Round Table Topic

There has been a slow and steady progression of the role that death plays in video games since the days of Ghouls and Ghosts and Super Mario Bros. No longer are most games tied to the old tenet of providing a player with an arbitrary number of lives, typically three for some reason, before the player is sent back to a previous spot in the game. This was a particularly brutal practice back in days of arcade games (which the aforementioned Ghouls and Ghosts and Super Mario Bros. were no doubt influenced by) where the goal is to punish the player and, more to the point, convince the player to insert an additional quarter or so. There were also extra lives given to players along the way to sort of feed the mini-addiction that these arcade-inspired games needed to feed in order to get players to cough up more change. This practice has persisted in the industry for ages to the point where designers still treat death and the process of player punishment much the same as we have for ages. Game developers and designers have largely abolished the system of giving players a finite number of tries or lives in a game, but so many games still cling to the concept of a player having a "life" to work with. This approach to handling the mortality or failure cases for the player can often lead to frustration (oh hey look at that) and that, for most titles, is not a desirable trait to aim for. There will always be the occasional Ninja Gaiden that will be released where designers very intentionally challenge player's skills. These kinds of games rely on a failure case as a means of reinforcing the lack of player skill or, alternatively, enforcing a certain in-game aptitude. There have been a handful of recent AAA games which attempt to make death less deadly. One of which is Lionhead Studios' Fable 2 which allows the player to die but then instantly resurrects him/her with permanent aesthetic scars applied to the player's avatar (also a minor experience loss). This mechanic still allows players to die and experience a failure case but it does not impede their progress through the game or, really, make for much player frustration (faux-challenge). The most recent and notable game which attempts to make its players think about death differently is Ubisoft Montreal's Prince of Persia. Prince of Persia treated death as a mere "misstep" and, essentially, gave it players an incredible number of mini-checkpoints that they would be restored to in the event of failure. Should games continually rely on death as a means of enforcing player skill? Is "death" the best way to do that? Is the current system of dying and returning to some designer-specified game checkpoint the best way of managing a player's progress through a game? Dying isn't fun, but is overcoming what caused a player to die before fun (and worth the annoyance of repeating a gameplay segment)?

The Death of Death

A game is often defined by its rules. If every game was played like Calvinball then there would be no common understanding of the game for people to play. Games without rules would play out like they were written by Kafka; one person would seemingly understand everything that is going on while another may feel constantly lost amidst faux-procedure. For a game to have any real comprehensible significance to its players, then, it needs to have some set of rules. And for these rules to have any gravity or meaning attached to them, the players of a game need to feel that the rules are there for a reason. This is the kind of "fact of life" that any child hears repeated throughout his or her life. If rules exist for a reason then it stands to argue: what if the rules are violated? If a rule of a game is that a player needs to keep himself above zero health in order to progress through the game, then something has to happen to reinforce the importance of this rule to the player. Or, in a similar fashion. And then there is the condition of gameplay where a player may want to just break the game rules as he sees fit for the fun of it. There isn't a game player in the world who doesn't at some point wonder: what happens if I do this? What should be the recourse for some sort of failure to obey (or, more generally, a failure case)? The role of these failure cases coupled with a reward (progression, character development, winning the game, etc.) forms the basis for the primary psychological structure fundamental to games: risk and reward. And like a lot of other design concepts, this is not one that is exclusive to games:

Denial and reward can encourage the formulation of a rich experience. In designing paths of travel, try presenting users a view of their target -- a staircase, building, entrance, monument, or other element -- then momentarily screen it from view as they continue their approach. Reveal the target a second time from a different angel or with an interesting new detail. Divert users onto an unexpected path to create additional intrigue or even momentary lostness; then reward them with other interesting experience or other views of their target. This additional 'work' will make the journey more interesting, the arrival more awarding. -- Matthew Frederick, 101 things I learned in architecture school

The role of death in games is, in the general case, the game designer's primary form of denial and reward. A traditional action game, for instance, may employ the occasional puzzle but its primary frictional force to hinder player progression through the game is the use of enemies and combat. Making a player invulnerable for combat encounters would allow for a guaranteed progression through the game, but the player may struggle to find meaning in the killing of enemies or toppling of non-mental obstacles along the way. Death is our answer to this. By allowing a player an easily-conceptualized form of punishment, we work to reward their eventual triumph and make the forthcoming activities more meaningful by that virtue. And, as a lot of the contributors to this round table surmised, death is less about any concept of "death" and more about the relationship of a number of universal gameplay concepts that designers must battle with in the development of any game across any genre: failure cases, challenge, and the symbiotic relationship between risk (or denial) and reward. Tobias Hoffmann starts out this topic's discussion with what he feels is a solid rule of thumb regarding the three life rule:

Losing the three lives are three non-fatal failures to the player, and it communicates the simple rule of thumb: If at first you don't succeed, try harder. If you don't succeed at the second attempt, try even harder. If you don't succeed at the third attempt, give it up. There's no need to make a fool of yourself

It's a cute rule, but it's one that simply doesn't apply to modern games save for those that intend to evoke the feelings of the days of platformers on the NES. Thomas Kiley shares a well-written summary of his feelings about the relationship between failure cases (death), challenge, and the balance that must be struck between them:

I think this topic ties strongly in with difficulty within a game. It is all about balance. I'll use two extremes to explain my point. In the first game, there is no punishment for failure. Death (or equivalent) is simply not possible as you have infinite health. In game B, if you die once, which is perfectly possible, the game resets. You have to start again, your save wipes. The problem with the second one is fairly obvious. If you played a game for more than about 20 minutes and you had to do all that again, you simply wouldn't. The problem with the first one is there is no challenge. If there is no challenge, there is no reward. Not only do you lose all sense of realism, and hence, immersion, the game ceases to be tense in any way. You no longer feel connected to the world or your character. If you can't fail, there is no point in succeeding. [...] So, like difficulty, it is about balance. In Fable (2), I feel no fear when rushing in, as I know I can't die. I don't think this is a straight out bad thing - your meant to be a super powerful hero whose going to save the world, I would find it very difficult to believe that some small time criminal can take me down. Indeed, any death breaks immersion because, in real life, you can't die then try again. For me, the answer is that the player should constantly fear death (depending on the game, but as a general rule) but never actually experience it [...]

Spoonbender attempts to argue the inherent problems with death in games:

[...] there's nothing immersive about dying either. Death will, pretty much by definition, always destroy immersion. No matter how you handle it. [...] Go and watch a typical action movie. The hero regularly screws up, experience setbacks and so on, but they don't die. They might have to start over from scratch, get booted out of the police force or whatever else. They get punished, sure, but they don't die. Perhaps games should do the same thing. Rather than killing the main character every time they screw up (and then ending up in a situation with no realistic/immersive way out), just let them, well, screw up.

In an attempt to frame the discourse of the conversation in such a way as to allow everyone to work off of the same definition of death and challenge (spoiler: most didn't work off of it), Tyler McCulloch gave a concise post on just that:

One of the things we must overcome is what we define as a challenge. Do we really need death in order for a game to be perceived as challenging? The most common challenge in games today is the avoidance of death. While this works for many games, this does not mean that we cannot think beyond this. Challenges and the difficulty of a game can become more grey instead of just a black and white, 'do-or-die' situation. Puzzles, World navigation, character modifications are all examples of challenges that don't necessarily need to involve death.

Tyler's point received a very well-worded counter-argument by Tristam MacDonald a few posts later (the emphasis is mine):

Tetris is challenging because you are trying to avoid 'death' (the playing area filling up), world navigation is challenging because you have to avoid certain situations that risk death, and character modification is challenging in that it affects your ability to avoid death. Most games quantify player progress in the form of resources - be it gold earned, stats/levels, or just time. Challenge is synonymous with Risk, and the player can only risk resources - challenge then is allowing the player to be deprived of resources, a mechanic commonly called 'death'. We can of course call it something other than death, and disguise it however we choose, but the core Risk/Reward mechanic remains.

The last sentence is one of the most poignant posts made in the entire thread and summarizes a lot of what this discussion is about while also pointing towards gaming's reliance on a profound failure case (death) to make up the "risk" portion of the risk/reward relationship. This is a concept that Gary Hoffman delves into:

Death is just one motivator that can be used in game's design. It's a state of failure and the player has to work to avoid it. You can call a failure state something else and wrap it up in a pretty bow but it's the same concept at its core. The player knows they will be punished if they fail to succeed at some part of game. Yes, the type and extent of the punishment varies, but its punishment none the less. This motivates them to achieve the game's goals while giving the player a sense of risk. Since, risk is often associated with fear & adrenaline (and hence excitement), well introduced punishment mechanics like death can make games more exciting while motivating the player to advance. The other powerful motivator in games, reward, can motivate the player but seems to lack the element of risk. A game that doesn't use death (or punishment in general) has to rely on the player seeking some sort of reward to motivate them to continue advancing. This reward could be something in game like completing a quest and getting gold or it could be personal satisfaction at completing a puzzle, etc. Getting a reward can lead to satisfaction and excitement (especially with random rewards like finding a rare item in a game) but still lacks risk and the feeling of danger about it. Personally, I don't like dying and having to repeat some section of a game. It lessens the impact, not because I "died", but because I'm now going through the motions I just went through simply because I made a mistake. I'll now defeat the challenge of the game but not because I played better as much as I simply know what's going to happen. Everyone's tolerance for death and having to replay parts of the game is different so I don't think there's a right answer but I think in the perfect (unachievable) world the threat and risk of death would exist but the game would keep the player on the cusp of death, always teetering over the edge of the cliff but never letting them fall.

Riffing off of the example games I listed in this round table's description, Orangy Tang takes on Bioshock's death mechanic (or lack thereof):

Fable [2] still actually punishes the player for dying (albeit relatively lightly), whereas in Bioshock there's zero consequences for dying - you're simply resurrected in the nearest vita chamber, which is never more than 30 seconds walk away. You lose no experience, ammunition, status or kudos other than the personal frustration of not hitting "heal" quick enough. Personally I'm not a fan of having death have minimal or zero side effects - not only does it make a mockery of people who actually play the game skillfully, but because it can mean a player can force their way through a game without really understanding it. I've witnessed this first hand when a friend was trying to play Bioshock like they'd play Quake 3 - run around frantically while shooting everything that moves. The result is that they miss the subtler stuff (like listening out for enemies or security cameras), and fail to learn how to use the weapons or environment properly. They end up dying repeatedly (and frequently) but because they're making slow (but painful) progress they never stop and realise that they're missing the point of the entire game, and end up dismissing it as too frustrating to be fun.

Thomas Kiley fires back with an argument about player interpretation of the important aspects of a game:

Interesting points, and Bioshock is a good example. My counter argument is this; in Bioshock the point of the game, for lack of a better phrase, is to explore and discover the cool under-water world. If the game was continuously challenging, the player would end up focusing on the combat and ignore the story. In a game like Halo, I focus on the combat, so I haven't collected any of the skulls, I don't explore the world, because I don't find that to be the focus of the game. By making death so irrelevant, they allow the player to explore the world fully without constantly worrying about dying and having to repeat an entire section. Without this mentality, many players wouldn't explore, as dying miles away from where you are supposed to be is worth than dying a few feet away.

This discussion yielded a very important point, if a very tangential one: what happens when there is a major, if understandable, difference between a player's intent with the game and the one of its designers? Bioshock is a game that had some of the most gorgeous architecture, interesting level design, and brilliantly-handled atmosphere in the history of the gaming medium, but it's also a game that, in my mind, ended up focusing far more on combat than I would have expected (or wanted). Its choice to allow players an infinite number of "lives" (resurrections through nearby vita-chambers) with no real punishment other than a minimal amount of backtracking and the chance that some enemies may have found a way to regain some health in the mean time was one that enhanced its players ability to make their way through the game. For players who may have been more focused on the game's aesthetic design and narrative than its combat and difficulty, then, playing through the game was more tolerable than, say, the Ninja Gaiden and God of War games of the world. Johannes gets the discussion back on track, or is simply kind of a conversational downer, by bringing us back to death:

The problem how I see it is that death [of the player character] is usually not an interesting gameplay event, mostly because it's usually is paired with a save system which makes it meaningless. Take an example. Assume a game, any game really, with the added game mechanic that periodically forces you to press some specific button, failure to do so within some reasonable time would lead to a reset of the game with the associated loss of progress. I think it's not far fetched to consider that particular game mechanic to be detrimental to pretty much any game. It's similar to functions such as eating or sleeping that get left out of other games because they add nothing valuable to game experience, and are instead implied to be performed automatically behind the scenes. This is analogous, although admittedly far-fetchedly so, to the save mechanic of modern games. "Death" means that the player is forced to resume from the last saved game. So as in previous example, why should you get punished for not pressing the quicksave button every 10 minutes instead of saving being done automatically behind the scenes? And if it is done automatically, popping the player back to the last safe position upon "death", why bother to model it is as death at all since it's quite clearly not the permanent end it implies? Then on the other hand, there are scenarios where death is a fundamental gameplay mechanic. Any multiplayer deathmatch game is a good example. There your death means the success of the opposing player[s], which is in itself interesting. It's not hard to come up with other scenarios either. Tetris for example, here the "death" mechanic is interesting because it marks an end of the game. The challenge, a meta-game in a sense, is to see whether you are able to play better next time.

This is a concept worth exploring: if death/rebirth is instantaneous and very little is lost in the process of dying, is the "punishment" of allowing another player to succeed enough? In a competitive game, the sole goal is generally not to get one kill (unless the player is really awful at the game) -- no, the goal is to win. Every death in a deathmatch game means that a player is further from winning than he/she was before, and in the realm of competitive multiplayer, that may be enough on its own accord. Most competitive games make death more punishing still through the removal of a player's previous weapons and, maybe, a "time out" period that they must wait in order to be resurrected, but the primary punishment is allowing others to pull ahead. GerardL brings up the topic of Mirror's Edge as a game that handles death well by splitting it into, as Thomas Kiley writes it, "compulsory punishment" and "optional punishment." GerardL makes this division by saying:

The normal campaign has an insane amount of checkpoints, so that you never have to do a large part twice, but broke up the challenges is small parts, that you did have to complete in one go. However in the speed run/time trail is where the game really shines. If you die, it is almost impossible to achieve the target times for the levels (because dying takes a little time). You can however still finish it the level.

This is a bit of a flawed example in practice as, in Mirror's Edge, the time trials are only unlocked by playing through the entirety of the game's campaign. While playing the story mode of a game may not seem like a poor activity to ask a player with, as it's the focus of Mirror's Edge in theory, in practice it is incredibly painful. The divergence of gameplay across two modes which enact the risk/reward model differently for the player, though, is a great concept that is definitely worth further discussion. Jacob Ensign writes one of the closing contributions to the threads which discusses the importance of "death" as a means of influencing player strategies and play styles through adaptation in a way that is beneficial to the integrity of a game's gameplay and mechanics:

There have been a couple people that have pointed this out in some way or form. A player's strategy is devised in response to the rules and goals of a game. A light death penalty will cause the player to create different strategies than if there is a harsh death penalty. A light death penalty might cause a player to develop strategies that are more focused on the completion of the game at any cost. This may even involve leveraging the effects of death to accomplish a particular goal if it is expedient. Take for example a room that is filled with deadly traps. Instead of the player taking their time to locate the traps, they may die repeatedly in order to discover the locations of the traps or bypass them using brute force if they are instantly revived every time they die. A harsh death penalty might cause a player to develop strategies that are focused on survival. This would include doing things like hoarding resources, building impenetrable defenses, or putting more consideration into the effects of their actions in different situations.

Are death and failure the best way to encourage a player to create unique and entertaining play strategies, though? At this year's GDC, Clint Hocking gave a presentation on influencing player strategies through a combination of intentionality and improvisation. These two high-level concepts formed the two tiers of a player's play style: one is, essentially, a strategy that a player has going into a scenario (intention) and one is a strategy that forms dynamically in response to a failed execution of that strategy (improvisation). The Far Cry 2 team encouraged this method of play through a series of gameplay systems that occurred while a player was fighting and none of the strategy-altering ones would, necessarily, cause a player to die, just to adapt his/her strategy; however, death was certainly a possibility of a failed strategy, but the team also gives a player a sort of "cushion" against death by allowing an in-game buddy to come and rescue the player. This gives a player an immediate second attempt at executing a combat strategy before he/she dies in the more traditional manner (having to reload from a saved game). I'd like to close out The Death of Death with a post made by Jason Kozak who, instead of providing lengthy arguments, made a very insightful observation regarding Orangy Tang's earlier post and chose to pose a few very insightful questions:

What is particularly interesting about this point, and identified in OrangyTang's player-story, is the failure of death as a game mechanic to actually force a player to adjust their play style. It is interesting because death as a mechanic is applied broadly across all methods of failure. A player who charges in recklessly receives the same response from the game as a player who runs out of ammo at a bad time. Which begs the question: Are some of the issues resulting from death as a mechanic a failure of a mechanic, or a fault of the broad application of it? Should we be applying separate mechanics based on the conditions of failure?

In the end, how a game implements its fundamental denial (or risk) and reward structure is completely dependent on that specific game. There is never going to be a concrete set of rules by which a designer can know the best way to make a player constantly feel as if he is performing a set of activities that are bordering on just the right amount of challenge to be meaningful for that player. What the concept of death in games really comes down to is how much of a punishment needs to be attached to a failure case? What kind of stakes should a player have to keep in mind whenever he is playing a game and how much caution does a designer want to impose on a player's style of play? This is a topic that is being experimented more in games in the last couple of years and, while there is no perfect definition, seeing non-standard takes on failures cases in games like Bioshock, Prince of Persia, Fable 2, and even a game like The Path, is great topic for further discussion in the field of game design. You can read the whole discussion at GameDev.net at Game Design Round Table 1: The Death of Death.