Fun questions to ponder about the future of subsims

[Pillar]

Imagine a strategy game that uses Dangerous Waters (or something like it) to resolve the naval portion of the overall gameplay. How do you think would be the best way for the naval game to interface with the strategic world?

For example, a persistent environment where vessels are crewed in real time, perhaps the same crews (like "guilds") crewing the same vessels...

Or maybe something like harpoon? (AI vessels directed by single people, either one player per vessel, fleet, or whatever.)

A mix of the two, where situations generate "encounters" or battles in DW? (In this case, the disadvantage is that getting an encounter is "meta" information that influences game behavior - e.g.,who is going to start an "encounter battle" at full throttle/transit and remain that way?)

In terms of the state of DW right now, what modification would be most well suited to multiplayer gameplay/battle resolutions?

What size is the community of multiplayer DW players right now?

[Molon Labe]

That's exactly what I want. I think a tactical sim like DW should have a campaign engine (http://www.subsim.com/radioroom/showthread.php?t=124314&highlight=campaign+engine) that tracks events at the strategic/operational level--something like a fleet command interface or the campaign engine in Falcon 4.0. This would allow you to use the tactical sim to resolve combat actions generated by player actions on the strategic sim.

How do you think would be the best way for the naval game to interface with the strategic world?

For example, a persistent environment where vessels are crewed in real time, perhaps the same crews (like "guilds") crewing the same vessels...

Having specific crews and the like would require a larger player pool than we can count on for modern naval sims. Would be nice to have.

Overall I think the best solution is to have a stategic interface that allows you players to log onto a persistant server (of course the clock would only run when people are in fact logged on), and then begin a tactical simulation in a platform and at a time of the players' chosing. AI reports would be provided to the stategic interface from platforms regarding contacts (sensors or flaming datum/distress calls) and such. And don't underestimate the value of anticipated contact (such as when platforms are nearing an objective). You don't get the "meta information" problem with this because you don't organize an encounter independently of the stategic sim. The tactical and strategic sims run in parallel, and the players move from one to the other as they please.

[SeaQueen]

That's exactly what I want. I think a tactical sim like DW should have a campaign engine (http://www.subsim.com/radioroom/showthread.php?t=124314&highlight=campaign+engine) that tracks events at the strategic/operational level--something like a fleet command interface or the campaign engine in Falcon 4.0. This would allow you to use the tactical sim to resolve combat actions generated by player actions on the strategic sim.

The thing is that for some reason, everyone who makes a naval sim locks themselves into the idea that they HAVE to make it capable of modeling any naval conflict in the world. And it's true that one of the cool things about Harpoon is that it can do that. The last naval sim I remember that was built around a specific theatre was Microprose's Red Storm Rising. Falcon's engine is neat because it doesn't HAVE to use the North Korean campaign, but you have to do something to switch it. It loads a whole different database to support it. I wonder why they don't do a similar thing with a naval simulation.

The other thing I wish they'd do is figure out how to model the interaction of ground units and naval units better. I'd also really enjoy seeing some amphibious operations.

[Molon Labe]

The thing is that for some reason, everyone who makes a naval sim locks themselves into the idea that they HAVE to make it capable of modeling any naval conflict in the world. And it's true that one of the cool things about Harpoon is that it can do that. The last naval sim I remember that was built around a specific theatre was Microprose's Red Storm Rising. Falcon's engine is neat because it doesn't HAVE to use the North Korean campaign, but you have to do something to switch it. It loads a whole different database to support it. I wonder why they don't do a similar thing with a naval simulation.

The other thing I wish they'd do is figure out how to model the interaction of ground units and naval units better. I'd also really enjoy seeing some amphibious operations.

It's not the "any conflict" that I'm after so much as persistence and context.

[-GrayOwl-]

It's not the "any conflict" that I'm after so much as persistence and context.

This Is No Global Sim Action.

Simply - this is tactical , operative limited area...

Never compare Harpoon and DW is is simply incompatible....

[Castout]

This Is No Global Sim Action.

Simply - this is tactical , operative limited area...

Never compare Harpoon and DW is is simply incompatible....


Hi GrayOwl what happened to your old account I see now your post count is starting all over again to 1?

Nice to see you back:salute:

[SeaQueen]

It's not the "any conflict" that I'm after so much as persistence and context.

And I think that's fair. That's what I liked about Red Storm Rising, because to a certain extent the campaign relied on scripted "random events" but the ground war was like some sort of random walk.

These days I think they could do a lot better than Red Storm Rising. Maybe if they built the game around a hypothetical China-Taiwan-US conflict set 20-30 years in the future. It'd be complete science fiction, but it'd be fun. There'd be the capability to load other campaigns, but the starting point with be that one.

I guess you'd have to build the campaign engine first, and then go back and build in the ability to "jump into" specific platforms. The thing is, since ships can be at sea for weeks or months at a time, you're probably not going to want to play just one vessel for the entire campaign. You'd probably want to be able to say, "today I'm going to be a 774 over here" and then when that gets boring, you could jump into the AEGIS cruiser way out because TBMs are coming over the horizon, or a DDG-1000 attached to an ESG because Marines have called for naval surface fires.

Of course, if you WANTED to you could just stay in one vessel and drive doughnuts in the ocean for days at a time between engagements. It's up to you.

There should also be the capability like in Falcon for players to make their own scenarios.

[goldorak]

Having a dynamic campaign a la Falcon 4 is not going to happen.
Be realistic. Even in flight sims, no one cares anymore for that kind of experience. Its sad but true, Falcon 4 was the pinacle of military flight simulation. Nothing since can even compete in the same league.
The real deal would be to have a dynamic mission generator, something much simpler than the Falcon 4 engine, something like the RSR engine. Updated for a modern scenario but still that kind of engine would be achievable with the economic constraints of naval sim game development.
Any other option is purely academic.

The SH series is an exception because WW2 has still a kiind of romantic vision. Also the fact that subs stayed most of the time on the surface makes the "graphics" part much more interesting and appealing to gamers.
There is nothing wrong with rendering 200-2000 ft deep abyss but since you're inside the tin can its not much of an appeal. Now if on the other hand we could have the interior of the subs modelled with precision, and the stations etc... rendered realistically, even if you never saw the outside who would care. You would be rivited to sonar screens, maps , tactical displays etc... The only sim I can think of that really nailed down this aspect was Sierra Fast Attack. A pity that developers never learn to copy or be insipred by the good things in other people's products.

[Molon Labe]

The thing is, since ships can be at sea for weeks or months at a time, you're probably not going to want to play just one vessel for the entire campaign. You'd probably want to be able to say, "today I'm going to be a 774 over here" and then when that gets boring, you could jump into the AEGIS cruiser way out because TBMs are coming over the horizon, or a DDG-1000 attached to an ESG because Marines have called for naval surface fires.

Of course, if you WANTED to you could just stay in one vessel and drive doughnuts in the ocean for days at a time between engagements. It's up to you.
Eh, that's what time compression is for. I don't see any reason why you couldn't play the whole war as a single platform, as long as you choose one that's going to have some action. You know, don't choose the sub that's doing a barrier outside a base that may or may not deploy any boomers, choose the one with some offensive ASUW and strike tasking, for example. The total number of "encounters" might be low, but that's just how it goes.

Having a dynamic campaign a la Falcon 4 is not going to happen.
Be realistic. Even in flight sims, no one cares anymore for that kind of experience.
Wow. I've only been gone from the Falcon 4 community for about a year, but when I was there all anyone ever played was campaigns, both offline and in MP. I have a hard time believing that has changed so quickly.

I don't think it's an unrealistic expectation in any case. Getting a new sim anytime soon is the pipe dream. But if we get one, I'd be surprised if it didn't have a campaign feature considering the disappointment caused by the lack of a true campaign engine in DW and the success of the campaign based SH series. I think a dynamic campaign engine is simply part of the consumer's basic expectations at this point in time.

[goldorak]

Wow. I've only been gone from the Falcon 4 community for about a year, but when I was there all anyone ever played was campaigns, both offline and in MP. I have a hard time believing that has changed so quickly.

I don't think it's an unrealistic expectation in any case. Getting a new sim anytime soon is the pipe dream. But if we get one, I'd be surprised if it didn't have a campaign feature considering the disappointment caused by the lack of a true campaign engine in DW and the success of the campaign based SH series. I think a dynamic campaign engine is simply part of the consumer's basic expectations at this point in time.


No you misunderstood me. What I meant to say was that no commercial developer has taken upon himself to make a product that not only equals but outshines Falcon 4. Of course players love Falcon 4, until something else comes along that delivers a 360 degree better experience that is.
And something like this is not going to happen, just look at all the military simulations.
Black Shark could have been a testbed for a dynamic campaign in the caucasus region, instead all you get are scripted missions and an editor.
There is a new WWI military air simulation. Same problem, all scripted missions and multiplayer. No dynamic campaign, not even a casual mission generator. I mean the games from thirdwire have a primitive random mission generator and no AAA software house can do something like that ?
It really means they are not interested. Why ? Who knows. Maybe players prefer simple multiplayer instead of these epic campaigns.

[Dr.Sid]

Interesting topic.

I think for most sims, dynamic campaign is another layer, another heap of work. Campaign has larger scope then missions. You play missions, then you jumo into campaign, which generally generates new missions.
It is true for flightsims, and for example for DW too.
In my sim I can have even missions with global scope, which makes campaign just a huge mission, with half of the fleet in SF bay, and other half near Taiwan.
Huge campaign is then just a solving of performance issues, by for example having some passive mode for units which are not in hot zones at the moment.
I too believe in freedom. In this model you don't even have to split campaign into missions. You could play a sub commander for while, empty its ammo racks, then switch to different sub (or destroyer or carrier) and just order the sub you have just left to head for rearm. Or you could be sitting on your flagship all the time and just give orders Fleet Command style, or even at higher level.

[SeaQueen]

Eh, that's what time compression is for. I don't see any reason why you couldn't play the whole war as a single platform, as long as you choose one that's going to have some action.

Time compression would be an essential too. These days, the US Navy is sufficiently small that EVERY ship is going to see some action, it just depends on when and what kind of action it's going to be. There's definitely going to have to be a campaign timeline, though, so in the very beginning of the campaign, the action might be delivering special operators, then as hostilities start it'd more in ASuW and strike, then as there are fewer and fewer surface platforms, it'd taper off in the ASuW arena, and finally as marines are landed it'd be about naval surface fires, mine warfare, special operations, and what not.


You know, don't choose the sub that's doing a barrier outside a base that may or may not deploy any boomers, choose the one with some offensive ASUW and strike tasking, for example.


That's another way players could control their "fun" factor. Ultimately, every mission in the campaign should have a challenge of it's own. In the campaign you wouldn't task an SSN to hold SSBNs at risk unless at least one was going to be at sea during the conflict. It's missions like that where I think it'd be best to be able to "jump in and out" of the platforms. In Falcon you could do that and I think there was a lot to be said for it. Air war is such that I didn't use the feature a whole lot, but I suspect in a naval sim, it'd be more valuable.

[Pillar]

What about DW as it stands now (with a mod, to be sure)...?

A campaign layer could easily provide inputs for scenario generation in DW right now, it just would have to be built by hand.

The last time I played DW multiplayer, it lasted about 4 hours of real time (you can't time-compress, or couldn't at the time - can you?) and it was pretty dull stuff.

Do you guys foresee any problems getting people to link up and play out such long scenarios on a regular basis?

[Molon Labe]

Why do we have to try to forsee what is already occurring with DW? Or did you mean forsee problems with the campaign engine? Because problems such as this are exactly what a campaign engine addresses--by allowing saves, reconnects, time compression, and the ability to break up a conflict into parts.

[Pillar]

What I mean is, if the campaign tracker existed but the battles still had to be resolved in DW *as is* (with mods), would there be problems getting battles resolved? If so what problems? (I gave the example of the time it takes to play out a mission in DW)

Last time I played MP DW we couldn't time compress... I'm not sure if that is still the case or how it works, so I'm curious about that. Saves, reconnects... are these not in DW MP now?

Thanks

[Molon Labe]

What I mean is, if the campaign tracker existed but the battles still had to be resolved in DW *as is* (with mods), would there be problems getting battles resolved? If so what problems? (I gave the example of the time it takes to play out a mission in DW)

Last time I played MP DW we couldn't time compress... I'm not sure if that is still the case or how it works, so I'm curious about that. Saves, reconnects... are these not in DW MP now?

Thanks
Saves are available for single player only; no time compression, no reconnect. But if you did have a campaign engine--such as something like fleet command--it could easily be designed to create a save file that records the scenario, platforms, positions, messages, etc that could be loaded again at a later time. The original game already did that for SP, why not have the same capability if it's used as a campaign engine?

Reconnection would be a given when you have a persistent server that players can log on to and off of at will. That's definitely a big step up from what we have, since the lobby would have to be integrated into the strategic level simulation.

As for using DW "as is" with a crude engine being used for scenario generation, I think the biggest problems you'd get would come from having tactical sim and strategic sim running in series rather than parallel (Series = stategic sim runs, generates tactical scenario, tactical scenario is played to resolution, data fed back into strategic sim, stategic sim resumes; Parallel = tactical sim runs at same time as strategic sim; both sims communicate with each other and update accordingly; multiple tactical encounters can be run simultaneously). You would have issues if there were multiple encounters possible in the same general time frame, because a second encounter might get "skipped" during the time it takes to complete the first, or you could have a problem involving what platforms to place in the tactical scenario based on proximity that may or may not have a chance to fight depending on how long the encounter lasts. This might be a minor issue but it looks like it's begging for glitches.

You would also have the "meta information" problem, because without the capability to switch between the strategic and tactical sims at will and run time compression, the point at which the tactical scenario is developed has to be deliberately chosen based on some fixed criteria---detection, potential detection, etc. And you also have to place opponents in specific platforms on the other side, which tells them who the hammer is about to drop on. Again, this is solved by running the sims in parallel instead of in series.

As far as player convenience goes, I don't think this is a major issue with this type of format. When I ran the GDT (http://www.commanders-academy.com/forum/forumdisplay.php?f=70) (which was essentially exactly what we're discussing here, but with me acting as the crude campaign engine/scenario generator) we had fixed match times of 2 hrs and completing scenarios in that time was not a problem. Meta-information was a huge problem, though.

[Kazuaki Shimazaki II]

They actually made a game that's like this sometime ago ... an entire series known as Great Naval Battles. I played IV, and I enjoyed it ... even though I kept getting sunk ... things seem to go well, then all of a sudden my ships get hit and my formation slows ... anyway...

It allows for almost all the functionality requested. You can play in all kinds of different positions, from Fleet Commander down to being the commander of a single turret.

All it lacks is multiplayer.

The Campaign Engine: While I hadn't played Falcon 4 (it won't run on my old computer, and I never found time to play it on the new - nice manual though), I think one should not underestimate how much harder this will be than even Falcon 4.

