Skip to content

Design Problems in Twitter Game Development

May 31, 2009

In the last week, a couple games appeared which use the Twitter “engine” – BackChatter and Spymaster. BackChatter is a conference game in which people submit three keywords and then wait to see the payout for said words in the #whatever channel that is covering the conference. Spymaster is currently in closed beta. I have not played, but judging my some of my friends’ Spymaster-related tweets, it’s similar to Mob Wars, level based, integrates an inventory system and lets players complete missions vs. others also in the game and on twitter.

As is my standard M.O., I am fascinated with the potential of any new platform. Twitter poses a couple interesting design challenges, though:

  • Channel flood: If the game or app in question causes one to spam the channel with messages like the one pictured, it begins to annoy people. These people, in turn, unfollow people or the channel until the game or trend has passed, but whether this is a desirable thing from either player or follower is questionable.

    While playing BackChatter and in an effort to test an exploit in the system, I asked my followers to retweet a particular message which contained my three key words as well as the #g4c tag. The result was dozens of messages floating through the #G4C channel interrupting actual conference-related material. Any twitter game whose primary mechanic relies on frequency or quantity will have similar issues. This perceived annoyance is enough for some players to stay away, yet this also appears to be the primary propogation mechanic used by these games. On Facebook, the annoyance of dozens of invites coupled with the flooding of the main channel, most recently with with every manner of top 5 quiz under the sun, was one reason many of us tired of Facebook in the first place. Mob Wars also updated the Facebook status feed, but because play required the player to wait X before accomplishing Y, the feed seemed less flooded. Since I haven’t played SpyMaster, I can’t peg the comparative math. A Parking Wars model which updates less frequently might prove a stronger model to follow. Since the feed in Twitter is the whole of the experience, attention to its flooding seems more critical than in Facebook.

  • Follower Advantage: In mentioning my flooding of the #G4C channel, I was able to leverage my 500+ strong followers list (@bbrathwaite), and asked Ian Bogost (@ibogost) to retweet my message to his 700+ strong followers list. Between the two of us, we probably have something like 400 discrete followers. The exploit that I noted previously relied solely on our ability to leverage this group of people. This is not unique to Twitter, of course. A larger network of friends allows you to succeed in many games. Typically, games tie your success to your willingness to propagate. By design, though, a game must consider how a player with 50 followers could possibly compete with a player whose followers number in the thousands (@djaffe).
  • No Desirable Opt Out: How do we not follow something in Twitter? In Facebook, I can notify it that I never want to receive an invite to a particular app ever again. In Twitter, my only perceived means of doing this is via unfollow.

I welcome your thoughts, as always, and please feel free to identify additional considerations that I’ve missed.

8 Comments leave one →
  1. May 31, 2009 11:03 pm

    I keep saying this everywhere I can – there needs to be some interest-filtering for lots of content, including twitter. I think the key is trying to leverage your social network and “share the load” so we all get better filtering.

    Technologically, how would you achieve interest-filtering in twitter? Tagging might be one way, but that’s gets old and cumbersome and I expect 5% of people (at most) will ever tag their tweets.

    Someone will likely write a plug-in for Google Wave that will draw your attention to things that usually interest you. Pretty Soon Twitter will be hidden behind layers of other systems and will become just a delivery mechanism.

  2. May 31, 2009 11:28 pm

    On the channel flood issue. I don’t have a problem with roleplays (like War of the World re-enactment) that occur in social networks. I just discussed the difference between these types of channel floods (in Twitter with @mikemonello, @imbri), and determined that an important factor that is missing from game updates is their entertainment value. The sameness of a game update is unintertesting, impersonal and repetitive, it is spam. So, I would be happy with game updates if they were entertaining in themselves. Their tweets become fun, dramatic, interesting, and not just promotional.

  3. May 31, 2009 11:34 pm

    Personally, I’m not a fan of games that rely on a larger friends list for the player to succeed. It destroys the beauty of selective social networking. I prefer to keep friends and followers whose opinion I trust, whose conversation I enjoy and to keep my list of acquaintences at a minimum. Yet the majority of social networking games make it extremely difficult for anyone with this objective to become overwhelmingly successful, even if they dedicate themselves to playing.

    I am playing Spymaster at present and, although I’m performing rather well at Level 4, I’m unable to take down some other players on Level 1 that, while they haven’t been playing long, have an overwhelming advantage through the fact they have a ridiculous number of followers.

    While I can continue to hack away and hope to one day “fluke” a win against these people, should I continue down my reserved line of networking I’ll be left by the wayside. In order to become successful, I’m expected to spread the word and get more people to sign up to the game who follow me, or simply begin following as many people as possible in order to build up a giant followers list myself. The latter is useless to me because it destroys the rich networking experience I currently have with people that I care about.

    What of a system that, instead of providing benefits that push the social “elite” into a greater advantage by default, the game incorporates a system based on percentages. That is to say, if 50% of your followers are playing, you’ll have a greater advantage against someone who only has 15% of followers playing, or that someone with 90% of followers playing has a greater advantage over both.

    For instance, if someone has 1000 followers and only 20 people are playing, they are at a disadvantage as opposed to a player that has 200 followers but with 170 of those playing.

    There would need to be more complicated restrictions within this system (so that a player with 1 follower that also plays wouldn’t be able to rise straight to the top) but it would ultimately help even out the field.

    In the end, this would help the game encourage promotion, too. Rather than rely on the larger numbers of followers to succeed, you would have to actively direct your followers to play the game in order to improve your stats, rather than just sit idle with power.

  4. June 1, 2009 3:39 am

    As I said on a comment in a previous post, I’m pretty concerned about the perceived spam issue occurring in many Twitter games. I’ve recently played Resistance 2018, a Twitter game made for the launch of the new Terminator episode.

    The game didn’t count at all on the number of followers you have. It just presented very small puzzles (mostly anagrams or fill-in-the-blanks) that you must solve by sending a reply with the solution and a given hashtag in a limited time.
    This way, not the number of followers, but the presence on the channel was rewarded.

    The problem is, a lot of people started flooding the channel with messages like “@Resistance2018 annihilation #RA:WM”, which of course are meaningless and perceived as spam. The game tried to add a bland story element, both by sending some background related tweets and by rewarding players with more point for the retweet of a given message. That didn’t solve the problem, to me. The tweets were not related and were almost as meaningless as the players’ reply answers.

    I think the only way here is to use direct messages more, or to find a clever game to give a sort of meaning (for example, a narrative meaning) to the tweet a player has to send to play.

  5. June 1, 2009 11:05 am

    While it’s neat to watch games for Twitter develop, it seems like it continues to be a terrible platform due mostly to flood. Follower Advantage could be beat by just using a changing modulo of the absolute follower number — right?

    It seems to me that the perfect tweet-game would be covert. Where the people who were playing would *get* that your tweets related to the game but your “muggle” followers wouldn’t perceive your participation as spam. Something like a conversation including a central hash-tag that had to follow some set of evolving criteria? TwitterNomic or something

  6. June 1, 2009 11:34 am

    I thought the Backchatter game at G4C was fun, and your attempt to flood the channel with your 3 words was clever. Although… why don’t I see you on the Leader Board?
    My thought about the game is that it would be better if they eliminated all words that appear in the conference title, and also the abbreviation ‘RT.’ Looking at the Word Cloud for G4C, the winning horse words to pick were ‘Game, Games, and RT’. ‘Change’ also ranked high.

    • June 1, 2009 4:20 pm

      Traci, I didn’t win because the wireless failed when I sent my three word tweet. So, I optimally gamed a system I wasn’t actually participating in. Ah well.


  1. E3 about to begin | Virtual Learning Worlds

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: