The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.
When it comes to ensuring that your game will make money, the conventional wisdom is that you are unsavory at best if you dispassionately incorporate elements primarily based on whether they’ve made other games profitable. The community tells you to go by your internal sense of ethics and your personal expertise as a professional, and not place too much faith in the objective measure of what customers pay the most money for because that measurement is so easily manipulated. The methods of hijacking human psychology are familiar, proven, and more widely understood by companies like King and EA every day.
Yet when it comes to designing what players can do within the game rather than how they can spend money on it, we find a parallel situation with opposite advice attached. Here, objective playtesting and player observation are lionized. The designer who produces a play mechanic or a control scheme or a tutorial level purely from intuition invariably creates a bad game. What you think players will like, want, understand, or do is invariably less accurate than what they say they will, which is, in turn, notoriously off the mark from how they actually act.
In other words, when designing a game, it’s smartest by far to base your decisions primarily on what players do and less on what they say they want or what you think is best, but when it comes to selling your game, following that exact plan — basing marketing and monetization decisions primarily off what players do pay for and less on what they say they want or what you think is best — is despicable.
I’m not sure why this is. Possibly it’s because human purchasing behavior is more susceptible to influence than play behavior is. At least, that seems to be a common first-order explanation. But consider:
- In designing Canabalt, Adam Saltsman deliberately allowed the player to jump for a fraction of a second after running off the ledge of a building, when he’s visibly in empty air, as well as jump cleanly over obstacles even if they’re so close that the character clips through the corner on the way up. Saltsman did this as a result of so many test players mistiming their jumps and getting frustrated at the game for being oversensitive and unforgiving.
- SounDodger and other bullet hell games have a convention where only the center point of your ship is vulnerable to collision, which leads to the satisfying feeling that you’re squeaking past enemies by the skin of your teeth when, in actuality, they did touch you, just not enough to count.
- City of Heroes and some installments of the Civilization series had custom random number generators to avoid outcomes that were mathematically valid but felt unintuitive or unfair to people, such as attempting something with a 66% success chance and failing twice in a row.
The common thread behind all these kludges is that players find games much more enjoyable when they blame themselves, not the game, when they fail or lose or die. Tweaks like the above are all accommodations to avoid the player feeling like the game is buggy, arbitrary, unpredictable, or unfairly difficult. Avoiding that has been a critical part of successful video games for as long as they’ve existed, something even the famed Shigeru Miyamoto recognized from studying Pac-Man and concluding that its success lay in convincing the player that she could have made it further if she’d only played better.
But the situations aren’t totally equivalent. Miyamoto increased Donkey Kong’s fairness by giving the player more bona fide power: he added jumping over barrels, which were otherwise frustratingly inescapable. In contrast, the games mentioned above accommodate the player by misrepresenting the way the game world works. They coddle. They fudge. They lie. And all because it playtested better.
So maybe designing games to be the most fun is, in actuality, precisely as unethical as selling them to make the most money. Maybe they’re both laudable. Maybe they’re both reprehensible. Both certainly can be more concerned with being believable than with being honest.
Or, perhaps, it’s not what you do, but why. Adding compulsion loops and pay-real-money-to-continues is selfish. It helps you. Allowing the player to jump for 8 more pixels after he runs off a ledge is magnanimous. It helps him. Friends of mine have made the analogy to a stage magician. His entire job is to fool his audience, yet they attend willingly and even enjoy the misdirection, and no one can legitimately claim that the nature of the magician’s profession makes him inherently unethical — unlike, say, a pickpocket, who employs many of the same techniques.
In any event, in the back of my mind now lies the specter that good playtest results are just as likely to mis-indicate that I’ve made something legitimately fun as high sales figures are to lead a producer to believe the product has been made better, not just more profitable. It’s not a feeling I enjoy.