Good Game Design

This is Part II of the Good Game series. It is actually a direct continuation of part I. If you haven’t read it, please do.

Game design is a complex topic, and one can buy shelves of books on game design theories (I personally own serveral books on game design theory). The thing is, there are many approaches to game design, some good, some bad. But ultimately, the game design has to center around the player, not the game. This is the Commandment for all game designers to follow.
This is where many games, Crysis included, goes wrong. The developers have been stressing so much about the open gameplay that they almost completely forgot about the gamer. If there’s one thing to know about humans, is that they form habits. This includes gamers in-game.

Point #1: Monotony in action kills
As I mentioned in my Crysis review, after some time, I developed a habit of dispatching the enemies, and soon enough the game became boring. I know that it’s an open game and you’re supposed to find different ways to dispatch your enemies, but humans settle upon habits quite quickly – once you found an efficient method that works for you, you tend to stick with it.

Where as in more linear games like Half-Life 2, the story guides you to do different things at different times, instead of doing the same thing many times over. Sure, you can’t walk out of Ravenholm (something I really wanted to do), but it’s well designed so the player doesn’t get bored. The player gets to crush zombies by means of cars, traps and saws.

So the point is: Monotony kills.

Yes, having an alien flying thingy suddenly falling on you because your AI team mates shot it down breaks the monotony, but also increases frustration because you died for no reason. Also, did anyone notice that the more interesting parts of Crysis (that is, towards the end) was actually quite linear?

If the player has to keep doing the same things over and over again, whether by linearity, or by force of habit, it is NOT a good game design.

Point #2: Players have learning curves.
This is the learning curve of Participant 1 in Portal (he was the only participant to have enough data points for a curve) -

I tried adjusting the time taken for the increasing difficulty of each level, but it doesn’t seem to be working too well. Anyways, I applied a power function for the curve, and the result is still an experience curve.

The learning curve in Portal is very well gradated. The game guides players to tougher and tougher situations without breaking the curve. This, is a good example of a learning curve implementation. On the other hand, we have Crysis, which expects you to be well adept at everything that is FPS related, driving included. Basically, Crysis throws you into the game expecting you to know everything already. This is a bad example of learning curves.

No matter how experienced a gamer is, the game developers must always implement a learning curve into the system. They can’t expect a player to just read the manual (which I’m sure most players will anyways, once they come to a point where they get stuck) and play. There must be allowance for the player to get used to the system.

Even if you compare very similar games like Quake III and Unreal Tournament 2003, where the difference is mainly cosmetics with the control scheme and everything else being exactly the same. Players still need to get used to the differences. Granted, being a fast-paced online FPS, the learning curve was designed (yes, those two games, especially UT2004 have what I’d call good learning curves) for fast absorbtion.

Basically, the game developers can and do in fact control the player’s learning speed. The better game developers like Valve and Blizzard and Bungie encourages the player to learn, while the poorer game developers like Crytek do not encourage learning, instead it forces the learning curve down your throat. That is not good. Most of the better developers also develop the player’s learning curve in tandem with the motivation to play the game.

Point #3: Motivation
Every player has a different motivation for playing the game. For an RTS player, his motivation in a single player campaign is usually the story. StarCraft and Warcraft III have very engaging storylines (and the fact that I’ve been playing Warcraft since its first incarnation helps cement the point). Command and Conquer: Generals didn’t have a very engaging storyline, and I didn’t really finish all the campaigns, though I did have a great time online with all three RTSes.

For an online game, the motivation is constant dominion. Because online play is played with other humans, it causes very dynamic scoring systems. One moment you might be king of the hill, and the next you’re at the bottom of the list when you press Tab. So the motivation is there for online games, hence I’ll not be talking about it.

But for an offline game, there is a constant need for motivating the player. As mentioned, I used complete newbies because I wanted to gauge the reactions. Players have very short attention span. The game must therefore motivate the player to continue.