Falcon 4's campaign engine deals with enormous formations and simple missions, which makes it easIER.

NK attacks SK. Just deciding THAT locks in a lot of decisions, such as the general strategic and operational directions the formations will take, and the list of missions for the units of both sides.

OK, so we have this mob that's supposed to be hundreds of thousands of troops and a few thousand tanks. Because the pilot's view is very limited, it is easy to use relatively simple formulae and norms to plausibly re-enact all those "tens of thousands of units" promised in the manual. Entire SAM units and bases can be aggregated and the algorithm for calculating attrition can be as simple as dice rolling. The advances of entire divisions against entire entire brigades defending can be similarly aggregated and resolved without wasting too much computer power. Add a dice of randomness and the pilot is not likely to get a continuous enough view of enough of the situation to know the result may be all BS.

When it is time to throw another mission to the human pilot, just check where the FEBA is and throw him a mission aiming at the right spot. Or task him towards a rear mission.

Generating the tactical mission itself when the pilot gets close is really a relatively sophisticated Random Mission Generator. De-aggregate the units closest to the pilot and plop down some targets and some AA, a mix of randomness and following some OPFOR textbook. Bingo, one dynamic mission.

Ironically, because there are fewer units, spread farther apart in the naval world, it actually becomes harder to write a good campaign engine. Starting on the tactical level, you can't solve most of the problems in an abstract, aggregated manner - there are too few ships, and they are often spread too far apart for aggregation to bring reasonable results. You have to solve them de-aggregated, which means AI must be running for each naval unit not currently being manned.

Further, on a higher operational level, the writing of plausible missions becomes very much harder. We have trouble getting entire human staffs to think operationally or strategically. Let alone a computer.

[Molon Labe]

Falcon 4's campaign engine deals with enormous formations and simple missions, which makes it easIER.

NK attacks SK. Just deciding THAT locks in a lot of decisions, such as the general strategic and operational directions the formations will take, and the list of missions for the units of both sides.

But, as SeaQueen's pointed out, there's no reason why the stategic-tactical naval sim can't be theatre specific. You could do a campaign where it's already decided that China attacks Taiwan, for example.

OK, so we have this mob that's supposed to be hundreds of thousands of troops and a few thousand tanks. Because the pilot's view is very limited, it is easy to use relatively simple formulae and norms to plausibly re-enact all those "tens of thousands of units" promised in the manual. Entire SAM units and bases can be aggregated and the algorithm for calculating attrition can be as simple as dice rolling. The advances of entire divisions against entire entire brigades defending can be similarly aggregated and resolved without wasting too much computer power. Add a dice of randomness and the pilot is not likely to get a continuous enough view of enough of the situation to know the result may be all BS.So if it's possible to reduce direct fire, indirect fire, SAM and AAM shots to die rolls in Faclon, why can't torp and missile shots be abstracted into a naval sim? Or why can't these actions be resolved as they were in FC?



Generating the tactical mission itself when the pilot gets close is really a relatively sophisticated Random Mission Generator. De-aggregate the units closest to the pilot and plop down some targets and some AA, a mix of randomness and following some OPFOR textbook. Bingo, one dynamic mission. Um, the units present in Falcon aren't random. The AA that is there is there because it is there. :) The position and movement of these units are tracked by the campaign engine.


Ironically, because there are fewer units, spread farther apart in the naval world, it actually becomes harder to write a good campaign engine. Starting on the tactical level, you can't solve most of the problems in an abstract, aggregated manner - there are too few ships, and they are often spread too far apart for aggregation to bring reasonable results. You have to solve them de-aggregated, which means AI must be running for each naval unit not currently being manned. Fleet command could handle it. And Falcon kept track of hundreds of ground units, each with dozens of invididual vehicles--the destruction of which was accounted for invididually (ie you destroy the SA-9s with a tank batallion and those SA-9 are still gone when you attack next time).


Further, on a higher operational level, the writing of plausible missions becomes very much harder. We have trouble getting entire human staffs to think operationally or strategically. Let alone a computer.The difficulty of strategic thought is what makes it interesting and worth playing.

[Pillar]

Really interesting stuff, thanks for the discussion here.

[Kazuaki Shimazaki II]

But, as SeaQueen's pointed out, there's no reason why the stategic-tactical naval sim can't be theatre specific. You could do a campaign where it's already decided that China attacks Taiwan, for example.

Being theatre specific would help, but would also limit the time we can play it. Sooner or later we'll want to do something new.

So if it's possible to reduce direct fire, indirect fire, SAM and AAM shots to die rolls in Faclon, why can't torp and missile shots be abstracted into a naval sim? Or why can't these actions be resolved as they were in FC?

I'm not too familiar with FC, but in general, all sims but Falcon 4.0 simulates units in "full AI form", one step in realism (and calculation work) down from "full human form" where there's a difference (you know, why it always seems the enemy planes can ignore some rules you are stuck to). This tends to create problems when the number of units start to increase, and why most sims have real trouble simulating on scales like that big Falcon 4.0 campaign.

Um, the units present in Falcon aren't random. The AA that is there is there because it is there. The position and movement of these units are tracked by the campaign engine.

Fleet command could handle it. And Falcon kept track of hundreds of ground units, each with dozens of invididual vehicles--the destruction of which was accounted for invididually (ie you destroy the SA-9s with a tank batallion and those SA-9 are still gone when you attack next time).

FC, like Harpoon, has the advantage of being a command sim, with no real expectation that you'll actually "visit" one of the units. That allows liberties to be taken in the whole model that would not be tolerated in a modern day sim. For example, in Harpoon II, I click on the Speed Adjustment and change "Cruise" to "Flank". BAM! The plane's at its new speed, with new estimated times to match. This won't be tolerated in even the crummiest flying sim.

Try this with your computer. First, boot up FC and have it run a complicated scenario with lots of moving and fighting units. Now start up DW. That's basically the kind of workload you can expect for your computer. If your computer is actually powerful enough not to collapse, neither model is being run at the maximum capability (realism) of modern computing hardware (as an analogy, you can probably run Harpoon 1 and 688 Hunter Killer at the same time without too many problems, because they so underutilize your computer).

So how does Falcon do it? AFAIK, it is like JPEG, smudging out the detail where you don't notice it. Take your tank battalion. The Campaign engine knows that it has 31 tanks and 4 SA-9s. Since this is not a tank battle game, top notch realism for the tanks is not necessary even under the closest scrutiny, while the SA-9s need to be handled more carefully. Nevertheless, if it seriously tries to run those 31 tanks and SA-9s as separate units even at relatively low realism, it'll take a disproportionate amount of computing power, and when one counts all the NK battalions the computer would crash. So instead, it treats it as a "tank battalion with SAMs", and assigns it an area of existence, another area of effect corresponding to its weapons range, and attack and defense values. This "blob" is now advanced as a single piece.

Let's say it runs into a SK tank company (14 tanks if I remember the composition of a ROK company right). Basically, that fight would be two "blobs" engaging each other. Since you don't see the fight there's no need to handle it in high detail, so it is resolved relatively easily. Say the computer figures the SK company is 100% gone, while the NK battalion loses ~30% So now, the computer eliminates the SK blob, and writes down the NK blob as having 21 tanks (with appropriately reduced ATK and DEF values). It might even remember to generate and store 24 sets of coordinates, so in the unlikely event you fly over the area you get to see 24 T-55 and M48 corpses. It can also handle it by making TWO blobs, and generating it as you fly in (after all, you never saw where they died).

The 21 tank NK blob continues south. Now it gets attacked by a squadron of aircraft. Same thing. Blob on blob, and both sides take some losses. The NK blob is now down to 15 tanks and 4 SA-9s, and its numbers re-written accordingly. Finally, you get tasked to stop it. When you get close enough, the computer breaks out the blob into individual tanks and SA-9s. You get a very realistic experience and kill 4 tanks and 2 SA-9s. You now leave.

The computer will store the coordinates for your 4 killed tanks and 2 SA-9s so it can "remember" to draw them in the next time you visit in the right places. Let's say the tank battalion stops. Now, you SAW where they were, so the computer has stores the coordinates for all the 13 surviving systems as well. Just storing coordinates doesn't take that much computer power - it is moving and working out their next moves that's the big drain...

Let's say in its stopped position, ANOTHER AI unit attacks it. Since you aren't there to witness it, the battle is resolved as a blob. NK losses: 2 tanks. Presumably the tank battalion would move now but if it didn't, it isn't very hard to randomly choose 2 sets of tank coordinates and turn their status to "dead and fixed".

The tank battalion (now really a heavy company supported by a section) gets back on the move. You get the mission to bomb it again, but by then it is far from its original place. When you get close, the computer "breaks" the blob of 9 tanks and 2 SA-9s out - you have no previous references and buy whatever is deployed.

The whole aggregation-deaggregation algorithm is probably more sophisticated and clever than described here, but that's the basic idea. Any other method would un-necessarily detract from your Air Combat Sim experience.

The problem is that such tactics don't nearly work as well in a naval sim. Naval warfare in some ways is not as neat and tidy as land warfare with its massed formations. A battalion of T-72s might be aggregated into one blip taking up a certain geometrical area. Ships just don't take to being aggregated as easily.

The difficulty of strategic thought is what makes it interesting and worth playing.

It's fun for you! It is hard on the designers, who probably don't understand the intricacies of strategy themselves

[Dr.Sid]

On the other hand, naval combat alone would have to handle much less units. And over long time scale. Battles 'all US against all China' are not likely. At any time, jut a few ships would be actually engaged in battle.

With land units thrown into it, it's all different. Carrier can run bombing campaign over lets say Iraq day and night, always having half the planes airborne. Those targets will move, and if you want to be realistic enough, support tracks are moving behind them, and you better simulate every single soldier, as long as he can pickup stinger or RPG.

I thinking about naval sim with limited ground units simulation. Most of the time you would want to win air & sea before going to land. So winning air and sea could be the content of the 'global scale to single station' naval simulation. If you succeed, you can boot Arma II :arrgh!:

[Molon Labe]

Being theatre specific would help, but would also limit the time we can play it. Sooner or later we'll want to do something new.
That's what expansion packs and mods are for.



I'm not too familiar with FC, but in general, all sims but Falcon 4.0 simulates units in "full AI form", one step in realism (and calculation work) down from "full human form" where there's a difference (you know, why it always seems the enemy planes can ignore some rules you are stuck to). This tends to create problems when the number of units start to increase, and why most sims have real trouble simulating on scales like that big Falcon 4.0 campaign. {snip}

The problem is that such tactics don't nearly work as well in a naval sim. Naval warfare in some ways is not as neat and tidy as land warfare with its massed formations. A battalion of T-72s might be aggregated into one blip taking up a certain geometrical area. Ships just don't take to being aggregated as easily.They call it "the bubble" sometimes. Inside of "the bubble" around a player in the 3D world (tactical sim), the fidelity of the simulation is at maximum. Outside the bubble in the rest of the 2D world (operational sim/campaign engine) everything is simplified.

I don't understand why twice now you've said that you can't do the same with a naval sim. Can you elaborate on that a bit?

From where I'm sitting, what I want from a campaign engine is to track the location and status of platforms so that it can place platforms into the tactial sim at the correct position, with the correct loadout, and with the correct level of damage. That isn't very processor intensive. The other important feature would be implementing detection rules. That's going to take some power, but if it gets out of control corners can be cut--you can create a "bubble" around platforms close enough to be potentially detected where sensors are modeled in full fidelity, and outside of that there's no need to model them at all. So you end up with only a small % of the platforms being tracked consuming processing resources at any given time.

Anything beyond that is a bonus, but F4 did a bit more and FC did a whole lot more, so more is certainly possible.


FC, like Harpoon, has the advantage of being a command sim, with no real expectation that you'll actually "visit" one of the units. That allows liberties to be taken in the whole model that would not be tolerated in a modern day sim. For example, in Harpoon II, I click on the Speed Adjustment and change "Cruise" to "Flank". BAM! The plane's at its new speed, with new estimated times to match. This won't be tolerated in even the crummiest flying sim.Well, in FC it did take a litlte bit of time to change speed. Even assuming speed and course changes are instantaneous, this is not a problem at the strategic level. There is absolutely no need to have accurate physics modeling outside of the tactical sim. That's one of those things that's just a bonus if you get it. You don't need it for the strategic sim/campaign engine to do it's job.


With land units thrown into it, it's all different. Carrier can run bombing campaign over lets say Iraq day and night, always having half the planes airborne. Those targets will move, and if you want to be realistic enough, support tracks are moving behind them, and you better simulate every single soldier, as long as he can pickup stinger or RPG.
In the tactical sim, yes, but in the strategic sim you just need to track the ground unit and account for who is in it. So yes, there is a record that the individual solder exists, but he isn't simulated as an individual until he gets placed in the tactical sim.

[Nexus7]

This is an exciting topic.

What I am saying is probably way off-topic, as it looks too much forward and without almost any substance.

What I am pondering, is the possibility, in the future, to model the whole world into a super-simulation/game but with a virtual world.

prototype 1.0 :

political situation:
- Civilization-like
simulators:
- Land: BF1942-like
- Air: dunno.
- Naval: DW-like

The "game" would run 24h/day 365days/year online.

When i log into the game i can choose to play at strategic level (Civilization-like) with random world and nations, or enter into more detail, i.e. sonarman into some multistation sub of some virtual nation...

As a comparison, something like the game "Real Life" but for politics/war :D

[EDIT] The player should choose a virtual nation to stay with for all the time being...

[Kazuaki Shimazaki II]

That's what expansion packs and mods are for.

We are still waiting for the expansion pack from Sonalysts. Look, realistically this game would be niche if it was ever made. You can't just ASSUME an expansion pack would be made. :D

In the event an expansion pack won't be made, it'll be a real pain to make a new campaign in the fan community. One thing about complicated scenarios (the essence of a "dynamic campaign") is that it gets much less customizable, at least to most of the public. It'll take a LOT of blood to work out how to get it all done. Think of how hard LWAMI can be, and that was poking at text files, or making a DW scenario.

They call it "the bubble" sometimes. Inside of "the bubble" around a player in the 3D world (tactical sim), the fidelity of the simulation is at maximum. Outside the bubble in the rest of the 2D world (operational sim/campaign engine) everything is simplified.

Yes. The problem with "bubbles", especially bubbles where outside is a highly aggregated world, is that they work better if they have some warning time. In Falcon 4 or some other flight sim, you are ultimately locked into a particular plane. The computer has plenty of warning time to get everything set up while you fly there. This game, apparently it should allow you to just go in and come out different platforms almost at a whim...

I don't understand why twice now you've said that you can't do the same with a naval sim. Can you elaborate on that a bit?

From where I'm sitting, what I want from a campaign engine is to track the location and status of platforms so that it can place platforms into the tactial sim at the correct position, with the correct loadout, and with the correct level of damage. That isn't very processor intensive. The other important feature would be implementing detection rules. That's going to take some power, but if it gets out of control corners can be cut--you can create a "bubble" around platforms close enough to be potentially detected where sensors are modeled in full fidelity, and outside of that there's no need to model them at all. So you end up with only a small % of the platforms being tracked consuming processing resources at any given time.

Oh, the detection rules ain't my worry. The problem is the whole campaign engine.

If you just want them to move and the locations and status to be stored, you don't need a campaign engine at all. You can just draw a complicated prescripted scenario. The problem is that you want them to react, and have randomness, and all that, without blowing up the CPU so you need a dynamic campaign engine.

Frankly, there several reasons:

Let's start with relative importance. In a fighter sim, even if you can control a few wingmen or strike package, your realistic ability to affect the operational or strategic level is quite small, no matter how good you are. You can wipe out a tank battalion, but there are plenty of those. You can put out ONE airfield, but it is only one and it can be repaired soon. So there's no way you can quickly devastate the script of the campaign engine (though, of course, if you do well consistently, the campaign engine is supposed to eventually reward you by showing you retreating NKs).

In a naval game, however, the chances of getting to operational or strategic effect is much larger, simply because there are so few units. Blasting away a ship at a particular position may be worth more than if you managed to somehow blow away an entire NK division.

When such disasters happen, the situation may be salvageable (so you will expect the AI not to collapse) but clearly some retasking has to occur. With a conventional ground war scenario (like Falcon 4), many units have the same general objective (advance south) so it is easy to retask. Even if a division is blown away, just stuff a second echelon division into the hole and tell it to keep pushing south. In a naval game, all the available units are probably already tasked on wildly differing assignments (one may be patrolling, another running in for a strike ... etc), and it is much harder to decide how best to redeploy the remaining assets. So not only are you much less likely to disrupt the campaign engine's workings in Falcon 4, but it is also generally easier to decide how to best compensate for it.

Further, aggregation is a lot easier with Falcon 4. Ground units especially in conventional war, tend to be composed more homogenously, are deployed over a smaller area and the deployment also tends to be more regular. This makes it much easier to do simple aggregating for combat modelling than ships. Ships are often so far apart, so irregularly spaced and so heterogenous in their abilities, that the aggregation assumption collapses. It is much more necessary to treat them as points like they actually are, rather than a aggregate blob.

Anything beyond that is a bonus, but F4 did a bit more and FC did a whole lot more, so more is certainly possible.

As I said, FC had the big advantage that it didn't have to be the little tactical sim at the same time it is the strategic sim. This new game of yours will REQUIRE this.

[Molon Labe]

We are still waiting for the expansion pack from Sonalysts. Look, realistically this game would be niche if it was ever made. You can't just ASSUME an expansion pack would be made. :D

In the event an expansion pack won't be made, it'll be a real pain to make a new campaign in the fan community. One thing about complicated scenarios (the essence of a "dynamic campaign") is that it gets much less customizable, at least to most of the public. It'll take a LOT of blood to work out how to get it all done. Think of how hard LWAMI can be, and that was poking at text files, or making a DW scenario.
LP expanded F4 with AF (Balkans). The community has made theatres for ODS, Vietnam, Isreal, and others.

DW as is already has a global map. Give the players a campaign editor tool that allows them to create/edit OOBs in new areas of the world and you're golden. It's just a matter of research and gruntwork.



Yes. The problem with "bubbles", especially bubbles where outside is a highly aggregated world, is that they work better if they have some warning time. In Falcon 4 or some other flight sim, you are ultimately locked into a particular plane. The computer has plenty of warning time to get everything set up while you fly there. This game, apparently it should allow you to just go in and come out different platforms almost at a whim...
You're wrong. F4 allows you to jump into and out of aircraft at will. You can jump into an aircraft that is already in combat (not advisable, but certainly possible). The CPU has no warning to this at all, other than the player clicking the button and the 3D world immediately loading.



Oh, the detection rules ain't my worry. The problem is the whole campaign engine.

If you just want them to move and the locations and status to be stored, you don't need a campaign engine at all. You can just draw a complicated prescripted scenario. The problem is that you want them to react, and have randomness, and all that, without blowing up the CPU so you need a dynamic campaign engine.
A single scenario is absolutely not capable of doing this, or we'd have it already! If you put the entire campaign into a single scenario, it would be completely unplayable. You have no time compression and you're stuck in a single platform which may not ever be involved in the fighting. No one in their right mind would want to play that, sitting at their computer for weeks or months without any breaks looking at an empty sonar screen.

The strategic sim allows you to follow developments in the war and order your forces into battle, while the tactical sim allows you to actually fight those battles, once they begin. Again, to accomplish this, the core capabilities are tracking platforms (location and status) and detection rules.

Randomness, reaction, etc. doesn't require any CPU power...that's provided by your team and your adversary. Want to add SP capability...fine...that's bonus territory. But we already have AI doctrines and scripted reactions to events in DW as is, and it is not processor intensive. There's no reason why it would be processor intensive in a broader sim. All stategic sims have this too, even primitive ones like PTO 2. The limitations come from the quality of the doctrines and scripts, not processor resources.


Frankly, there several reasons:

Let's start with relative importance. In a fighter sim, even if you can control a few wingmen or strike package, your realistic ability to affect the operational or strategic level is quite small, no matter how good you are. You can wipe out a tank battalion, but there are plenty of those. You can put out ONE airfield, but it is only one and it can be repaired soon. So there's no way you can quickly devastate the script of the campaign engine (though, of course, if you do well consistently, the campaign engine is supposed to eventually reward you by showing you retreating NKs)...........

Doesn't matter. The sim doesn't only track your actions, but the actions of all forces on the OOB as a whole. And the success of those forces most definitely does effect operations going forward, which means the AI has to react to it just the same whether or not the events were caused by human players.


Further, aggregation is a lot easier with Falcon 4. Ground units especially in conventional war, tend to be composed more homogenously, are deployed over a smaller area and the deployment also tends to be more regular. This makes it much easier to do simple aggregating for combat modelling than ships. Ships are often so far apart, so irregularly spaced and so heterogenous in their abilities, that the aggregation assumption collapses. It is much more necessary to treat them as points like they actually are, rather than a aggregate blob. You'd track ships individually, there is no need to aggregate them. Again, F4 tracks literally hundreds of individual ground units, each of which has dozens of vehicles. Suppose the stategic sim tracks a few dozen, or even a hundred plus ships in a conflict. F4 and FC prove that it's manageable.



As I said, FC had the big advantage that it didn't have to be the little tactical sim at the same time it is the strategic sim. This new game of yours will REQUIRE this. FC was a tactical sim with an operational command interface instead of a station-based interface. FC, SC, and DW all have at their core the same NavalSimEngine. FC did do all the little tactical things at the same time it was tracking platform movement and accepting commands. F4 also managed to run an operational sim while running a tactical sim at the same time.

[Nexus7]

You are producing dust.

[evaamo]

Another interesting case to review containing a full featured dynamic campaign with limited theater-size/land cover is Enemy Engaged: Comanche Vs. Hokum(EECH). The source code is available, so for the programmers reading this thread, it might be entertaining to research how they did it.

It is attack-helo oriented, but has a combined arms feels to it, since a full all-out war is taking place in real time all over the theater, including pretty cool ground engagements, close air support, SEAD and strike missions from fixed-wing air assets, SAR, transport missions, recce/BDA from helos and of course, some basic amphibious ops from an ESG close to the shore (I haven't seen a full CVBG, but it might be included in some other theaters).

The separation between 2D and 3D is much less obvious than in F4, but then again, I don't think it has the same level of complexity than F4's campaign engine (I could be wrong tho') so less computation might mean a more tight integration between "tactical sim" and "operational interface". I believe strategic is a too high level for this sims to properly handle, since it becomes more human (read: political) and might be too abstract to bother modeling - not that much added value to players if you want to offer a whole strategy-to-very tactical experience to a user and yet very complex to design properly.

cheers

[Molon Labe]

Thanks! Never heard of that before, but after reading a few reviews, it looks quite impressive. I didn't find too much information on the campaign engine, but it sounds similar to the one F4 uses.....

[Hawk66]

Nice discussion^^.

I've wanted to create a campaign generator for DW last year but I've abandoned this idea - because of too many problems (AI, limited possibility to create units on the fly with a true dynamic position etc.)

I've read some threads from developers in the Black Shark forums that they haven't created a dynamic campaign engine due to economic reasons. (Costs too much for a first version).

Concerning naval sims: I heavily doubt that there will be a modern naval platform in the foreseeable future with a dynamic campaign engine.

Probably a non-commercial naval plattform might be feasible: No 3D view but more a RSR style with the option later to add graphical stations like sonar etc. The heart would be the gameplay (incl. campaign generator).

And what IMO is very important and more or less completely absent in DW: interacting with the crew(different skills), getting reports and so on - not only contact xyz was detected or lost.

I think that would generate so much atmosphere that I wouldn't need 3D grahpics. What's your opinion?

[evaamo]

If you feel like trying it, it is being republished by the guys over www.gog.com for less than US $6.

Then head to http://simhq.com/forum/ubbthreads.php/topics/2733487/STICKY_EECH_and_allMods_1_12_r.html

and download the official community mod which includes LOTS of fixes and new stuff. It is a very nice sim. I know the source code for the community mods is available as well, but don't know where it is located. Might be worth contacting the "locals" over simhq's eech forum.

cheers!

[evaamo]

Hawk, I believe you are correct in that 3D is not necessary at first. But then again, 3D is what sells these days, so like you say it may be a matter of a community effort.

Global Conflict Blue might have been such a good option, sadly the owner decided to close access to the source code.

On the other hand, such a complex campaign engine was planned for a community effort called Combat Simulator Project (http://csp.sourceforge.net (http://csp.sourceforge.net/)), however development has been slow and I'm not sure if at this moment, the scope of the project still includes such a complex item.

I believe a dynamic campaign is such a complex and non-profit oriented project, that it may take too long to develop and to stay commited, unless you work for a defense contractor in the distributed simulation business. :haha:

cheers