DW on operational level - an external program

[Janus]

Hello,
yesterday I read through this old thread Studying the AI & fully dynamic game sessions. At some time the point was raised that Dangerous Waters is not suitable for long scale dynamic missions but only for rather small tactical engagements (see this post from GrayOwl). So far so good.

Now after I had read this Fighting Steel in the North Atlantic came into my mind.

Fighting Steel in the North Atlantic (FSNA) is a strategic campaign for Fighting Steel.
Similar to the campaign mode of GNBNA, it will allow to control task forces in the North Atlantic to either attack allied convoys or to prevent the German task forces to do so.
It is a separate program that will generate a scenario file for Fighting Steel when two task forces engage for a battle and launch Fighting Steel to play the battle.
When the battle has ended and Fighting Steel has been left, FSNA will continue the campaign.

You can download this application and play it without Fighting Steel, but you'll need a very old version of a Java Runtime Environment (1.1.x I think) and it is a very early version.

A similar application for Fighting Steel is Thunder at Sea. It is a commercial program which basically does the same as Fighting Steel in the North Atlantic. There's also a demo available which cannot be played without Fighting Steel though.

Another example is Subwolves which does something similar for Silent Hunter III in an even bigger scale.


What I want to know is if anybody has ever thought about doing something similar for Dangerous Waters?
I think technically it is feasible; DW scenarios are plain text files which can be generated automatically (SuBB and kage have done something similar before?). Of course it is not trivial and even developing a prototype of such an application would take rather months (or even years) than weeks depending on the complexity (does it need AI, is it multiplayer compatible or only singleplayer, is there a fixed area of operations or is it possible for the player(s) to load different parts of the world similar to the DW mission editor...).

How complex do you think an automatically generated scenario for DW can be? Do you see possible problems due to shortcomings in the DW AI and scenario editor features?

How open (modable) such an application could be (would it be possible to add new areas of operation easily or would this be limited to only one like in Fighting Steel in the North Atlantic)?

Do you think this is feasible in the first place (technically and does this Dangerous Waters community have the manpower and Know-How to pull this off)? Where are the limitations?

[Werewolf]

I definitely think this is a great idea since I am sick and tired of small scale little non-in-depth scenarios whose outcome doesn't matter a thing. Unfortunately I have not got a clue how to do it......what are the programs that you mention exactly doing? Are they a kind of "strategic overlay" to the mentioned games that generates missions on basis of the outcome of the previous scenarios? In that case it is kind of similar to Cuban Missile Crisis which I had for free when I bought DW.

[SeaQueen]

Originally Posted by Janus

How complex do you think an automatically generated scenario for DW can be?

It depends. I sincerely doubt that they'll be the kind of "submarine novel" scenarios that some people like. Depending on how the campaign is written they might be as complex as, say.. a good photo recon mission, though. A lot depends on how well the campaign is written and how well the role of each platform in the campaign is understood. I assume, essentially, the "scenario generation" software would be a wargame in itself and one could "jump into" individual tactical engagements as an alternate means of combat resolution.

Do you see possible problems due to shortcomings in the DW AI and scenario editor features?

Maybe. Suppose you sink a ship in DW. How does the scenario generator know that platform is sunk?

How open (modable) such an application could be (would it be possible to add new areas of operation easily or would this be limited to only one like in Fighting Steel in the North Atlantic)?

It might be worth it to create a framework where one might be able to add new theaters in the future, but I wouldn't worry initially about building every campaign into it. You could do that later. It might actually be interesting to see other people's spins on the same campaign. Like in Falcon there's two versions of the North Korea campaign that take place in different time frames. I would start with a China-Taiwan scenario or a World War III that never happened scenario since it would most likely be the largest and then figure that if it could handle that, it could handle anything else.

[Janus]

Thanks for the replies.

FSNA (Fighting Steel in the North Atlantic) works like this:
You start the game (a Java application) and play it. The area of operations is the North Atlantic and the North Sea (this would roughly translate to the GUIK area you can choose in the quick mission mode in Dangerous Waters) where you have some naval bases. Originating from these naval bases you can set up task forces and give them waypoints, set their speed etc.
As time passes by and your task forces move across the ocean you might receive contact reports for enemy convoys, coastal patrols and task forces. The sources of these contact reports may be submarines, airplanes or whatever (this is not specified in FSNA as it is only an alpha or beta version compared to the commercially distributed Thunder at Sea which is a fully developed product but the basic game principles are the same).
If one of the player's task forces comes into range of an enemy task force the game pauses and writes a scenario file for Fighting Steel based on data from the Java application. For example if your task force consisting of the Bismarck and 3 destroyers on a southerly course makes contact with a British task force consisting of the Hood and 1 light cruiser on an eastern course, FSNA writes a scenario file which resembles just these conditions and also takes weather state and other environmental conditions like time of day into account when writing the scenario file.
What you do then is start up Fighting Steel and open the created scenario and play the engagement. Afterwards the results are read by FSNA and you go back to play the operational game until the next contact occurs and the next scenario file is written.
This is the basic principle...

Of course there are many possibilities on how to design this for Dangerous Waters. Decisions to be made are for example: will the external program simulate a given fictional or historical conflict in a predefined area like for example the Falklands War 1982 or will there be something like a scenario editor? A scenario editor will be pretty hard stuff to begin with so I'd say it is best to concentrate on a fixed conflict which can be covered with the units available in Dangerous Waters which would make the historical Falklands war not feasible with the available platforms in DW. So maybe a Falklands war redux...

------------------------------------------------------------------------------

Let me brainstorm this possible operational scenario:

Operational scenario Falklands redux

let’s assume Argentina never occupied the Falklands in 1982 and thus the Falklands war never took place back then. There still was a rather small British garrison on the Falklands which enabled Argentina to occupy the Falklands 20 years later – among others – with the help of naval equipment bought from Russia. Thus we have the Kilo and Akula controllable platforms on the Argentinian side and all the west controllable platforms for the British (I don’t have any halfway reasonable explanation why the Brits use OHPs and 688Is but it’s only an example anyway and who cares if not everything is fully reasonable now…). I don’t want to talk about technical aspects like will the operational level use a hex based map or a vector map, will it be played in singleplayer or multiplayer, will it be implemented in Java or C++ and so on.

So the operational scenario starts with the Falklands (now being called Malvinas ;-)) occupied by Argentina and with a British task force with aircraft carrier supply ships and troop transporters underway in the North Atlantic to the Falklands. Peacetime ROE are in effect for both sides as the political leaders still try to find a diplomatic solution for the problem.
During this first part of the operation the goal for the British is to reach a naval base somewhere in the middle of the Atlantic where final preparations will be made for the final march to the Falklands. The Argentinians have to find and shadow the British task force. The starting point of the British task force and the naval base they are heading to are known to the Argentinians so that this phase does not turn into a massive search for a needle in a haystack. There need to be several goals each side has to achieve during this phase of the operation which can trigger a tactical phase which will be played in Dangerous Waters itself (it is nonsense to trigger a tactical phase each time both forces come into each others bubble in this phase). But due to player’s action during a tactical phase the next phase of the operation can be triggered: war. This might for example happen if Argentine subs come repeatedly too close to the British task force or to make it very obvious: if either side starts shooting the other. If the war phase will not be triggered due to any kind of this reason, it will start automatically at a given date or when the Brits have reached the naval base and departed with final course for the Falklands.
A sub-goal for this first phase for the British could be to insert Spec Ops on several points around the Falklands. Those positions are known to the Argentineans (but maybe some of them are bogus: e.g. there are 3 possible insertion zones for Spec Ops but the Brits only have to insert Spec Ops at one of them). So the Argentineans have to shadow the British task force and observer or even prevent insertion of British Spec Ops on the Falklands.

From this second phase on war time ROE are in effect for both sides. The British declare a 200 mile exclusion zone around the Falklands in which they are free to engage any unidentified contacts. Outside of this zone the ROE are a bit tighter, so that contacts need to be identified first before they may be engaged. This can be regarded as standing orders. I don’t know any specific standing orders for Argentina at the moment but it would be good to have some as well. The operational goal for the Brits is now to regain the Falklands while the Argentineans have to prevent this. There are 2 game variables for both sides which resemble national and international support for either side (morale so to speak). Let’s say this is measured in 100% for both sides individually. Every time any of the sides looses a ship, plane or anything morale will drop. If Brits kill a neutral Freighter outside of the 200 mile exclusion zone their morale will drop (international community does not tolerate neutral victims) whereas the morale and international support for Argentina would rise in this case. If any of the both countries morale/support drops below a certain threshold (influenced by a difficulty setting maybe) the operation is lost for this side. Argentina could achieve an instant total victory for example if they manage to destroy all the British troop transports before they could invade the Falklands – British morale would drop to 0%...
During the war phase there are several sub-phases or sub-goals to achieve; for example the British can only start an invasion at a limited number of fixed areas which allow an amphibious invasion. Those areas are known to the Argentines. So the first sub-phase would be to start or prevent the amphibious landing of the British infantry. Once the British infantry has landed the land war starts which the player cannot control directly but can influence by naval actions. For example both sides need to supply their troops on the islands. The Brits have to rely completely on naval convoys they have to protect while the Argentineans can use both freighters and transport planes from the nearby homeland. Both sides need a fixed number of supplies each day/week or whatever cycle. If this minimal number of supplies is not met (because the freighters are intercepted by the enemy…) the morale will drop and land war will go badly for the respective country.

Other possible sub-goals:
• British TLAM strike on Argentine airbases in South America from where naval bombers depart to bomb the British task force…
• Argentine submarine raid on British amphibious task force Falkland sound during the invasion of the infantry…

So far for an example operational scenario. I like the Falklands setting quite a lot, it has a lot of potential. I did not talk about too technical aspects now, like will there be a time limit in the tactical phase in dangerous waters (I think for certain mission types this would be reasonable); is it possible to generate such scenarios automatically and will they be unique enough, not too boring etc (with respect to what SeaQueen said about the complexity of DW scenarios and sub novels - I doubt it would easily be possible to generate something like this fully automatic...).

[Molon Labe]

I can't speak for the technical side of creating a program to do this, but I've done the work myself before. And it is a LOT of work. Getting a program to do that instead of a human would be full of awesome and win.

[Janus]

I think the silence in here says it all: no interest in this kind of project because of its complexity

[Hawk66]

I had a very similar idea this summer. Random mission campaign / framework by generating doctrine and mission files.

The generating part is not the problem but mainly the AI/missing API. I think it is not feasible to test the suff directly in DW since you have no API to connect to and manual testing was no option for me.

Perhaps it is manageable but you need unlimited time ;-) and need to be more determined as I was :-).

[Janus]

Originally Posted by Hawk66

Perhaps it is manageable but you need unlimited time ;-) and need to be more determined as I was :-).

Very true. Those are the main constraints...
Hawk66, just out of curiosity: did you get beyond the idea and already started coding something or did it stay in your head as an idea only?

[Blacklight]

You know.. this is a fantastic idea and will really open up a LOT of new possibilities with this sim ! The only catch is that the external program would have to read the mission's completed objectives somehow or otherwise, we'd have to write down the outcomes for all the mission objectives and plug them in manually which might be a pain. On the positive side, I still think that this is a fantastic idea ! It will open a lot of doors and there is an audience who would really appreciate something like this.

[Janus]

Can't completed mission objectives be read from debrief.txt located in the Dangerous Waters main folder?

[Hawk66]

Originally Posted by Janus

Very true. Those are the main constraints...
Hawk66, just out of curiosity: did you get beyond the idea and already started coding something or did it stay in your head as an idea only?

Before I forget it: Happy New Year to all of you guys :-)

No, I didn't code anything in that area.
Do you plan to go forward with your project?

[Janus]

Well I would not call it "my project". It is an idea which I wanted to present here to find out if there is any interest and especially if there are people with programming- and mission design skills who would participate in such a project.
I definitely don't have the time (and skills) to pull this off on my own. Scenario generation for DW would only be a part of the application (maybe even an optional one if there would be automatic resolution of combat events...). And therefor you first have to reengineer the structure of DW scenario files, I don't know if anyone has done this before and did document it in a useable format?
Then there is the whole operational application itsself of course...

[Hawk66]

Re engineering of the structure of a scenario is not tough...I've done some 'tests' in the past...the keywords/structure are very clear and it is pure text-based. It's almost readable as a XML schema.

My 'dream' was/is to generate (some) doctrine files and bound it to the scenario (it is possible, there is a DOCTRINE(?) command, which is not accessible in the editor but it works. With this approach, you could realize a carrier group with a moving patrol screen. It is not 100% reliable but anyway it would be a new experience IMHO.

I think the investment could be worthwile but you would need a couple of guys doing it and the question which bothered me most: What is the remaining lifetime of DW?