decision making within real-time combat simulation isn't going to be fully interactive. Also, Retalin-enhanced button-press spamming is likely to just relegate winning to the domain of people with macro keyboards.

But if "defensive moves" were implemented similar to "fatality moves", it could be visually entertaining. For instance, if an attack failed in sword-vs-sword combat, play a parry sequence. The animation then becomes a scene involving two players, rather than just one.

[Edited by - AngleWyrm on August 28, 2008 6:07:05 AM]

Quote:

The Witcher tried to implement defensive moves in their bar-fight mini-game. It boiled down to: 1.Click Attack 2.Hold Defend button until enemy attacks 3.Repeat Did you enjoy the mini-game?

Exactly. There are only two choices and they don't form a Scissors/Paper/Rock relationship (Intransitive if you want the lingo).

If they had also implemented a "Break defensive move" action, like a throw, then the combat would have been more interesting.

Quote:

I like the idea of defensive moves. Martial arts have them. But I'm not so sure about making a player choose between dealing damage and damage reduction, because the second option is generally just postponing an inevitable death. That may have to do with using hitpoints.

The choice is not between dealing damage and damage reduction, it is between the different types of defences.

the idea is that you use defences to break the inertia of your opponent.

Just say you have an enemy that keeps flinging a quick set of fireballs at you. You are capable of moving out of the way and dodging most of them. However, while doing this you are unable to get close to them as it wouldn't give you enough time to react and dodge out of the way.

But, if you throw up a shield/defensive move and reduce the amount of damage (or eliminate it), then their attacks lose their momentum and you can charge in and make your attack without fear of taking a big hit.

Now they are on the defensive and you have the momentum.

If you can attack and be attacked at the same time, your analysis holds, but if you have to chose between attacking or defending, then momentum comes into it and breaking this momentum by using defensive moves and your analysis fails.

Quote:

Moreover, can you really judge which attack, and execute the appropriate response in a fighting game? Combat button clicking is at a rate of around 4~8/sec, but Eye-Hand response to a simple go/nogo signal is about 1~2/sec. So decision making within real-time combat simulation isn't going to be fully interactive. Also, Retalin-enhanced button-press spamming is likely to just relegate winning to the domain of people with macro keyboards.

This also comes down to momentum. A player will react in certain ways to an attack. A good player can exploit these reactions and set up a chain of attacks that force the player to react how they want.

An example with scissors paper rock is one where a player responds to one game by playing what beat them in the last game. A good player can spot this kind of behaviour and then make their choice so as to exploit it to win.

An example using the system described for this game could be that a player has a character with a strong power in fire attacks. They constantly use these (or use them more often) attacks. A good player could exploit this by assuming that their opponent will be using fire attacks. They place up a defensive barrier that is most effective against fire based attacks, even if it is not their most powerful defensive move.

They then know that their opponent will switch to their non fire based attacks, and so they quickly throw up a defense against this form (which might just be their most powerful defense). Their opponent now is at a disadvantage as the time it will now take the second player (who threw up the defences), can make an attack against them.

The second player exploited the reactions of the first player (that they would switch to an attack form that would defeat the first defense), and use it to swing the momentum of the fight.

Now to do this, the second player might have worked out a sequence of moves and then repeats them when they see the first player start to throw out those fireballs.

However, if the second player has this as a macro, then the first player can exploit this pattern and knowing throw a fireball at the second player, then throw another 2 fireballs at them knowing that they at first, the second player will be throwing up a defence that is not effective against fire, but then also know that the second player will be launching an attack and be completely open without any defences.

Randomly spamming button presses is not much better than predictability. If an opponent is just button mashing, then I can equal them simply by spamming the same thing at them. And, if the player has warning, they can actually defend against the attacks coming at them and so easily exceed the button masher.

Quote:

Original post by AngleWyrm The Witcher tried to implement defensive moves in their bar-fight mini-game. It boiled down to: 1.Click Attack 2.Hold Defend button until enemy attacks 3.Repeat Did you enjoy the mini-game?

ie. The combat system in Oblivion...

Quote:

Original post by Edtharan In the book Game Architecture and Design, the example they use to demonstrate the concept is one of a martial art fighting game. They use the system: Leg Sweep beats Forward Kick Forward Kick beats Stomp Stomp beats Leg Sweep They also go into how you can change the ratios between them by introducing a cost to use them.

The problem is, they do go into a lot of interesting depth here, but ultimately it just yields a system with a fairly trivial optimal strategy, even if that strategy is a mixed one.