At my first job in the industry, working in the single-player department at the now defunct Stainless Steel Studios, we were strictly forbidden to create scenarios with anything even vaguely resembling defeat or loss in the story-lines. So, being the stubborn sort, after we finished our first game, I immediately sat down and made a scenario, for myself, that was all about defeat.

I chose the Spanish Civil War for a setting, pitting the player against Franco’s overwhelming invasion force. The goal was to let the player rewrite history, a glorious triumph against all odds. I placed fourteen major cities across Spain, each and every one of which, except the capitol of Madrid, would go over to Franco’s side when he marched against them. There was nothing you could do. You could not fight, because your troops would defect in the face of enemy fire. You would get the hint sooner or later, and pull the rest of your troops farther and farther back, until only Madrid was left. A siege was announced, and the player was then told to hold out for so many minutes until European reinforcements arrived to help your cause – and you could go on to kick some serious Fascist butt.

My goal was to take the player’s expectations for a strategy scenario, dash them, and then make it absolutely clear that they were utterly doomed, before showing them a single glimmer of hope. I’d like to flatter myself that the scenario was moderately successful, both fun and dramatic (it did fare well on the fan mod sites). At any rate, I have never stopped being proud of engineering that retreat – of capturing a little whiff of the tragic moment in a game, in order to make the subsequent victory that much more glorious.

It’s easy to understand many designers’ knee-jerk reactions against putting dramatic defeat situations in their games. It’s not very hard to get such setbacks muddled with player failure. The line is pretty darn thin. The iron rule is that if a player so much as suspects that they could have avoided a bad turn of events, or that it was a fault in their performance as a player, you have not only pissed them off, but you have killed the drama you had worked so hard to build.

Books and movies have a huge advantage in not incurring regret in their audience. Their “players” have no agency; as much as they may dislike a twist in the plot, it’s not their fault. As game designers, we must reckon with regret. Our players have to do more than like the story; they have to accept each turn of events and roll with them, and never wonder if they should have gone back to get it right.

At Stainless, we steered extremely wide of these dangers. “What’s the fun in losing?” and “The player should feel successful at all times” were our ruling directives. But this reasonable precaution comes at such a price! Without real, down-and-dirty reversals, you could never have Lando Calrissian turn on his old friend; you could never have Edna Mode, in the Incredibles, turn to Mrs. Incredible and say, ominously, “Do you know where your husband is?”; and you could never, as in Braveheart, after the battlefield betrayal by the Scots nobles, have the man at the English King’s side raise his helmet – revealing none other than Robert the Bruce. (Not to mention having Boromir try to take the ring from Frodo!) Setbacks provide some of the most intense and moving moments that humans have learned to produce in storytelling. They are a cardinal station on Joseph Campbell’s mythic wheel, indispensable to the patented Hollywood movie formula, and the spur that is required for every hero to rise to their greatest deeds.

Fable by Lionhead Studios

Games are no exception. Some of the most memorable moments in games depend heavily on reversals to kick their dramatic arcs forward, from Planetfall to Fable to Beyond Good & Evil to Deus Ex. And yet, as an industry, we clearly have a lot to learn – and a lot to invent. So, then, how do you draw a clear line between player failure and dramatic reversal? It is a question well worth pondering. And, I contend, it is one that we can begin to answer.

The earliest, simplest method of creating dramatic setbacks in games would be the cut-scene ex machina. One of my earliest memories of this venerable technique is in short clips between levels in the very first Ninja Gaiden, but the same effective ploy can be found in the likes of Diablo II, not a few Final Fantasies, and to wonderful effect in Grim Fandango. It’s safe. You are not likely to think you failed in a scene you had zero control over, especially as they tend to take the form of rewards for completing a section of the game.

While there is nothing wrong with this device, we have, as an industry, been moving away from cut-scenes, or at least cutting back on them. There seems to be a general consensus that cut-scenes are fine, but placing your dramatic moments in-game is far better – less disjunctive, more immersive, and so much more powerful. And power is exactly what we’re after.

Of course, this brings us back to the problem that moving these reversals into the gameplay is not a simple affair. Under the hood of every game is the same simple mechanism: we give the player a set of skills to master, and then run them through the paces, demanding effective performance of that skill-set. To know that they did well, the player requires feedback. If the player thinks they screwed up, but the game doesn’t send a clear message to that effect, they will understandably be frustrated and put off. Not to mention that the drama will not be read as drama. This means that player’s actions – and their results – are not good contexts for setbacks or reversals.

Of course, a setback that relates to gameplay is a good thing. In a game that involves a lot of jumping, making the player ‘unfairly’ fail a jump would be bad; but blocking the player’s path of egress by causing pillars crumble behind them is good. In fact, it’s got class. Other things under the player’s control, possession, or purview are also generally off limits: power-ups, special skills, equipment, accumulated points of any kind. Don’t take them away and don’t stop them from working, not for the purposes of the story. As with all things creative, there are surely ways to successfully break these rules – but generally speaking, it’s not a good idea.

Setbacks that call on the player to use their skills can be excellent. This is especially true when the challenge comes as an objective. Objectives, after all, usually function both as gameplay goals and as points in the game’s plot. This dual nature gives objectives and quests great potential, but often plagues them with an array of different awkwardnesses.

Depending on a game’s genre and the rigidity of its objective system, there is a wide range for what you can get away with when it comes to dramatic reversals, or, in fact, storytelling in general. Generally speaking, a setback in the plot at the start or in the middle of a mission can be very effective. It helps create the miniature plot arc that should comprise every mission. Introducing a reversal right at the end of an objective runs the risk of invalidating an objective. Say the player is sent to rescue someone. If they are not where the player was told to find them, what does that do to the player’s sense of accomplishment?

While you can massage the phrasing of the objective (e.g. “The Feds are after Jane. Look for her in the old barn to warn her”), or push the reversal off to the beginning of the next objective, not every case is so easily fixed. Even a reversal near to the end of a mission can get thorny, especially in the middle of a boss fight (or the equivalent). In fact, any story event in the pitch of action is a problem, because it is competing with the gameplay for the player’s attention.

Which brings up an interesting point. Curiously, the most vivid and memorable scene in the opening of Half-Life 2 is the one that cripples the player, dashes their goals, and generally messes with the player’s ability to properly play the game. I am speaking, of course, of the apartment building scene, where the police raid the building, and chaos erupts, sending you and the building’s tenants scrambling for safety and escape. You run and you run, but eventually they surround you. There is no way out. The scary masked guards begin to beat you, you fall, the screen blacks out… and then you hear what might just be some form of rescue.

When the screen fades in again, the guards are lying senseless around you, apparently the work of Alyx, a young woman who introduces herself and then leads you away to safety. Whew! This entire sequence is a real tour de force, and it seems to break most of the rules laid out above, not to mention many more that are standard to our trade. What’s going on? The simple answer, of course, is that somehow they’ve convinced me (and, I wonder, everyone else as well?) that I wasn’t screwing up, that this is how things were supposed to go.

Come to think of it, the entire game up until that point is rather odd. You don’t really have objectives. It’s as if there were something entirely different going on instead of traditional, objective-based gameplay. As you play through it, these subtle cues tip you off, and it makes this impressive dramatic reversal something that you, the gamer, are willing to swallow.

Valve Software's Half-Life 2

This scene in Valve’s recent masterpiece is the best example I know of, illustrating something new that’s creeping into our industry: moments and passages where a different set of dramatical rules guide the play experience. Just like the opening of my in-comparison pathetic Spanish Civil War scenario. Or like nearly the entirety of The Indigo Prophecy (Fahrenheit, to Europeans), which may or may not be a game at all, but instead, one extended interactive dramatic sequence.

I believe there are many more examples of what I’m talking about here, all testaments to designers whose creative intuition won out over common sense. Of old, we have divided games into core gameplay and cut-scenes, into action and story, into interactive sequences and passive sequences. There is a way to break out of this paradigm, but it’s not going to be easy. We can make games where you play the game, and you play the story – and when we do this successfully, the result will be unequivocally powerful, deeply moving, and truly inspiring.

What is needed is the creation of a language of game drama, in the academic sense: a set of established conventions that will allow the player to read a setback as good storytelling, and not a slip-up on their part. A grammar of subtle cues to create the distinction. In any game with a story, where the player has sufficiently broad control of their character, those controls can be used to participate in story events as well as in gameplay. You can run madly to help a friend on the battlefield, only to arrive at the moment of his death; or desperately pull your troops back the moment you recognize that it’s an ambush; or struggle vainly as magical tendrils grab at your ankles and wrists, resulting in inevitable capture. In each of these examples, the player could read the failure as a skill failure or a dramatic twist. How are they going to know the difference? The answer is a coherent, consistent grammar. The cues are ours to invent. A change in camera angle. A musical motif and a change in the lighting. Excessive motion blur. Voice over, either inside the character’s head or from a narrator. A gentle slowing down of time.

If you don’t believe in the power of this sort of convention, go back and watch a few movies or old television shows from, say, the seventies or earlier. My favorite example: telephone conversations. You used to find what I can only imagine were nervous film makers, who weren’t confident that their audience could follow both sides of the phone call. A diagonal line would divide the screen in half, showing each person holding the phone. Now go watch The Departed, with cell phone conversations all over the place, as natural as two people talking face to face. It is very likely that those early film-makers needed to hit their audience over the head when it came to phone conversations. We 21st century moderns are very well trained.

I would argue that it is just the same with gamers today. Not only do we need to invent a language of game drama, but we need to teach gamers to speak it fluently. We’re not talking about a revolution overnight. The language wants time to develop. First, players must be weaned off of the constant positive reinforcement of classic gaming. We must start with relatively minor setbacks, or a few big, memorable ones, until gamers are acclimated, thus opening the door to more nuanced and authentic dramatic experience. The current crop of gamers is impatient of stories and their trappings, because our stories are mostly bad and we tell them awkwardly. But when they get a few more tastes of what games can be, they’ll be back. They will be back, and they will beg us for more.