GDC10: Sporadic Play

 Posted by (Visited 17346 times)  Game talk  Tagged with: , , ,
Mar 102010

Liveblog of talk on Sporadic Play, Bryan Cash & Jeremy Gibson

Sporadic play describes a game where mechanics intentionally limit how often a player interacts with a persistent game world. We’ll talk about history with it, why it is a good thing, design concepts, and a bit about the future.


Most obvious place to start — Facebook. Look at the top 20 games here. Farmville is #1 and uses sporadic play. Mafia Wars, Petville, YoVille, 16 of the top 20 are sporadic play games. If you add up the MAU, you get 332m people, which is kinda BS, but just in the top 20.

It has also been around in console and traditional PC games. Animal Crossing, a game where as time went on in real life, events like Christmas happened in the game too. Seasons, real time, limiting what you can do based on the real time of the real. Another one was Kingdom of Loathing in 2002, a web game, limited number of things you can do based on action points.

Another recent example is play by email or play by forums. You may be familiar with Gaia Online — a game in the form of a forum. There are also RPG forums and PBEM RPGs.

And Tradewars 2002 — which came out in 1984, when 2002 was THE FUTURE. You flew around to different planets, traded, had fights. And at the end of your actions for the day, you had to leave your spaceship online while you had your day, vulnerable. So you would park your ship in the deepest darkest corner of space.

29.9m players of fantasy sports in 2007… pick rosters, teams, and then you get delayed results and watch what happens. Has been around since 1960. You are playing a game as a metagame around another game.

Going further back… play by mail. Peaked in the 1980s, but started as mail rules for Diplomacy by John Boardman in the 1960s. There were three magazines devoted to playing games by mail.

And then there is chess, by correspondence. Since you had so muh time you could do research, look up moves in books, and pick the optimal period. On this sheet people would mail back and forth there were choices for “I can’t read your handwriting” — a 2 month delay in a move! This goes back to the middle ages, kings sending moves back and forth.

The conference game this year is BackChannel, which is betting on keywords that will be tweeted during the show… a high tech sporadic version of Buzzword Bingo.

The stock market is also premised on sporadic play. Limited interaction, persistent world… been going on since 1602 in Amsterdam and 1309 in Antwerp.

And evolution: from a species point of view it is a sporadic play game. You get limited interaction, then 9 months later, a result… slow feedback… but hey after billions of years, we think we won.


It does not mean casual. People lump them together, and there is overlap, but consider Bejeweled Blitz. You can keep playing it as much as you want.

Key differences…

* depth of gameplay — stock market is a deep game.

* Variability of learning curve. Casual games have smooth ones… sporadic games vary a lot.

* Potential time commitment. Bejeweled Blitz can eat 4 hours… but a sporadic play game intentionally CAP you, so the potential for time commitment is lower.

* Persistence… the world persists in some way, and does not reset.

* Sporadic does not necessarily mean asynch multiplayer. It focuses on the relationship between the player and the game. It frequently USES asynch multiplayer of course.


Millions of people play sporadic games on social networks. People tend to talk about these in terms of viral tricks and we wanted to talk about it differently.

Also, reaching broad audiences. We were traditional gamers who didn’t have time anymore but still want to play.

Then there is the multitasking gamer. It is more common now to find gamers who are playing multiple games at once.

Then there are non-traditional gamers. They are looking for “something to do” with low time commitment and barrier to entry.

All of these audiences can be served by the same sporadic game.

This is also a new lens for designers — lens in the sense of Jesse Schell’s lenses in ART OF GAME DESIGN. Many games now are sporadic but we don’t think of them that way.

We believe sporadic play can be added to existing games and allow them to reach broader audiences.


We made this back in 2006 at ETC at CMU. In 2008 it won an award at the IGF, and also some awards from JayIsGames.

You play a character, build up skills, you have a plane. Buy weapons, etc. You fly places, you click on the combat icon, and fight some pirates in real time action. But mostly, you just fly. A screen with the plane bobbing up and down. Flight in the game is in real time. You trasport goods like fish from one town to another. But the entire time for this trade route is 11 hours and 32 minutes… you can see a timeline of a traderoute with multiple stops stretching into the future.

The game used to SMS players when an attack happened, interrupt you while you were doing your taxes or whatever. Fights with pirates accumulate, so you can catch up whenever you come back.

So it is a game where you spend the majority of your time not playing it. But there’s a persistent world, with brief interaction. My avatar in the game keeps doing my action queue while I go on with my life.

Now, Skyrates is not perfect. But it taught us a lot about these sort of games.


Respecting player’s time. (WoW screenshot… two guys who have played for 3 weeks… one an hour a day, and one 20 hours a day… “I HATE this latter guy”). But in sporadic play, even if you have 20 minutes a day you feel competitive. Also, it isn’t about player position, but about velocity — how well are you using your time every day, how efficiently you are using your time.

We did some analytics, showing players the other players using the same plane they had, comparing yourself that way, on efficiency.

Managing obsession. Everyone loves cake, but it can be too awesome. Players often play to the point of oversaturation… they get exhausted, who burn out beause the game allows them to play as much as they want. They play until they are sick of it. Remember the old showbiz rule: always leave them wanting more.

Developer benefits: it extends the life of your content. With 20 minutes of time, they run through content at a metered rate. It gives you more time to develop. And you an develop more based on consumption trends — watch where they go and what they consume and develop in response to that. You can “prune” your procedural content towards what players want.

Some examples… in Kingdom of Loathing or FB games they use metrics to decide on content. And in WoW they added daily quests, adding ways to block the rate of consumption. Indirectly turning it into more of a sporadic game. These are elements allowing you to play it sporadically of you need to.

It lets you internalize a game… a constant drip feed of interaction reminding you to re-engage. Email, Twitter, Facebook, SMS, “I should check in.”

EVE Online demonstrates that sporadic play does not need to be casual. Skill learning in EVE is sporadic — from a few minutes to a month to learn something. There are players who just research, and don’t play the game in any other way.

When a game is internalized, it becomes a regular activity… like checking your email, but it takes less time. Check in between classes, or at work. Keeps it in the back of your mind all day. It makes your life a little more exciting.

Sporadic play gives us new revenue models. Time is the most valuable resource, and we let people pay for the fourth dimension! Time is worth way more than their money. Having a more lucrative use of the time you have is really powerful. I might pay for a 50% luck bonus or 10% boost to Xp for a day.

Also paying to avoid the parts of a game you don’t like. Buy a mining robot to avoid the dull mining. So people pay you to play your game less, which also frees up server load.


* Time currency. Some sort of point that is gained over time. Action points, moves, health. You can play with the rate it accumulates. Mafia Wars gives you a point every five minutes. Others like Legends of Zork gives you 30 once a day.

Can you go past a maximum? Save them up? Can you adjust the rate with variation (more on weekends?)

Basically, this is about managing the obsession, you decide how often you want them playing with this mechanic, and how long an experience is.

* Scheduling things. Groups of actions that occur as time passes. You can sequential — queued movement. Parallel actions — like cultivation in farming games.

A question is who is doing the scheduling. You can make it directed and tell the player, or give them the option to pick different schedules.

And when is the player notified of rewards? As soon as you finish the action, and then you have to wait for action points to accumulate? Then there’s no suspense. On the other hand, if you hold off the feedback on success, whether the action worked, that means two hours of wondering what the outcome will be.

Then there is the question of how many items can be scheduled. How many fields do you have, how long a queue?

* Uncertainty.

Weighing the chances on whether an action may happen. Predicting the state of the game in the future… since this is a persistent world, the state can change over time. So if I send things in a direction which will take 5 hours, there’s the possibility that stuff changes in the 5 hour interim.

A question to ask is how much do you want players to look ahead. You have to ask yourself how much complexity you want. And how powerful do you want to make prediction? In Skyrates everyone was moving goods… so you tried to move goods to somewhere that was a good place to sell, but other players could see it too…

There is also real life uncertainty, whether real life obligations can interfere.

* Punctuality.

Player attention to the schedule. It ties into how players make choices. You plant crops, if you set up the game so that they get double crops if they show up on time, or spoil if they take too long. If it takes a ouple of days to spoil, that is a lot easier for players.

These elements add risk and analysis to the game for players.

The design question is do I want to reward punctuality, punish for being late? Determines how hardcore the game feels, how risky decisions are.

And is there insurance? You could make a really bad deal in Skyrates where your guy did stuff poorly because you set it up in advance and things changed… so they added an insurance mechanic to prevent a bad decision by the bot.

* Multiplayer.

How do player affect one another? Competition via leaderboard is one way. Trading in Skyrates was indirect comptition, or Mafia wars attacks when you are not looking.

Cooperation is possible too — ways to help people via asynchronous play. The funny thing about the current games is that some of them do this via asynchronous synchronous play, as in Mouse Hunt. Tournaments happen on a schedule, people who are on at the same time do make a difference. It forms some player relationships.

Roleplaying is another form… an event that triggers synchronous activity as a periodic or optional thing.


Engagement customization… a game that understands when I have time to dig in, or when I don’t. Most games just split this up. Games that adapt to how much YOU want to play. This goes beyond single player and into multiplayer. You see some games where there is a core that is there for every fifteen minute timer, and others who play at longer intervals. Is there a way to trade time, loan things, or otherwise play together within limits and be important to one another?

Elegantly handling the desire to quit. Netflix lets you put an account on hold, but lets you set a reminder to come back in the future. A way to maintain a bit of a connection to them. Let’s say you are going to drop out for three mnonths, can the game play itself for you in that time so you come back thinking that it has still been there for you?

Shared time currencies is the notion that action points could be shared between games or gamers. Ways to use the points you earn across multiple titles from one developer.

Or you can’t play again until your whole group finishes?

Or real life metrics — you play the weight loss game, lose one real life pound, that gives extra action points?

Community consensus. You can join a faction in wow, but so? Identity as part of a larger gorup isn’t that important. But think of voting… it is sporadic in real life, and based on that our community arrives at consensus. Can there be a sense of working together in this sporadic sense to shape the notion of what a group is?

Multi-engagement gaming. The idea that the same player can play differently on different devices is not new. “Keyhole interaction” — complex game, like stock market, but simple interactions, buy and sell.  But multi-engagement means adding a companion to existing IPs. Casual players supporting hardcore players, for example. Think about a Call of Duty World Conquest strategic game on Facebook that ties into the CoD FPS? Add a Yellow Ribbon mobile app — add my gamertag to your yellow ribbon on your phone. If you launchy the app daily, it means you are thinking about me, it boosts my character.


Sporadic play has been around for hundreds of years. It can be integrated into many types of games. It can be used to create companion experiences for existing games. It can help small developers extend their content, and it respects the players’ time.

Some links:

From the questions segment… adding chat helped with frustration when you wanted to play but were out of points because it gave something to do.

  18 Responses to “GDC10: Sporadic Play”

  1. ahem. i do believe as of 4 years ago, it’s called “passive web gaming”, mr. cash and mr. gibson.

    passive web games are brilliant because they fit into your life in the same slot as ‘checking your email’ or ‘checking the sports scores’ or ‘checking your portfolio.’ it’s hitting a site. clickity, click. ‘okay. my turns are done for the day.’ and you’re’s persistent. it’s massively multiplayer. it’s comptetitive. it’s social. it’s portable. it’s passive.



  2. As with most new (or recent) behaviors, there is often new technology at the root of their emergence. With sporadic play, I think that technology is mobile web browsing, in the broadest sense. The past 5 years has seen laptops get significantly smaller and lighter, phones get significantly smarter and more capable; people are generally able to be online longer, more frequently and with less investment.

    All of this enables sporadic play. Being online all the time—or being able to get their quickly and easily when you need to—makes it possible to game in small sessions (given a supply of games that fulfill that desire, of course). People can then choose how many sessions/day/game they would like to dole out based on their availability. It not only helps keep busy, ex-hardcore-gamer folks like myself gaming, it also allows people who previously were excluded from gaming because of the steep initial time investment to play. Think of all the people who have jobs/lives punctuated with lots of short breaks who have been taking up social/casual gaming in recent years—parents, nurses, receptionists, night-shift workers, etc.

  3. Nifty m3mnoch! Although just to remind, the original iteration of Skyrates (focusing on keyhole gameplay and sporadic play) was done in 2006 as well. It sounds like we were on similar wavelengths.

    I’m sure there’s room in the world for all sorts of naming. Although I’d hesitate before necessarily tying the experience to the ‘web.’ I think that a number of features of sporadic/passive gaming can occur in games that are separated from the web. Take Fable 2 for example, where you had the characters earning money when the player wasn’t playing the game, or Animal Crossing which only uses the console’s internal clock, or a number of iPhone games that never touch the web but are still persistent for that player.

  4. Games that are entirely sporadic lose me after a brief time. My farm is overgrown with weeds, my pets have gone to animal control, my cafe is empty.

    But sporadic elements in immersive games can enhance the experience. My favorite resource extraction system to date was that created for Star Wars Galaxies. You could spend hours prospecting the perfect resources for whatever you wanted to construct, but once you found them, you put up a harvester and then went on with your (virtual or real) life. It took out the “tedium for the sake of tedium” aspect that makes most mining/resource systems vulnerable to an influx of macro-wielding exploiters. (As importantly, it revealed a design approach that was not apathetic or actively hostile to crafters, something that seems increasingly rare).

  5. There is a hint of this in behavioral science, the concept of ‘free-feeding weight’. It is a measure of stimulus consumption given the effect of the stimulus on the subject. In the case of small animals, particularly cats, rewards in the form of food was enabled all the time until the test subject reached an ideal or target weight. Then the rewards are cut back to 80%. This had the observable effect of obtaining consistent target behavior even when the subject had recently sated on the reward.

  6. […] (11/3/2010) […]

  7. This dovetails nicely with a conversation I had the other day about Kiva feeling like a facebook game.

  8. And now I’m going to go and design a sporadic browser-based game 🙂

    Thanks for the notes!

    (I’m going to be there next year, hopefully. But looking at the timetable, I don’t even know if that’s a good idea; I couldn’t possibly choose between what’s on offer 🙂

  9. Dude…

    TradeWars was one of the best games ever and I just saw it relaunched recently as a web version at

    It’s back and alive – official remake… Check it out, very cool 😉

  10. Oh forgot mention… They’ve also got a Facebook version too

  11. Thanks for all the positive comments, everyone, and a *huge* thanks to Raph for liveblogging our talk.

    I’m working with Bryan on getting the slides ready to distribute…basically removing the fancy transitions, moving it from Keynote to something more universal, and adding notes to each slide describing what we said about it. It’ll take a little time, but I think the slides will make a lot more sense with the notes.

    — Jeremy

  12. Re: m3mnoch:

    I’m also aware of Justin Hall’s PMOG (Passive Multiplayer Online Game @, which eventually became the Nethernet. Justin was a grad of USC before I joined the faculty, but he and I discussed a little about the differences between passive gaming and sporadic play at his thesis show in 2007.

    One differentiator, I think, is that in passive gaming—at least in the sense of Justin’s game—sets up actions (traps) for other players to trigger and then waits, passively, for that to happen. Sporadic Play, on the other hand, still requires active participation, even though that participation is spread across a player’s day. Passive gaming also doesn’t really cover the same historical ground as sporadic play, with games like Animal Crossing, TradeWars 2002, and correspondence chess falling under SP but not really qualifying as passive.

    — Jeremy

  13. […] Neptune’s Pride is a browser-based multiplayer strategy game that falls somewhere between Galactic Civilizations and Risk. Eight empires fight to be first to conquer half of the galaxy’s ~180 stars. While the game ostensibly progresses in real-time, moving a fleet from one planet to another takes about 16 hours (before speed upgrades) and your economy only produces funds once a day. Thus players only need to check the game a few times a day, a style sometimes referred to as “sporadic play”. […]

  14. […] have to make choices (they are strategic choices rather than real-time, but so what? Games have a long tradition of slower play). The choices require knowledge and skill (the skill is what gets derisively called […]

  15. […] have to make choices (they are strategic choices rather than real-time, but so what? Games have a long tradition of slower play). The choices require knowledge and skill (the skill is what gets derisively called “spreadsheet […]

Sorry, the comment form is closed at this time.