February 18, 2011
It happens on a walk home after a long day’s work. I’m listening to an episode of Irrational Interviews as I pass through the exit of the university car park: in particular, to an episode with Brian Michael Bendis, comic writer extraordinaire. A short conversation within the episode barely latches onto my mind. I furiously rewind the podcast as memories of a recent conversation come flooding back.
“No-One Thinks That Spiderman Is Really Gonna Die”
Brian Michael Bendis
Cut to a week earlier, and I’m sitting in a room with four other students, each of us from various arms of IT. Tension drips off me like the invariable sweat of the 35 degree* day that lurks outside. It is the second meeting of the new year for the Sydney Interactions Forum, and I’m anxious to say the least.
*Degrees Celcius, for our American friends
Lets back up a bit. The Interactions Forum is a group I formed near the end of last year, with the sole purpose of increasing the conversation that was occurring on the topic of computer games, their design and production. While groups (notably the IGDA’s Sydney chapter) exist to support the game development community, I wanted to start a group that, both, met more frequently and which would encourage people in my particular university (who are notably absent from such community events, despite a fair number of students being interested in games) to start actively developing.
My nervousness came down to two key points: that the group had nearly vanished last year, taking quite a few members with it; and that we were trying a new format, whose success would partially determine future attendance and thus the fate of the group.
In reality, the danger of such a result was minimal, but in my mind, the stakes were high. And the topic of the night was stakes.
“(We’re Looking For) Stakes Outside Of The Fight”
Ken Levine
Those of you who follow this blog know that I have written about stakes a couple of times: those who’ve talked to me about games will know that its a subject which I find both intriguing and perplexing. In summary: Save and Load mechanics are vital for gameplay learning as players must be allowed to fail and try again. Yet because death is the consequence of failure, this learning model cheapens it, to the point where life and death stakes are arbitrary and meaningless. With this in mind, the question we posed for our second Interactions Forum was “Can we have failure in games without player death?”.
As I finish explaining the concepts, I pause. I wait.
And then, with a rush, the discussion comes.
Braid is the first candidate for consideration: with its ‘rewind rather than reload’ approach, is death really the stake of the game? In short, yes: failure to step on a goomba leads to Tim’s death, even if it is temporary. However, Braid’s highest stakes are more grounded in puzzle solving: will you ever find out what actually happened between Tim and the Princess? Interestingly, this reality is echoed in the puzzle-solving gameplay: failure delays the player’s acquisition of (quite literal) puzzle pieces of the story, sometimes to the point of having to restart the current level completely.
Yet this, we later determined, was a factor of many puzzle games: failure simply means a delay of victory or discovery. Braid’s innovation is that its story and stakes echo its consequences of failure so damned well (as opposed to Tetris, where the biggest stake is often how long you can procrastinate before you die and have to consider starting actual work again). So while puzzles forego death in failure, they often do so by replacing a dilution of meaning with a complete lack of it. This is not quite the answer we were after.
Neither are the sometimes effective but ultimately antithetical consequences of failure that Mass Effect’s sprawling dialogue tree have. These range from the emotional to the life-and-death, all of which affect the game permanently, and most of which affect NPCs rather than the player. The power of these is that you can unintentionally kill a friend suddenly and permanently; the antithesis is that these people are constantly dying and coming back to life in the main, combat-based, gameplay mode. Death is cheapened and yet important in an incredibly arbitrary way which speaks to the limitations of design more than the intentions of the creators.
“You Can’t Have Spiderman Die, But… You Can Have People Around Him Die”
Brian Michael Bendis
Perhaps the answer lies in a completely different kind of game, like Uplink, where if you fail, you simply have to play more game. Failure in Uplink means you have to cover your tracks to stop yourself from being traced: its a slightly different modal shift on the same gameplay, with heightened tension and our good friend death** as a consequence. Later in the game, failure actually leads to you skipping to the final level, which you may be woefully unprepared for. These two consequences are a marked departure from the norm, but still have the death penalty inextricably woven within them.
** In this case death is not actually physical death but being caught, but the same basic conundrum applies: when the highest stake is being caught, but doing so ends the game in a way that expects you to reload, being caught in itself becomes meaningless.
Perhaps we have to look at the Racing genre to find our solution. Most games have a lack of advancement as a consequence for failure, or simply a lack of new funds for new car parts. Many of these are both non-linear and without storylines, meaning that these stakes are only gameplay-based and have no emotional connection past win/lose and completion. Forza 3 and Project Gotham both involve a true concept of time, which ensures that a loss is a truly lost opportunity which may not be repeated.
Looking all the way back to 1996, we found what is possibly the best fit for what we termed ‘failure without death’ in a game called Death Rally. Death Rally simply made you pay to enter races, meaning that if you continually didn’t win, you could easily endup without any money to enter a new race – eventually, loan sharks might actually take out your car. Failure on a gameplay level is not final in and of itself, but leads the player slowly towards a final death in a way that amps up the tension. The final death is never cheapened as it is the consequence of a number of smaller failures.
“We Don’t Even Need An Explanation Of Why They’ve (The Player Character) Come Back”
Ken Levine
At this stage, there was a bit of a pause: with Death Rally’s system, failure without death could be achieved***. The topic of conversation transitioned to one of adaptability: sure, this system was useful, but could it be applied to genres other than racing?
** *I’m not entirely sure how the fact that you could die within a race was neglected within our conversation. However, its easy to see how this system could be (and lets face it, probably has been) implemented in a game without in-race death.
First we looked at the FPS: the classic genre for meaningless death. The obvious problem is eliminating death as an option in generic gameplay. And here we hit a snag: the basic gameplay of a shooter needfully involves death, both as a consequence of your actions and as a possibility in failure. Some games circumvent the latter by saying you are ‘unconscious’ or ‘in the hospital’ (GTA I’m looking at you), however this stretches the world’s logic to the point that to the player, they essentially died.
It should be noted, however, that GTA, alongside many RPGs, feature a loss of XP or money upon death that mirrors Death Rally’s lost race entrance fees quite well in their ability to amp up the tension. The difference lies in where death occurs: either with the loss of money/XP, or as a possible, long-term result of it. The latter gives death, and the stakes associated with it, a greater importance.
In fact, it can be characterised that what we were looking for in this discussion was a game whose stakes were less called upon. The result is almost an inverse of Checkov’s gun: you’ve taken the time to establish that the gun’s in the room, so: don’t pull it straight away; don’t then ignore that it was pulled; and don’t then rinse and repeat like a washing machine on steroids. Established stakes in a game should be called upon in a long-term manner, while short-term gameplay elements should lead to recoverable failures which affect, but don’t often call upon, the long-term stakes in question.
The trick to all of this is matching all this to both your gameplay and your narrative elements in such a way that the entire outing makes some sort of sense.
“We Have To Find Ways To Make New Stakes That Are More Meaningful, Or We Can’t Have Emotional Engagement”
Ken Levine
Which brought us to the last game type discussed: creative games. In games where the goal is simply to create (Sim City, Minecraft and Rollercoaster Tycoon are just a few), your creation is the stake. Short term failures in these games are almost never fatal, but failures can result in the destruction of your precious items which have taken hours of gameplay time to create. In this genre, the stakes are necessarily meta in nature in that they are more related to they player’s input and sacrifice rather than any particular character’s. Maximal engagement is achieved not only by putting the player’s own emotionally-linked creations at risk, but by not needlessly flaunting, and thus devaluing, that risk to their faces.
It is quite fitting that we ended with this on that sweltering summer’s day, as it was in fact my creation, the Interactions Forum, that was on the line. And like a well-plotted narrative, a single failure on that day would not have spelt ultimate doom, but would have simply raised the tension for challenges to come. Yet failure wasn’t on the cards that day: the Forum was an exciting boiling pot of ideas, an excellent beginning to what will hopefully be a long-lasting tradition.
As I listen to Ken Levine and Brian Michael Bendis talk about the stakes in Spiderman and how they relate to games, I remember all this, and I smile.