There are few reviewers on the Internet (myself included when I do reviews) who use the 1-hour rule. That is, if the game manages to constantly attract one’s attention for over an hour, the game is worth playing. So it is important that the game keeps the player motivated.

One way to keep the players motivated is through the use of an incentive system – like unlocking some new (bikini) costumes for Kasumi in DOA. Failing that, there is also the use of a reward system – like in Portal, Half-Life 2 and the like, you have an Achievements section of the game, or like in the Medal of Honor series, where you actually win a Purple Heart. These are called extra-game motivation system. They reward a player outside the game.

A good example of an in-game reward system would be Diablo and its ilk. Diablo randomizes what ‘rewards’ the player gets when killing monsters or doing quests, which makes things more random, and the player would be more motivated in order to get more things. Also, MMORPGs have the same kind of reward system – kill monster, get gold, get rare item, sell rare item, get more gold.

In FPSes, stuff like Quad Damage act as motivators to the player as well. Getting a Quad Damage means you get to kill more with a single shot, increasing your incentives to shoot and score more.

Point #4: Good Playing Field Design/Level Design
What does de_dust (from Counter-Strike) and Big Game Hunters (from StarCraft) have in common? They’re the most played map in their respective games. Why? Simply because the maps are good. The designers of the maps have the player in their mind. Big Game Hunters conveniently places the player’s start point next to the resources, making resource gathering easy. With resource gathering at the back of their minds, players can focus on their offence instead. De_dust is a map suitable for a small number of players and a larger number of players (whence they can actually play tactically). Compare it with cs_assault, which frankly, isn’t very optimal as it is too large.

Also, does anyone remember a game called Oni? It was by Bungie studios and to this day, it sticks in my mind, because its levels were very very very well designed. I remember reading somewhere that Bungie got real architects to design that place, so stairwells and such are not in strangely inconvenient places, but rather, everything was intuitive and naturalistic. Oni won my respect for Bungie Studios as a good game developer, and as much as I love Half-Life 2, I say Valve could pick some of Bungie’s brains in terms of level design as there were some level designs in Half-Life 2 that made no sense.

A good level design unconsciously guides the player to their goals. Thus. it would be good, if the level designers studied graph theory, instead of just opening 3DS Max/Hammer/Unreal Map Editor/Crysis editor and just making a level. If every level designer designs like “Ooh, here looks like a good place to put crossroads to confuse the player”, then the game world will be very chaotic. Big game hunters and de_dust guide the players by means of choke points, gently forcing the player to take a different route around.

Of course, for some more naturalistic feel, level designers can randomize some non-critical part of the level (like adding crates, trees, burning tires…)

Point #5: Use Innovation Carefully
This article started out as Innovative Gameplays. I talked quite a bit on TimeShift, and Portal. Then I deleted most of everything and rephrased a lot of things. Anyways, you can’t find a modern game that is uninnovative from one another. There used to be a time (remember Quake III A and Unreal Tournament) where games didn’t try to outdo each other in innovation, but nowadays its all different. You can’t sell a game that is virtually identical to another game.

You have the gravity gun in Half-Life 2, the nanosuit in Crysis, the Aperture Science Handheld Portal Device in Portal, and the Beta suit in TimeShift. The RTS scene is increasingly moving towards the zero-sum approach of Rock Paper and Scissors format.

Point is, games nowadays are not short of innovation. However, games nowadays are short of good implementation of innovative features. If you’ve ever played TimeShift, you’ll know how bad the game is, despite it having relatively innovative features. Similarly with Crysis – good idea, bad implementation, except Crysis had an advantage – graphics, which made me finish the game.

Innovation mustn’t make a game boring. Both Crysis and TimeShift managed to bore me. TimeShift eventually put me into Max Payne mode the whole time. Habits form. Again, I say, the game must guide the players, if anything, it should guide the player gently into doing different things – like towards the end of Crysis, after the Core levels, or the whole of Half-Life 2 for that matter.

Coming up next: Avoiding Pitfalls