Consistency is one of the simplest and yet most crucial aspect of game development. Alas, it is also one of the easiest to neglect and is often a casualty of the developer/publisher relationship. But there has never been one single case of successful, inconsistent game in history. That alone is worth spending a few minutes understanding what this is all about.

Consistency means that every aspect of the game, background, key features of gameplay, art direction and technical orientations, are all focused and coherent one with another. This is important because:
 a consistent game is a better product,
 incoherencies may harm the relationship between the player and the game,
 and because the development of a well-defined product is less risky.

A better product

In most (all?) development shops of the world, there is a latent antagonism between marketing folks and developers. The former are often accused of not being genuine gamers and not understanding what it is all about. The latter are said to be uncooperative, and to not understand what it is all about. Developers are probably the most defensive, and many will certainly find the idea of creating a good product absurd, as if profit was the only important aspect of gamemaking.

GTA: San Andreas, PC, 2005

Walking around with machine guns.

But a good product is not just a game that could sell well. Mostly, it is a game that is legible, easy to understand, comes quickly to mind, easy to discuss. If a game is a good product, chances are that it responds to a real need of the public. And if what if offers is easy to understand, it may become, in the eyes of a public, the best answer to that need. If this is well exploited, this may lead to commercial success, but in any rate, this will bring recognition to the developers.

So what is a good product?

Zelda: A Link to the Past, SNES, 1991

One of the many secret passages in the game

Nomad Soul, PC, 1999

The "adventure" mode.

first, it’s a game that is easy to describe in a few words. Like, "the story of a young gangster with complete freedom of action in a state not unlike California" captures GTA San Andreas, or "a game where the hero must both defeat monsters, action-game style, and solve intricate puzzles in a fantasy world" accurately describes Zelda. Suppose that you don’t know any of these two fine games: those words give you enough to imagine what the game they describe could look like. If you like whacking monsters with a sword, you feel you won’t be disappointed by Zelda, and if you like shooting people, you can pick a copy of GTA with confidence.

GTA: San Andreas, PC, 2005

Another day on the freeway...

A good counter-example is Omikron - the Nomad Soul, which describes itself as "a game that brings the best of the adventure, FPS and fighting genres together in a science-fiction world". We’re not advanced. That’s where consistency is needed.

Second, a good product has a strong evocative potential. Games like GTA and Zelda are open doors on a powerful universe. . Because these games are very focused (and also very well done) not only do they immediately come to mind when thinking of those themes, but just their names brings to mind so much about those universes.

A good product is not a game stripped of its soul. In fact, best-sellers are, more often than not, good games. But the reverse is even truer: good games are all good products (whether they sell well or not is beyond the scope of game production proper, but at least they can). Take any game from the Gamefaq’s top 10 games of all times, and you’ll find that they are easy to fully describe in a nutshell, and that they all have a strong evocative potential.

Stronger relationship with the gamer

If you have read my other articles, you will know that this is the aspect of game development that I find most important. If a game can reach the player and grab their attention, then your job is done! This is what gives a game its right to exist and to live through in the collective memory.

We’ve seen that inconsistency makes poor products. That happens before the player decides to give the game a chance, so that’s quite important. But now, let’s see what happens next, once a player has decided to get the game and play it.

Getting into the game requires some effort. The player goes through a phase of learning - the game settings, the characters, the game mechanics, the game goals, etc. All that learning effort has a scientific name: the cognitive cost. The more effort is needed, the higher the cost. And in what currency is this cost expressed? In attention, of course. In other terms: any learning effort the player has to deliver are a strong obstacle to the ease with which they can enter the game.

This is something of which developers are generally aware. Lengthy, wordy manuals have disappeared from the game boxes, replaced by simple leaflets and fancy interactive tutorials. Difficulty has also been toned down. Developers, under the gentle pressure of console manufacturers, also are increasingly better at adopting game conventions, so that players can switch with ease between two similar games. But these changes only address the problem in surface, by removing formal obstacles.

Even a very playable game of great technical quality and which adheres to the most up-to-date gaming conventions will be very difficult to enter if it is not consistent!

Nomad Soul, PC, 1999

One of the FPS sections... why was it mandatory?

Vagrant Story, Playstation, 1999

Not sure that’s what he meant in the original version!

Ico, PS2, 2001

That’s where you start. No one will hear you scream.

Ico, PS2, 2001

The rooms are huge...

Ico, PS2, 2001

That intense light makes you feel like a helpless bug

Ico, PS2, 2001

A typical outside scene

Ico, PS2, 2001

Together, you feel stronger.

