Conflict simulation, peacebuilding, and development

MMOWGLI almost ready to sail

An email from the MMOWGLI team went out today to those who have signed up for the forthcoming playtest, indicating that the Office of Naval Research’s experimental crowd sourcing/simulation experiment will launch next week:

Dear mmowgli player,

Thank you for your interest in mmowgli-a groundbreaking experiment in collective intelligence. You’re officially on our team of pioneers and early adopters as the first people to see mmowgli in action.

mmowgli officially launches next week with a lightning round of fast-paced gameplay.

What will YOU do to turn the tide of Somali piracy? Contribute as much as you can, but of course the more you play, the more you can change-and win-the game.

Watch your email early next week for the exclusive launch time and online invitation.

With best regards,

The mmowgli Team

Public playtesting had been delayed by the sheer number of people who had signed up to participate—more than 14,000 at last count. Indicative of the public attention all this has received, Rear Admiral Nevin Carr, the Chief of Naval Research, was on MSNBC yesterday to discuss the launch.

In fact, the Moves Institute at the Naval Postgraduate School is currently running a pre-public playtest of MMOWGLI at the moment. The game’s interface is a refreshingly clear and intuitive one, and the feedback and scoring system is designed to reward the most stimulating contributions (although whether those turn out to be the most thoughtful ones remains to be seen). The first stage of the game is all about generating short, tweet-sized micro ideas:

In later stages of the game, players will develop more detailed action plans, which other players can rate and comment upon. These too will be scored. Throughout, the game moderators have the capability of issuing rewards for particularly good ideas, which should help with quality control.

There’s also a leader board that allows you to monitor other player’s scores. There’s a risk that it might generate metagaming, however—not collaborating with a rival because you might push up his/her points total?

The biggest challenge, however, will be handling the potential traffic on the server. I don’t think any of the IT folks will be getting much sleep the day it launches publicly…

8 responses to “MMOWGLI almost ready to sail

  1. Neil 02/06/2011 at 1:18 pm

    Some people argue that the crowd might not have useful input.

    There are games in existence (eve-online being one) that are very naval like. Tactics developed based on ship capability and mixed fleets for interdiction etc. amounted to strategically placed blockades, massive confrontations, organized attack forces and small roving bands of pirates across vast stretches of space.

    Resources, like those in real life, are very limited in these games. When you get killed, you lose your ship and it’s armament, ammunition etc. and if you are on the frontier (0 sec space in eve) scarcity and the logistics to deal with it are very important. If you aren’t independently wealthy it can take days to get back into the fight.

    Like real life, in addition to organized navies (usually in alliances, much like NATO, only privately owned and ran), you have to deal with pirate forces.

    A good fleet commander in Eve can put any sized force to very effective use and play to the strengths of the available fleet to get results, break up pirate camps, etc. While rare, good fleet commanders definitely exist, and they come from all walks of life, not just military.

    I can see these types of skills being used in unexpected ways in this game that might help real life fleet commanders and admirals deal with small bands of raiding forces much better.

    It’s not always about how much firepower you have, it’s how you use it and there are fleet commanders in these games that have excellent tactical instinct and real time organizational skills. you can’t always teach that type of stuff at the Naval Academy.

    there’s nothing like real time conflict simulation to see who’s got the right stuff…

    Someone mentioned that the scale of having 14000 players might be an issue. This issue could be alleviated by splitting the player population among game world “shards” (instances, servers, or parallel universes) so that there’s no more than 150 players (or whatever number is deemed realistic) in any “shard”. Realistically you can’t have 7000 navy ships in the gulf of oman and expect a useful simulation to come out of it.

  2. Stephen Downes-Martin 01/06/2011 at 11:26 am

    I just received the following email from MMOWGLI:
    “Hi Stephen. Thanks for your interest. We have you on our waiting list. For the time being we are carefully controlling capacity to prevent overload, and we hope to increase the number of players soon. At that point we will send you email.”
    From the game blog and player instructions it seems like scoring depends on quantity oif activity (number of cards played) and popularity (how many other cards get linked to yours), plus some adjudicator opinions of what is interesting. An intriguing way of running a mass brainstorm. My initial scepticism (the normal state of my mind) is still there, but my interest is definitely growing.

  3. Rex Brynen 30/05/2011 at 9:58 pm

    As far as I understand it, the adjudicators are only a (relatively small?) part of the scoring system. However, I have to say that based on my very limited experience of only one round of mmowgli, I wasn’t convinced that it was generating anything more innovative than a BOGSAT would have.

    We’ll see how they do with 14,000. I would worry that the signal:noise ratios will diminish sharply in favour of the latter.

  4. Stephen Downes-Martin 30/05/2011 at 9:31 pm

    But it is precisely predictions that we are after … if we know a priori what are good ideas (as appears to be implied by the MMOWGLI reward structure) then we don’t need to use MMOWGLI, the adjudicatots should just have a brainstorm or a traditonal seminar style war game. If we do not know a priori then what I understand is the reward mechanism is flawed and will interfere with the game, but more importantly the ideas generated become predicitons about what could or would work. And further moves in the (prediction market) game by players playing pirates, Somali clan chiefs, etc reward the players who made the earlier predictions about what might work. What one shoudl not do is reward the ideas as they appear based on adjudicators’ opinions. That is equivalent to providing instant approval/disapproval feedback during a brainstorm.

  5. Rex Brynen 30/05/2011 at 5:47 pm

    That provides quality control on predictions, Stephen, but I don’t think it provides a mechanism for generation of innovative, non-predictive ideas in the first place. For example, what sort of initiatives would be needed to increase the willingness and capability of regional states to prosecute pirates as a matter of universal jurisdiction? To what extent can the negative externalities of using private armed guards on merchant shipping be mitigated (etc, etc)? As interesting as predictive/futures markets are, I don’t think they’re naturally well suited to that sort of brainstorming.

  6. Stephen Downes-Martin 30/05/2011 at 3:47 pm

    The answer to your question was provided several years ago when evolutionary game theory was instantiated by applying prediction markets to subjects beyond stock markets, i.e. to subjects such as counter terrorism, politics, etc. Reward and punishment is provided by financial incentives and losses (REAL ones, that REALLY focuses the mind and weeds out the losers). I do not know how closely MMOWGLI mimics a prediction market (if at all), but I suggest the Navy would have done better using prediciton market technology. See for example

  7. Rex Brynen 30/05/2011 at 2:47 am

    You’ve highlighted a critical issue in “crowd-sourcing,” Stephen: because the crowd aren’t necessarily especially knowledgeable about the issue, there’s no guarantee that their contributions are particularly insightful. How then do you encourage out-of-the-box thinking, without a sort of populist “mob sourcing” in which bad ideas win approval from participants not because they’re good, but because they appeal in other ways? Conversely, how do you avoid imposing a sort of orthodoxy on the discussions which undermines the innovative purposes of the experiment?

    I never got past Stage One in the MOVES pretest because they ended up restricting it to those with NPS accounts. I do think the reward/approval/gamification system they used for this part of the process was reasonably good, BUT that was with 75 participants—at the moment, I can’t imagine it working very well with 14,000.

  8. Stephen Downes-Martin 29/05/2011 at 5:06 pm

    Reference “Throughout, the game moderators have the capability of issuing rewards for particularly good ideas, which should help with quality control.” But what constitutes a “good idea”? One that the moderators have already decided a priori is one? One that matches the moderators preconceived criteria, or one that turns out later in the game to have good consequences and a minimum of bad ones? Having the moderators reward ideas as “good” breaks the fundamental rule of a brainstorm. It will be intreresting to see what effect it has in this environment.

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: