Wargaming Small Wars (merged thread)

[Ken White]

--or depths. I am seriously not touching that sheep...

[William F. Owen]

. I am seriously not touching that sheep...

Good. You can get your own!

[Rex Brynen]

Good. You can get your own!

No comment.

[AmericanPride]

"First, don't try to sell me on the virtues of EU2, HOI, etc. I've played those and others like them. While they are deep as games, they are paper thin as simulators. Playing EU2 constantly for a year won't make you Metternich or Marlbourogh. It only makes you an expert at playing EU2.

Point taken. I only aimed to assert that complex games modelling different theories of power, war, or what-have-you simultaneously are possible. Of course the specifics will take on a different shape in a COIN simulation -- the level of depth and complexity necessary is nonetheless attainable. Some kind of balance between "game" and "simulator" is necessary -- too strict of a simulation will not provide, in my opinion, sufficient flexibility to explore possibilities and contingencies. Too much of a game will, of course, push the realism envelope into absurdities.

Frankly, I would prefer to be able to drop people into the game for a week or two weeks at a time, with the ability to accelerate time so that they can experience 4-6 months of game-time.

That seems like a practical suggestion, given the training intent of a COIN simulator.

More than that, the cause-effect linkages need to be nebulous so that participants can't learn 'the system'. In other words, NPC reactions should be unpredictable within a certain range of outcomes. Perhaps the 'rules' should change at random intervals. We want to ensure that it doesn't teach the participants that there is a 'right way' or any such thing as a 'silver bullet' in COIN.

Having unknown variables which are difficult, or impossible, to know would provide, I think, a layer of unpredictability. GPS, for example, lets you know the basic personality traits of key leaders, but there's no certainity that Person A will, for example, accept a bribe, or having accepted it, do what was requested of them; or having rejected it, leak it to the media. I don't think a wild variable need be as dramatic or fundamental as "rule-changing", as it is possible for the program to roll a die so-to-speak.

Fourth, how do I monitor their progress and give them feedback? Frankly, I don't trust the idea that they will pick up lessons or insights on their own. I need some way of having 'the trainer' or 'the mentor' either in the sim or able to teach outside the sim.

Point taken. I should have specified my original comment or worded it differently. My mistake. Players shouldn't, of course, learn in isolation; I meant to say that players can learn different things dependent on what course they pursue provided its in a sand-box type simulation. The lessons will come out afterwards to when the players/trainers explain why the end-state looks the way it does, and whether that's desirable or not. Modelling a specific theory of insurgency will only tell players that the other models of insurgencies are invalid.

I am assuming the vast majority of entities in the game will be 'bots. Are we talking about a single, continuing universe to play in? Or will their be a number of separate 'zones' or 'instances'? If so, how do we monitor each separate zone and populate it with role-players? How do we ensure that the game world doesn't gradually morph into an environment not conducive to our training objectives?

Valid questions and I think that would depend on the model of simulation being used in the first place and practical considerations of who will be particpating, how many of them, and for how long. Some games have gazillions of bots, some have a few handful doing menial undecisive tasks, and some have none at all. I would imagine that the option to tailor these kinds of variables to the particular training objectives desired would be the most effective.

Question: do you think Operation Flashpoint is a realistic, useful combat simulator?

[AmericanPride]

If I were tasked with designing a COIN simulator, this is what it would look like:

1) Sand-box mode: an established framework to define player interaction, simulating a country similar to the Arab states; preferably a browser based system to mitigate admin bias, but a text-based game would be effective also.

2) Each player would control a faction; there would be kinds of factions: political, paramilitary, and conventional military. Factions would be measured a credibility variable. More on that later. Factions would have a prescribed set of policies/causes which can be changed at great cost to credibility.

3) Each faction type would have unique options and units. Player action would take place through these units. There would be many unit types with different capabilities, but each will be measured by common variables such as loyalty, leadership, religion, ethnicity, etc. This would allow for a wide combination of units (i.e. a US military faction and native military faction would share an Infantry unit type, but with different variables for them) and different kinds of interactions. This also allows factions to take on various shapes as the game develops. Some units could also be made available/unavailable depending upon a faction's policy/causes. Units can be killed, captured, turned, etc. Better units cans be recruited/assigned/purchased with higher credibility.

4) There would be one country divided into regions, each region with different population elements defined by similar variables as units. This would make it more difficult for Faction X to operate in Region Y if it does not share with it a common identity. However, this can be mitigated (or exasperated) by a faction's credibility rating. The greater the differences between factions and the population, the more credibility is required to operate effectively. Each region will also have a prescribed set of "wants" similar to a faction's causes/policies.

5) Holding local and national government positions (through units) would give factions more options; i.e. guiding infrastructure projects, lawmaking, etc. These options will also have an impact on a faction's credibility.

6) If two or more player actions contradict (i.e. how to use a common resource or defining local policy), the decision always rules in favor of the unit with more (fire)power (for better or for worse).

7) Game runs in real-time. No pauses. No time to think unless the player was able to create sufficient space and privilege for himself. Game ends when active players agree to call it quits.

[BayonetBrant]

If I were tasked with designing a COIN simulator, this is what it would look like:

(snip)

You might want to include some succes criteria and/or goals for each faction. Are they all after "control of population" or are they after something different? Maybe they're after 'removal of another faction' or 'control of a certain area' or 'resource'...

You need to define success criteria/victory conditions FIRST and then look at the capabilities available to accomplish it for that group. And it wouldn't be the first time a group tried to accomplish something completely outside its capabilities...

[AmericanPride]

You might want to include some succes criteria and/or goals for each faction. Are they all after "control of population" or are they after something different? Maybe they're after 'removal of another faction' or 'control of a certain area' or 'resource'.

That would be measured by the prescribed policies/causes. Depending on the intent of the simulation (training or otherwise), the player can define them upon joining, or the game gods can define them (or a mix of both). The player would then be free to shape his faction however he thinks is most effective for fulfilling the cause(s).

EDIT: Over the next couple of days, I'm going to start learning PHP. A basic COIN simulator will be an interesting project down the road.

[Ron Humphrey]

If I were tasked with designing a COIN simulator, this is what it would look like:

1) Sand-box mode: an established framework to define player interaction, simulating a country similar to the Arab states; preferably a browser based system to mitigate admin bias, but a text-based game would be effective also.

2) Each player would control a faction; there would be kinds of factions: political, paramilitary, and conventional military. Factions would be measured a credibility variable. More on that later. Factions would have a prescribed set of policies/causes which can be changed at great cost to credibility.

3) Each faction type would have unique options and units. Player action would take place through these units. There would be many unit types with different capabilities, but each will be measured by common variables such as loyalty, leadership, religion, ethnicity, etc. This would allow for a wide combination of units (i.e. a US military faction and native military faction would share an Infantry unit type, but with different variables for them) and different kinds of interactions. This also allows factions to take on various shapes as the game develops. Some units could also be made available/unavailable depending upon a faction's policy/causes. Units can be killed, captured, turned, etc. Better units cans be recruited/assigned/purchased with higher credibility.

4) There would be one country divided into regions, each region with different population elements defined by similar variables as units. This would make it more difficult for Faction X to operate in Region Y if it does not share with it a common identity. However, this can be mitigated (or exasperated) by a faction's credibility rating. The greater the differences between factions and the population, the more credibility is required to operate effectively. Each region will also have a prescribed set of "wants" similar to a faction's causes/policies.

5) Holding local and national government positions (through units) would give factions more options; i.e. guiding infrastructure projects, lawmaking, etc. These options will also have an impact on a faction's credibility.

6) If two or more player actions contradict (i.e. how to use a common resource or defining local policy), the decision always rules in favor of the unit with more (fire)power (for better or for worse).

7) Game runs in real-time. No pauses. No time to think unless the player was able to create sufficient space and privilege for himself. Game ends when active players agree to call it quits.

The majority of these reflect a particular application that my former place of employ was trying out. They are a reasonable way of approaching the issue and offer a lot of good developmental steps towards an even more in depth program. The ultimate goal would be that it should be able to be small enough for smaller groups to work yet be able to expand in scope and detail to the point at which some of the products would be able to interact with existing training systems.

Et All: I would think there are enough simulations of everything from economy, to peace negoatiations to admit that such a thing is definately doable. The very fact that it needs to avoid silver bullets makes it more so.
I think the real questions are the following.

Is it affordable? (not only in the sense of cost but also in consideration of is it worth it)

Who should build it? (this really would determine whether what you get is what you need. Way too many issues with those building it having a somewhat biased approach on whats important vs whats extra. This however is probably unavoidable)

[AmericanPride]

The majority of these reflect a particular application that my former place of employ was trying out. They are a reasonable way of approaching the issue and offer a lot of good developmental steps towards an even more in depth program.

That's interesting because the points I outlined constitute the framework my buddies and I are using to build a simulation of the Dune universe. We're building the game using PHP coding which apparently can do anything. It's cheap (but not so easy).

Is it affordable? (not only in the sense of cost but also in consideration of is it worth it)

I've always wondered what makes military simulators expensive (off-hand, I read somewhere that to operate a carrier in a simulation runs up to a million dollars a day). That's absurd. Is most of the cost in the hardware? I've found very sophisicated online simulators that are very cheap to make and operate.