Some think that the game’s mission is to achieve "suspension of disbelief", prevent gamers to realize that after all, those gorgeous graphics and music and animations and what not are just bits and bytes. The fact is, if the player doesn’t believe in the game illusion, this comes as a strong attention shock. All game incoherencies will go that way. Whenever the player would think "it’s not logical", "this doesn’t make sense", "why do I have to do it" or similar questions, they are not using their brain to believe in the game, but rather to escape the game. If the game design is revised in the middle of the production phase, for instance, it will show! Again, the Nomad Soul was a splendid game when it was released in 1999. But having to switch from adventure to tekken-like combat to FPS mode without a fair justification were just as many occasions to wonder, "why do I have to do this now?" and to get away from the game! Games with a complex story line also suffer from this. In the Western versions of Vagrant Story, one of the few games to ever have been honored with a perfect 40/40 by Famitsu, the translated dialogue text was much longer than its kanji equivalent. Rather than finding a better way to display text, it was decided to chop the end of most dialogues. The result is a storyline that doesn’t make sense, which hurts the game eventually. Finally, games with an art style that doesn’t fit their story tightly, which is unfortunately the case of the majority of modern games, lose a way to get the player involved.

Conversely, games with a strong consistency have a much easier time to get the player into the game. Take Ico for instance. It may be difficult to find one single thing wrong in the design of Ico, but what really makes the game stand out is its consistency. Everything fits well together. The simple story is adequately conveyed in the short intro, but it really starts to make sense in the first gameplay minute. As you try out the controls, the hero yells for help in an immensely empty room, yet no one answers his plea. Each gameplay moment reminds the player of the same feeling: Ico and Yorda have been unfairly sentenced to death, but together, they will escape from the castle. The art style of the castle is extremely coherent, and it serves the story well: each room feels as if it were two miles long, in each outside scene, the light is overwhelming, almost blinding, reinforcing the feeling of loneliness and vulnerability. The music is subtle and quiet, so when the heroes shout they can be heard distinctly. The gameplay also carries that combination of vulnerability and hope. All "big" jumps look impossible to perform - so they are all leaps of faith! in combat, Ico is always outnumbered by indestructible opponents, but never killed. Miraculously, he manages to make them flee! Because the interaction between the heroes is important, so it is in the controls: the player has to keep a button pressed during most of the game to "actively" grab Yorda’s hand (else, she’s on her own and ends up being attacked by dark spirits). Developers probably thought about giving Ico or Yorda stronger abilities, so that they could "pwn" the monsters through their trip. How about a crossbow for Ico or some offensive magic for Yorda? That would have made sense from the outside of the game, but it would have strongly disturbed its core principle: "2 weak kids against the odds". Likewise, an uninspired producer could have required the development to add some architectural diversity. Again, that would have made sense in itself, but would have much lessened the message of the game.

Better, safer development

By then, we know that consistent games are both better products - which have stronger commercial potential, and better games - which are better at drawing in the gamer. Those are two good reasons to believe that publishers and developers alike should focus on achieving, and protecting through the development process, that precious consistency.

Yet, anyone who lived that will tell you that it’s very seldom the case. Why?

Consistency is never what publishers or developers are after in the first place.
 Publishers want titles that sell well, which means title that marketing people find easy to sell, in other terms games with high visibility and with several unique selling points (USPs).
 Developers want peer recognition, which they believe is attained through technical achievements, like in the old days of the demo scene. They also appreciate creative freedom.
 For this and other reasons, their views on gaming are not always compatible, and most titles are subject to debate.

Ideally, when a developer shows a prototype to a publisher, it is consistent and in tune with a genuine market need. More often than not, the publisher will want the developer to add features which are announced in a competing game, or make editorial changes to the storyline, etc. Theses changes are not bad in themselves - the publisher usually has a better vision of the market. At any rate, the pre-production phase is here to experiment different hypotheses. But by the time this phase ends, both parties should agree on what’s important in the game, and what is less important, as this has strong implications on the development priorities and production orientation.

For instance, if both agree that the title is a fast-paced action game, they could postulate that no matter what, the frame rate must never go below 60 FPS, that the main character should be very detailed, the NPC less so, the levels must be spacious and not overcrowded. If the title is rather an adventure game with hints of action, then the frame rate may fluctuate, but much more attention will be spent on the levels and backgrounds, perhaps less so on the animations of the hero. Such high-level decisions will have far-reaching effects on the game development.

If it becomes clear that the game cannot be consistent, focused around simple but strong propositions, which the publisher should know before the end of the pre-production phase, the best to do is to pull the plug and stop the project. Stopping a project in pre-prod is not a failure: it’s always a learning experiment for both parties, the opportunity to explore new ideas and play with new technology. Publishing a poor game, however, is not an accomplishment. Unfortunately, it is very difficult to estimate whether a game is good enough to be published early enough.

The worst that could happen, however, would be to request to add or substract features of the game which would directly threaten its integrity. While making minor changes to get the game more in tune with the times (such as integrating a new special effect, improving the controls or the interface) only pose project management issues (time, resources, priorities), changes that require the game to change its nature, to depart from what was agreed in the pre-production phase, will have much more severe consequences.