Introducing
visible points and progressive levels is one way to make a system feel game-like. But points & levels are meaningless without a clear set of rules that outline how points are earned, and what it takes to attain new levels of achievement.
Rules are a foundational element in game design: they setup the fundamental challenge of the game, and give direction and focus to the players. The same applies to social systems that track points and levels. For example, here are Amazon's
rules for earning Reviewer points and becoming a 'Top Reviewer'. Clearly, you 'win this game' by publishing reviews that other people find useful -- which is a smart social game for Amazon to be running. Similarly, eBay's Feedback Forum includes
clearly-outlined rules about giving and recieving feedback and
earning new achievement levels. Dig deeper and you'll find
eBay's policies for dealing with feedback abuse-- which are basically people's attempts to game the system.
Making the rules explicit makes the 'game' of earning points and levels easier to understand -- AND inevitably opens up the 'meta-game' of trying to earn points in alternate ways (e.g. cheating). As long as there are games with rules, there'll be people who test the boundaries of the rules -- and the greater the percieved value of 'winning the game,' the harder people will try to subvert the rules to their own ends.
Google is a great example of this ongoing dynamic. If you think of PageRank (Google's search results ranking scheme) as a complex linking game, then each search result is an ordered list of sites that are 'winning the game.' As Google's popularity grew and appearing at the top of a Google results list became more valuable, people started gaming the system in various ways - which caused Google to
change the rules of the 'game' in order to identify and subvert cheating. Although Google doesn't publish their rules, people still play the Google meta-game to gain exposure and publicity for their sites.
Achieving a high ranking in Google search results is a form of competition for scarce resources -- which is another powerful way to makes a system feel game-like. Ranked lists like Google search results and the Amazon Top-Reviewers list stimulate competition among participants, and reinforce the rules of play. They also provide the 'seeds of culture' by advertising what qualities are valued within an environment. Orkut, for example, publishes three
Leader Boards (the gaming term for 'ranked lists of winners') that correspond to social statistics that the system is tracking. This gives people a concrete game to play -- as evidenced by the emails I've recieved from several folks on Orkut, begging me to take actions that will help them increase their standing in these lists.
I'm sure you can think of other examples where rules and compeition add a game-like flavor to social software (e.g.
Technorati), and I'd love to hear about 'em! My main goal here is to underline the power that a simple ranking system with some well-defined rules can have on the social and cultural dynamics within a networked environment. I'll conclude these musings tomorrow with some thoughts about the game-like nature of exploration and discovery in a networked social environment.
1. Marius on March 4, 2004 7:59 PM writes...
Great article. Waiting for tomorrow. The Games will go on...
Permalink to Comment2. paolo on March 5, 2004 11:03 AM writes...
Today is March 05, 2004.
Permalink to CommentBut the date of this post is April 02, 2004.
There should be some bug somewhere.
3. hylton on March 5, 2004 12:06 PM writes...
Thanks Paolo - fixed.
Permalink to Comment4. Ross Mayfield on March 5, 2004 12:51 PM writes...
Orkut is a recent example of changing the rules. At launch, profile, network and community pages pictured the top 6-9 by friend rank. This encouraged preferential attachment and was driving towards a power-law distribution. Then they changed the game by randomly displaying pictures on these pages, but introduced a stats page to highlight the very very top ranked. Kind of like killing off the middle class.
Whats also interesting is how part of the game has become figuring out the rules, notably what puts people in jail. A political analogy could be either facism or an unwitten constitution.
Permalink to Comment5. AJ Kim on March 5, 2004 2:03 PM writes...
Great example, Ross --I love your analogy about killing off the middle class :-) This really drives home the point that Leader Boards are a powerful lever for tweaking collective behavior and shaping cultural values. I hope that Google re-introduces randomness into their system somewhere --kinda like an 'I'm Feeling Lucky' feature for finding people, rather than websites.
Yep, Orkut's 'you're in jail' meta-game looks like the result of an unwritten constitution. They clearly didn't think this through before launching - but hey, it's Beta, so they're figuring it out on the fly. I'll bet that Google will learn quickly from this experience.
Personally, I think Orkut has a chicken-and-egg problem with sense of purpose. At this early stage, Orkut is wide-open and experimental -- so many cool little features with lots of potential! -- and thus it's hard to focus on using the tool for any particular purpose over time. To write a good constitution, you need a clear sense of purpose and direction for the environment you're building. Perhaps that will simply emerge over time - that would be very Google-like :-)
Permalink to Comment6. Gregory Narain on March 5, 2004 3:05 PM writes...
I would add Friendster as a potential example as well. When Friendster initially launched, they were completely open and allowed anyone to find anyone, within the 4-degree universe.
Recently, they have announced that they plan to add more controls to the system which will allow your profile to be secured. I'm all for the security, but nonetheless the rules have gotten more restrictive and protecting the user, however, I'm sure many ofthose playing the game may not like the boarded up windows being added to member's profile pages (or the lack of new friends made).
Permalink to Comment7. AJ Kim on March 5, 2004 8:16 PM writes...
That's a great example - thanks Gregory. It's interesting... both Friendster and Orkut started out as low-profile, semi-private, social systems -- and now, as they're scaling up and become less private & 'insider-er' they're adding constraints progressively over time.
In game design, you often think about competing for limited resources EARLY in the design process -- because that's part of the fun, part of the challenge. Constaints create more interesting (and potentially meaningful) decisions.
Permalink to Comment8. Dan on March 9, 2004 3:45 PM writes...
Metagaming.
Cheating isn't necessarily metagaming. The metagame is a game on top of a game. If you think about it in game theory it is playing the game at hand with some historical knowledge of what has gone before but is not relevant to the game at hand. The other game player's rule set is also different from your own and doesn't necessarily haev anything to do with the ruleset of that particular game.
Imagine your playing a series of games of Risk and someone you beat the game before keeps attacking you so that you lose. And this goes on. They're only playing to make you loose. That's a metagame. They win not because they win any game, but because they make you loose.
Metagaming slashdot for example is not about the points gained, it's about what you do with them after you have gained them. The fact that when karma whoring was prevalent it was easy to accrue points meant that this could be repeated and the metagame as rewarding for those who played it.
Permalink to Comment9. AJ Kim on March 9, 2004 9:24 PM writes...
Thanks for your comment, Dan. In gaming circles, cheating & hacking (and generally bending/breaking the rules) is definately considered ONE TYPE of meta-game -- but as you point out there are others as well. My overall point is that, in any particular environment, players will often pursue a goal that's other than the goal set by the designers (e.g. the FakeSter craze on Friendster).
Permalink to Comment10. Jamie Pitts on March 23, 2004 4:12 PM writes...
Just commented on this in the SW blog:
Traders in markets such as the NYSE participate in what is probably the largest-scale implementation of social software which uses game play to achieve, through rules governing behavior and information flow, the proper balance between cooperation and competition.
Permalink to Comment