Asynchronous gameplay is a popular phrase for describing various forms of online games that connect players but don’t require simultaneous play. Many eminent commentators have talked about the possibilities for this kind of gameplay, and how it might be the future for games.
However, in a fascinating debate on Gamasutra initiated by Ian Bogost, Raph Koster and I ran across a confusion of terms. Raph said that asynchronous games have existed for hundreds of years, citing play-by-mail Chess as an example. Except I think play-by-mail Chess is synchronous. When talking about synchrony, we actually meant two entirely different things.
Where many people casually talk about synchrony in relation to whether players are together in real time, I think it means games that require players to be in sync with one another in game time. This article elaborates on that idea and describes how real and game time intermingle.
Game Time or Real Time
Many games are played in real time. Sports, action games, console shooters, racing games and more are all real time experiences in which one or more players acts and reacts to a world in motion. However, like weight and mass, game time and real time are not the same thing. They only appear to be so under certain circumstances.
A single player game might be played in real time, such as Portal 2, but then you have to go away and cook dinner for your wife. So the game is paused or saved, to come back later.
A single move in play-by-mail Chess might not receive a response for several days. A tabletop roleplaying campaign may start and stop over a number of discrete sessions of real time spaced out by long periods where the game is not being played.
Game time is always internal to the game world, and it may not tick forward on a second-by-second basis. Whereas real time is all around us all of the time and always ticks at the same rate (relativity not withstanding).
This is an important distinction.
Synchrony
Synchrony is a property of all games:
If the attainment of a win as a part of a loop depends on the actions of more than one player, that loop is synchronous. If not, it is asynchronous.
Synchrony is often used broadly, such as different game modes. Single and multiplayer modes are the most common example, but there are some subtle mixes also. CityVille, for example, is a mostly asynchronous game with the exception of the gating mechanism that requires you to hire friends to work in a building. Gating requires their action to complete the loop for you to win, so that is synchronous.
That’s why I describe synchrony in terms of loops rather than the whole game. A loop is a basic molecule of gameplay in which the player takes an action and the game reacts in some way, determining whether the player experiences a win, a loss, or another loop. In an asynchronous loop the reaction comes from the game itself, but in a synchronous one it is dependent on other players.
Loops also vary in length. In team deathmatch Halo the loops are in real time, but in PBM Chess the game time only ticks forward when a turn is taken. So a single loop might take days to close. And yet the need for involvement of other players is the same. So that’s why synchrony is related to game time rather than real time.
Temporany
The more casual definition of synchronous gameplay means ‘playing at the same time’. The dependency on other players that my definition implies is not present, so it could mean players playing alongside each other within a common space, or competitively, or in parallel.
For me that’s just too hazy. Designing with player dependency in mind is very different to that of simple simultaneous presence. YoVille and deathmatch Quake may both be played by groups of players online at the same time, but they are completely different. It is not enough to say they are both ‘synchronous’.
In the debate on Gamasutra, I suggested that the casual definition of synchronous or asynchronous is actually describing a different property to synchrony. I labelled it temporany. The definition of temporany is:
If the play of the game contains the simultaneous presence of two or more players, it is contemporal. If not, it is atemporal.
So your Counter Strike session or Texas Hold’Em games are contemporal, but your Mass Effect or Chu Chu Rocket games are atemporal. Wandering around Playstation Home or YoVille is contemporal, but visiting a friend’s restaurant in Restaurant City is atemporal.
Temporany is about whether players are in the same place at the same time or not, irrespective of why they are so.
Where the Twain Shall Meet
The fun part (for me at least) is how synchrony and temporany interact with one another.
For example: You visit your friend in Farm Town, and he is online at the same time. You help him clean up his crops and chat to one another. Is that contemporal? Yes. Is it synchronous? No. It’s contemporal because both of you are in the same space at the same time. However it’s asynchronous because he doesn’t actually need you to help clean up those crops. Your help is not a requirement of his win.
On the other hand: Your friend in CityVille sends you a gift request. He needs Marble to help complete his community building, and asks you to send him some. Is this contemporal? No, it’s atemporal because you do not need to be online at the same time to send or receive the message. Is it synchronous? Yes. He needs your help to win.
Synchrony and temporany form a grid. There is contemporal synchrony, contemporal asynchrony, atemporal synchrony and atemporal asynchrony. More conventionally, although less accurately, you might call them: multi-play, parallel-play, serial-play (or turn-based-play) and single-play. Quake, World of Warcraft, WeeWar, Portal 2.
Not all combinations of synchrony and temporany work equally well in all situations, and it’s important to understand why. Each tests the tolerance of players differently, and so synchrony and temporany actually form one of the major constraints on all games.
Tolerance
Tolerance is the degree to which your game tests the patience or goodwill of players before they will stop playing.
Everything from teaching the player the rules of the game, to how it asks for money, tests player tolerance. The general rule is that the less engaged the players are, the lower their tolerance.
Tolerance is an unavoidable factor in all games, but you should always try to reduce it where possible. Synchrony and temporany test tolerance in many ways. Largely this is because of the fallibility of individual players, but it’s up to you to remember that players are not ideal and design for that.
Lexulous is a game cited by Jane McGonigal in Reality is Broken as an example of what she calls an asynchronous game with millions of players. Using the basic rules of Scrabble (and originally named Scrabulous), the game was a major hit on Facebook. However, contrary to when Jane wrote her book, it has lost practically all of its 5m users. As have the official versions of Scrabble that followed it.
The reason is Lexulous is synchronous, not asynchronous. Players either forget or don’t want to take their turn, which leaves many individual games orphaned. It’s not fun for other players relying on the forgetful player either, so they are more likely to drift away too. Their tolerance has been reached.
It’s an example of how atemporal synchronous (i.e. turn-based) play can have major issues – but also of how, for the more engaged players of Scrabble, it’s not a problem. All three games may have low MAU, but they have very high engagement rates (35-50%) compared to most other Facebook games. If you love Scrabble, you’ll remember to take your turn.
Another example of how games overcome issues of synchrony and temporany is drop-in/drop-out play in Left 4 Dead. The game is based around cooperative play and, unlike most cooperative shooters, you really can’t just go it alone if other players let you down. If you do you will die.
However the major problem with this design is how players leave or join the game. If the game is reliant on cooperative play then what happens when a player disconnects mid-session?
Left 4 Dead solves this by always being ready to drop an AI character in if a player drops out, and the AI is good enough to make the experience nearly seamless for the remaining players. The AI character is also ready to drop out again if another player wants to join the game. And so tolerance issues are minimised.
What tends to work least well (in terms of tolerance) is designs where players have to depend on each other. In social games the need for scale is paramount for this reason. Similarly, online team sports (such as Football Superstars) also have significant issues because of tolerance.
Generally speaking, the more co-ordinated a group needs to be, the more likely it is to represent only a fraction of players in the game. And if the game is wholly reliant on a co-ordinated group of players, the more likely it is to have a smaller - but passionate - audience.
World of Warcraft is largely played asynchronously, for example, and only a small subset of players get into the guild side of it because it’s a hassle (as satirised wonderfully by The Guild). Texas Hold ’Em is by far the most popular card game online in part because it’s easy to step in and out of. More serious card games tend to have much smaller audiences.
Finally, easing tolerance is the basis of many freemium business models. In CityVille you can pay for that Marble instead of waiting for your friends to send it, and in Ikariam you can speed your game up by paying. It works because players don’t want the aggravation that comes with waiting, and a small amount of cash seems an easy price to pay in order to get away from busywork and back to having fun.
Conclusion
Fun comes in many forms, and everybody wants to have some. Some of those forms ask more of players than others, however, and some of them have to account for issues of synchrony and temporany in ways that others do not. By clarifying what those two properties are, we are more easily able to see how they interact – and thus account for likely pitfalls in our designs.
So the point of getting the terminology clear is that it makes us better worldmakers.
(Today’s image comes from Wikipedia.)
I disagree that any work on definitions helps one actually design a game. It helps discussion, but not design.
These distinction exist whether they are defined or not. To point to them and talk of them and discuss and aknowledge them is useful, but to define them is useful only to those who know the definitions and only when they discuss them.
It does not make us better game designers. Or - ironically - as you define it, worldmakers.
The search for definitions does not increase possibility, By compartmentalising and categorising, an analogue space becomes a binary one. What was a scale is now a series of discrete steps.
These definitions are useful after the fact perhaps, to academically dissect what has been done, or what may come. But for the game designer, they are limitations, not keys.
Posted by: Sorrell | 06 August 2011 at 05:56 AM
Typos? No edit function?
DAMN THESE HANDS!
Posted by: Sorrell | 06 August 2011 at 06:08 AM
It sounds like you are against seeing patterns in the craft of design because it somehow hurts some sort of free thinking... That's just silly really. Definitions are the backbone of understanding and are useful tools in any task; design or otherwise. Keep the insightful definitions coming.
Posted by: Robert Massaioli | 06 August 2011 at 07:37 AM
@Tadhg - Chu Chu Rocket is best enjoyed with four players and feels a bit like multiplayer Bomberman only manipulating crazed Lemmings. Think you may want a different example for an atemporal game.
@Sorrell - I disagree, except perhaps if you are a sole creator. Terms like "loop" are essential for discussing how a player experiences a game and I need these terms to communicate a design upon which several team members are working. Essential for feeding-back effectively, too. Taxonomies are also useful eg. Bartle's Explorer / Killer / Socialiser / Achiever types which helps me to remember to cater for different audiences.
Posted by: Emmeline Dobson | 06 August 2011 at 10:18 AM
I think the definition of "synchrony" most often used is the one you would call "contemporal/atemporal". That's the accepted definition, and it's the one people consider most important.
It's most important because it makes a huge difference when programming the games. If you try to make a synchronous ("contemporal") game, you have to synchronize the player sessions; that usually means complicated network code that is prone to failure (lag, etc.). (Or you can do, for example, split-screen, but that's less common nowadays.) Asynchronous ("atemporal") play is much easier to program; it's about as difficult as sending an email.
Now, it's possible that these problems have been solved so many times already that we're no longer worried about them; we might be able to choose our terminology based on something other than how difficult it is to program a particular feature. But you'll probably still face resistance if you try to change the terminology. ;)
Posted by: John Evans | 06 August 2011 at 01:13 PM
Hi John,
I think it's accepted only in so far as common conversation. We as a creative community have a habit of doing that, and it leads to a fair degree of confusion. For example, 'game mechanic' is also a standard term but ask anyone what exactly a game mechanic is and you'll get 100 different answers.
So I'm comfortable with suggesting that perhaps the community is wrong and there's a more exact way. I get ribbed plenty for coining terms anyway (thauma, numina, storysense, etc) so that I don't mind :)
The code question is interesting. Surely a synchronised game is any one that requires players to be kept all in the same state regardless of real time. The hourly tick clock in Planetarion needs to keep all of its players in sync just like the FPS server. The only difference being that it doesn't have to correct lag on the fly quite so often. However keeping players in sync matters, so it's the same problem at different scale.
Whereas a shared space in which that doesn't matter (such a chat lobby possibly, or visiting a friend in Restaurant City) doesn't. That can afford to be lossy even though it's, as I call it, contemporal?
Thanks for the thoughts!
Posted by: Tadhg | 06 August 2011 at 01:56 PM
Hi Tadhg, really interesting read. I like the idea you're positing here. Unfortunately I'm afraid the words you've chosen for it aren't too likely to receive wide spread use in the way you suggest.
A/synchronous isn't exactly a casual word to begin with, but more of a term of art. In fields like education and media studies, it's already well established as a term indicating interaction in real time or over time. Hence, the casual definition is essentially the definition, while your definition is one that no one else really uses.
Meanwhile tempo, the root from which you've created temporany doesn't have much connection to place, so that also feels like a stretch. Again, I think the ideas you're positing here are really important, but I don't think your word selection helps those ideas to come across.
Posted by: Mosesoperandi | 07 August 2011 at 01:41 PM
Thanks Moses.
As I say, I'm comfortable with the definition back and forth for now. If the idea holds true then it will be adopted, if not then not.
What troubles me more is the conflation of many kinds of play under just one banner with no differentiation. Regardless of terms, the issue is that.
Tadhg
Posted by: Tadhg | 07 August 2011 at 04:03 PM
Sorry. Don't agree. You raise an interesting point about time, but use it only to justify your interpretation of asynchrony, despite the fact that others on the original blog entry disagreed. These are, in my opinion, linked but not causal or mutually exclusive. Nor are, as you imply, others confusing what you think synchrony to mean. They just don't agree.
Access to knowledge as much as ability to act denote synchrony. If I'm sat opposite you across a chess board much information is simultaneous, even with turn taking. Making it PBM means that someone has more information for a time. Even the lag of a slow connection to an FPS multiplayer match has to be accounted for. It's a continuum. Maybe that's where the problem with shared definitions lies?
Posted by: DoctorMikeReddy | 08 August 2011 at 01:42 AM
I disagree with your disagree Mike :)
What kicked this off was a confusion of terms, not just as "that's the wrong word" but rather a "Wait, that's an interesting take on it. What do you mean?". It stumbled upon an interesting discontinuity between kinds of play that happen online at the same time. So it's not just about the words. It's about what's behind them.
Access to knowledge is a part of the sync/temp grid. Game time is not always real time. It can be sessional or continuous, it can be fluid or measured in ticks/turns, and of course that affects how the game is both strategised and played.
PBM Chess is an example of atemporal synchrony. Moves are taken and there are sometimes long gaps in between to study the board. This happens at each player's pace.
So while I may take my move and see the board first, you still get to see the board for as long as you like before taking your move, and back and forth it goes. The co-presence implied by temporany is not a factor. Yet the game still requires moves from both sides to proceed. Hence: synchrony.
Chess played on a board by two people, on the other hand, is an example of contemporal synchrony. We are in the same room at the same time, so we are watching the same board. The co-present effect of temporany is a factor, and the game still requires moves from both sides to proceed. Still synchronous.
Speed Chess even more so, where time actually becomes one of the conditional rules of the game.
Thanks for the comment, as ever.
Posted by: Tadhg | 08 August 2011 at 02:14 AM
Continued from twitter -
I think my problem with this isn't the definition of player and game synch being different things, it's that you're taking the normal, accepted definition of synchronous - happening at the same time - and redefining it to mean having dependency - which is already covered by the word synchronicity.
Temporany - it hurts me just to write it - seems to be the difference between multiplayer and social multiplayer. One is the conventional imagining of multiplayer, one is the toddler definition. All in the sandpit, all playing different games.
The thing that bugs the ever-living shit out of me isn't pointing out these differences, it's naming them arbitrary and uninformative things. Temporany will mean nothing to anyone who hasn't read this. It contains no useful information aside something about time. And in doing so, it also defines the important thing about this difference to be about time, when it may well not be. A single game can easily contain both kinds of play, the literal and social multiplayer aspects. WoW quite definitely does. It could easily be that the purpose of such designs is to manipulate the social interactions in the game and nothing to do with time whatsoever. In fact, I'd imagine it's probably more likely.
The fact that you define a single-player experience in these terms you just made up, where they have no meaning at all, feels like someone inserting a screwdriver into some of the better parts of my neck. How is this useful?
The search for clear and interesting thinking about games has to be accompanied by clear and useable words to describe them. And the describe bit is key.
"multi-play, parallel-play, turn-based-play and single-play" are useful and accurate terms. They're well understood. "contemporal synchrony, contemporal asynchrony, atemporal synchrony and atemporal asynchrony" are not just painfully pretentious, they also contain no useful information whatsoever.
Clouding an interesting set of definitions with horrible made up words bullshit makes what you have to say less accessible, less useful and does sterling work to make me ignore the meat underneath. Please stop it.
Ahh, it's like the good old days, innit?
Posted by: Sorrell | 08 August 2011 at 03:46 AM
Have to agree that the terms as stated in the main peice are confusing neolgisms that need their contexts stated explicitly to have any use, when there are, as Sorrell mentions, perfectly fine definitions for these types of game/pattern of play relationships.
Sorry Tagdh, but these words are pretty redundant to me.
Posted by: Tony Coles | 08 August 2011 at 04:07 AM
Master Sorrell,
Indeed it is (side note for anyone else reading this: Mark and I used to work together.)
There's basically three reasons to kick back on the assumed meaning of synchronous that's appeared by convention rather than definition.
1.
The first is that it's technically wrong. 'Synchronise' in any context means things in alignment with one another, and that means dependent on one another. Synchronising watches doesn't mean watches that are running at the same time, nor does synchronising files via Dropbox nor keeping a game state in sync.
Synchronicity is also not a word with a totally different meaning from synchronous. It's a term invented by Jung to describe when two apparently unrelated events happen together (though not necessarily simultaneously) in a meaningful manner. A kind of deliberate coincidence if you like, with the stress on their relationship rather than their strict timing.
Even the root 'chron' as justification (meaning time) is not great because its usage tends more toward the recording of time (chronometer, chronicle) than time as a state or a thing in motion. 'tempo' is much closer, as in temporal, contemporary. Especially with the meaning of the word 'contemporaneous' literally being 'existing or occurring during the same period of time' and the related 'simultaneous' meaning 'Occurring, operating, or done at the same time.'
Aside a bit of dictionary fun, this matters because the more normal the terms that we use are in relationship to the language that everyone else uses, the easier it is for them to understand what we're talking about. So it's easier for students, investors or anyone to get where we're coming from. Even apparently sophisticated words that can be guessed at are better than words whose meaning inside the walls is essentially the exact opposite of it usage in the outside world.
2.
The second reason is that, from a design perspective, it's vague. To take McGonigal's example, she cites Lexulous as a great example of 'asynchronous' play and that this is why it's so popular. Only her example is fundamentally flawed: Lexulous is not at all popular. Yet many single player games (which is about as asynchronous as it gets) are.
So the lesson is not that Jane is wrong, but that the term itself is vague. Lexulous, just like any multiplayer game, needs players to play it. It just doesn't need them to play at the same time. So it's got a lot in common with Quake Arena, yet not at the same time.
3.
The third reason is to foster new thinking. I make no bones on this blog about saying that there are plenty of dead ends in thought around games, and that a lot of that is to do with assumed or borrowed language.
I'm comfortable with saying that some things need to be knocked down in order to go forward, and some things need to be coined afresh. It may seem like neology Hell, or even a self-referential exercise at winning the game of Internet, but it's neither.
Whether you agree with a new term or not, its invention can spawn new ideas. McGonigal's fiero, Godin's use of shenpa, O'Reilly's appropriating of Web 2.0 and more are all contributions. Some stick, some don't, depending on whether the body politic finds them relevant or not. All to the good.
---
As to usage, you mentioned that single player games could contain both. I agree. I defined synchrony at level of the loop rather than the whole game for this reason and cited gating in CityVille as an example.
However I'm right in saying that breaking it down into multi, parallel, turn or single is less accurate than the sync/temp grid. It's easier to grasp because it relates to terms you already know (yay) but those terms are not explicitly defined either (boo).
Single and multiplayer are well understood by virtue of many years usage and examples of games (although by your own admission, it's not always 100% accurate).
But parallel? Raph categorises foot races as parallel and also social gaming on the basis that they are played side by side.
I think that's wrong because foot races decide winners and losers based on placement (so are synchronous) while social games don't. Both may be played side by side, but are fundamentally different because wins in one depend on other players, but not in the other.
Turn based play is also fairly hazy. There is the clear example of Chess, where turns are sequential, but then there's Cliffski's Kudos, which is both turn based and single player. Or Minesweeper which uses discreet actions (I.e. Turns). Or VGA Planets where all turns are resolved at once. Or Planetarion where time moves forward in ticks and the player can choose to take a turn or not, and all turns across the whole game resolve every hour on the hour.
So the terms I use may feel uncomfortable at times, but they have a point and that is to get at that clear thinking you mentioned, and to do so from a fresh stance if possible.
Where would Einstein have got to if he'd called relativity 'gravity time'. Those were known words of the day and concept of relativity is related to both, and yet it's not quite right either. Sometimes you just need a big and new word to be specific.
Some might say I do that a lot, but it's only because I find the current set to be lacking for on reason or another. ultimately it's everybody else who'll decide whether some of my ideas are on the right track or not though.
Brilliant talking with you, as always.
Posted by: Tadhg | 08 August 2011 at 07:03 AM
Bravo, Tadgh. I agree with the detractors that coining new lingo when old words already suffice is presumptuous and annoying. I disagree with them that there are old words that work for the concepts you are describing. So you are perfectly justified in proposing new lingo. Conflation of words/concepts is a huge enemy in any sphere of thought, and especially so in emerging fields such as videogame design. I've been very inspired by your aggressive stance against conflating the words 'story' and 'world' in the past, so I'm happy to see your continuing battle to clarify concepts.
Posted by: Joshforeman | 23 August 2011 at 11:02 AM
Thanks Josh!
Posted by: Tadhg | 23 August 2011 at 12:13 PM