Wednesday, June 10, 2009

Screw the Rules, I Have Good Design

This is a super nerdy article totally focused on making games. I've been irritated lately by things that follow this mold, so it's only natural that I mention it. If you're not a super nerd planning on making your own game someday, you can probably skip this. It has some value for GMing though, since a lot of GMs make house rules.

One of the thing that annoys me in game design is when people insist on making things a certain way because they need to conform to existing, previous design decisions.

I don't quite think I can explain this very well, so I'll use an example to illustrate.

Waaaay back in the 90s there was this popular fighting game called Mortal Kombat. A lot of you guys have probably heard of it! The characters in MK were very similar. They all had the same normal moves with the same hitboxes, and they all had the same number of special moves. One characters had an exception (Johnny Cage's groin punch). This game was really, really not that fun design. Most of it was about playing the best character (probably Sub-Zero; Scorpion had an anti-projectile teleport though, making it not so clear-cut) who had slightly better versions of the same watered-down moves.

Mortal Kombat never really strayed from this design much. Characters continued to have very similar basic attacks and fairly similar special moves. Although different characters became the best, it was still clearly in favor of characters that had broken gameplay elements or slightly better versions of the same boring special moves.

By contrast, around the same time, there was Street Fighter 2: Championship Edition (CE for short). CE was an older, more busted/abusive version of SF compared to the newer versions of the game (HF, Super SF2, ST, and HDR) but it was much more refined than Street Fighter 2. The moves in CE were very different from each other. Ken's low medium kick was vastly different than Sagat's in both reach, power, and utility. Projectiles in the game fired at varying speeds, recovered at different times, and had very different amounts of hit stun and damage. Special moves had radically different properties as a whole - Bison's torpedo (psycho crusher) behaved much differently than Honda's torpedo (sumo headbutt) despite being similar moves.

SF2 would definitely be the harder game to balance. It's had a lot of iterations whereas each Mortal Kombat game, with the exception of 3, has had only one (3 had a version called Ultimate Mortal Kombat, with a lot of new characters and the same basic gameplay). While MK was never really ever balanced at all (MK vs. DC sure isn't) it would have taken the developers much more effort. Why didn't they?

The problem really is rules. Moves in the old MK universe (not sure if this is entirely true but I suspect it still is) follow a certain design guideline. Moves deal x amount of damage if they're a certain kind of hit, and put you in a certain type of stun. Moves that knock you up or down or launch you around all launch you in a certain way. Rather than fine-tune these values, the MK designers set up a bunch of guidelines for how they'd make abilities, and just sat down and made a bunch of skills and threw them into the game. This is from what I have seen still the case even in the more recent MK games, although there's a lot more types of attack now so there is more diversity.

This isn't really for fighting game design though. It's far more of a pen and paper or MMORPG type of design. D&D 4th suffers from it a little bit, and Returners FFRPG has a lot of it there. These systems are good uses of rules to create lots of cool things, but Returners in particular is kind of bad about having not enough flexibility (particularly in the equipment area).

There's another type of design that totally throws rules out the window. In fact, this is pretty much the entire design behind SF2CE and most SF games in general. This design says, design a lot of broken stuff, make your characters really unique, and hand the broken stuff out, a few to each character. This is also the method WoW uses, notably. After you've totally thrown your game on its head you can then adjust things.

Most games use this method, and it is also bad. The main reason it is bad is that you can very easily create metagame shifts - if you give some class something you don't think is that bad and it then becomes overpowered, many people will switch to playing that class and abusing the tactic you gave them. Even worse, when you fix the imbalance (nerf it!) people get pissed.

The only way this design strategy is even sort of good is if your players can easily adjust to changes. A good example of doing this is Guild Wars. When skills get changed, GW players can easily change out their skills for different ones and there are numerous quality build options for any given class in both PvE and PvP.

A bad example of this is World of Warcraft. Classes are forced to take skills (or be less effective) so nerfing them is generally bad stuff. For instance, if one were to nerf a rogue's crippling poison so that it was less effective, the rogue could not change that ability out. While there is some ability to change talents, any change to a class can resonate strongly. Also, talent changes cost the player valuable gold and are quite expensive, especially for lower level players who are more likely to make mistakes.

A really bad example of this is Ragnarok Online. In RO, if you mess up your character's build with an errant misclick, it costs you real money to fix. Prior to the introduction of the cash shop, which lets you spend real money for things in the game, you could not fix these errors at all. One misclick meant you'd need to delete your character and make a new one. If the developers ever change the way the game is played by altering skills or adding new equipment that forces you to re-stat your character, you have to pay real money to fix the problem or delete your character.

The moral of this story is really to never play RO.

So there's a third way to design, and it involves rules but only sort of. It's way harder to start but way easier to balance. That method is using a design skeleton. It's a made-up-ish term and I feel sort of nerdy using it. Oh well.

What I mean by design skeleton is that you take a base character and give him some things he can do no matter what. This stuff should be common to everyone - some classes or characters might not have these traits in favor of powerful other tricks but the basic character skeleton is shared by all characters as a default.

To illustrate this somewhat I'm going to bring up a game that no one actually played, Tabula Rasa. In TR every character could use a ranged weapon with a certain base amount of skill, so everyone could defend themselves. They could also melee with their ranged weapon too. These abilities were NOT sub-standard. I made a support character (healer type) and I had a significant number of weapon switches, including shotguns and rifles. I did not even use my class specific weapon and instead relied on basic guns as backup to my class powers.

Even better in my opinion is that every class in that game had some access to self heals in the form of repair kits and modules to repair damaged armor. The healer classes were way better at it and could do many things like area effect repairs and large single target repairs, but everyone could recover from lots of damage if they took cover and used repair kits.

This did not exclude my healer class or the combat classes. My friends played assault classes, and I was able to greatly aid them with my repair packs and 'magic' spells, which were better than the assault characters' repairs and magic, and let them focus on shooting things with their assault class guns. Still, there were times when my repair kits weren't needed and my mana was low, and I happily pulled out my rifle or shotgun and did some blasty.

This design is absolutely awesome. I especially liked the fact that assault characters could aid themselves if I couldn't get to them, either due to reloading (med/armor/shield packs took a long time to reload, so if you ran out of uses in the pack you'd have to wait a few seconds) or because I wasn't close enough. The main failings of TR was a bad beta filled with bugs (which led to bad PR), and characters that weren't unique enough. The end game classes were really sexy, though.

With a game design like TR's where baseline characters are able to function, you have more wiggle room for broken design elements. I would take it one step further, though.

Another really good example of design skeleton is hidden in World of Warcraft. That element is the "PvP trinket." Once per couple of minutes, any character can use this trinket to get out of a control effect for free, and it protects them from control effects for a very brief period. Every character has equal access to this trinket for very minimal effort. I know it's one of the cheapest items to buy with honor points, although I got mine for getting ranking (which was also very easy).

This trinket lets people break free of chain fears or stuns for a moment and gives them a chance to turn a battle around. It doesn't give them a complete pass, since it lasts only a few moments and has a long cooldown. This means you've got to use it intelligently, but it means that if you fight against a complete stun lockdown or fear/charm chain, you've got a few seconds to break free and stop them from keeping the lockdown going. This means you've always got a shot in a fight - a rogue can't ambush you from behind and stun you forever with no way to fight back until you die (It's worth noting that rogues in particular are kind of good at re-stunning you immediately after it runs out, but warlocks can have a hard time dealing with the trinket).

Other good examples include universal mobility options. These options need to have counters, but things like dashes, flying, multiple jumps, or whatever are pretty good at forcing the fight to go in different directions. A player should always have the avenue of escape if it's a fight they can't win. On the flip side, these options should not be entirely uncounterable. A good example might be a stamina meter so that dashes cannot be used to just flee forever. Another good example might be snares/movement slow that some characters can employ to keep others from fleeing.

Within this design structure you do need to look at all the pieces though and design counters around them. For instance, if you have absolute methods of escape such as teleportation, you need to design a counter to that, such as a class that can lay down anti-teleport fields or has a debuff that cancels teleporting skills. To that end, you also need to be able to counter those elements, so the anti-teleport field needs to be small enough to be escapable, or needs to be able to be destroyed. The teleport debuff should be able to be removed by friendly players, and so on. This creates strategy elements that enhance the game, rather than subtract from it.

Another important thing within the design skeleton is giving people abilities which obsolete elements of the design skeleton. If you give everyone a dash for escaping, some characters might get a better dash ability or a flight or double jump. These characters probably won't use their dash very much in favor of using their other skills. Another better example is the repair kits one, where normal characters can use repair kits to heal themselves, but healers can do a much better job on both other characters and themselves. These things are perfectly normal and natural elements, and should be encouraged.

If you make characters decent without needing special powers, you'll find people that will experiment a lot more with those special powers. This is good, fun, and enhances your game. And when you make changes, people won't feel as much like their character is useless, because even without their special powers at 100%, characters can still contribute.

This does not preclude the use of easy 're-skill' options that let players fix their characters. Always, always let players fix their mistakes.

No comments:

Post a Comment