That was funny, made me laugh out loud enough that I had to explain to the curious wife what I was laughing at... wife didn't get the joke

Hmm, how old are you? I'm shure your local recruiting office can help you ;-)

Ok Ok, I'll try to be more serious. =)


The Problem with tanks sims is, that a tank is only reals sensable/useful in a unit. For a FPS you can always claim to your alter ego is steely-eyed-special-forces-dealer-of-death, and therefor act alone or a VERY small unit.

For a tank the fire-unit is a platoon, the manoeuvre element a company. So if you want to play it as a tank commander in a campaign setting
(assuming that this would be single-player/offline), you'd have to have a company CO and 3(4) platoon commanders played by the computer...and able to give senseable orders to YOU the player.
If AI ever gets that good, we(humans) are DOOMED...DOOOMED I say.

The other way would be to "railroad" the campaign. Which I think would be bad, as there will hardly any unsepected development of the situation.
Thats why I believe the only way to play a tank-SIM(or should I say mechanized warfare SIM) is in multy-player networks sessions.

The other problem/wish mentioned, the equipment span.
If you want to make EVERYONE happy, and include ALL fracking tanks/IFV/APC of the world from 1914 to now, it would be a #%&*$# of modelling to do!! Me thinks that would make it quiete costly for dev. companies(and hey, what kind of sales can you expect for a simulation?) Not to speak about the problems on getting the needed informations in the 1st place!! Or you cut on the model quality and make some kind of generic FCS for all that differnt vehicles.
That would be a bit dull, wouldn't it? I mean what use are the different looks the, if you have the same Fire controll for all of 'em?

The key to company/platoon command is making simple, easy to use controls for those units. Point and click to move them on the map. Shift-point-click for waypoints. Set speed and formation. Go.

I agree with most of your points, but especially as someone who's just now getting into the tank sims (grabbed Achtung Panzer, Panzer elite, t-34 v Tiger) after a LONG lay off, (I think last tank sim I played was M1 tank platoon?), I want to be the newbie member of the platoon. I'd rather have some else give directions, while I try to survive for awhile, not step right into the tactical thinking part of being a tanker.

The AI orders don't have to be great and all-knowing, I'd love to be put into harms way at times, where you have to consider whether or not you are going to follow that order. I'd imagine there were guys at one point in time, that probably wondered wth Michael Wittman was thinking, and damn if I'm going to do that!

I do agree that the best way to play is MP, but that takes a leap after you've played a SP learning campaign. It's much less hard on your friends when you die 2 minutes into a 2 hour mission... for the 8th time in a row.

For the time span... If the engine is sound, then they can do different theaters/time periods incrementally, it doesn't have to be released that way. Granted, you are absolutely correct, it would take forever to get all the appropriate vehicles in.

Since this is a thread about "dreaming", there's not much to say about it from a developer's perspective. What you want is a complete world simulation, and from a commercial perspective I just don't see it happening. The scope simply is too big to ever finish such a venture. An engine that would support such a wide span of different systems and activities as would be needed for a combined arms simulation spanning 100 years in history (and why stop there? Wouldn't we love to charge the Waterloo battleground with a platoon of Leopard 2s? Why not have Tiger tanks in the mix to save the Roman legions at Cannae? What about Montezuma's palace guard equipped with MP5s when meeting Cortez? Axe-wielding Vikings invading Japan?).

A Panzer III has pretty much nothing in common with an M1A2 SEP except the basic concept of a tracked, mobile platform with a gun mounted in a turret. Engagement ranges have increased by a factor of ten since then (and up to a factor of 100 for defending infantry). You want a modern FPS's level of detail for the world, and at the same time fast air to drop JDAMs. Even if we accept the rather arbitrary limitation of the last 100 years in historical scope, there's still an awfully large spectrum of highly diverse combat systems to be considered if it were to be made truly universal.

The next question is then about simulation fidelity. If you want every system modeled with the same level of fidelity, that would be a rather low level actually in order to complete the work in the time span of, say, less than two decades. You'll need an underlying damage model that can cover the vast span from a single bullet, steel full caliber AP shots from 20mm to 75, AP rounds with some HE filling, canister, beehive, long rod penetrators, HEAT rounds with varying degree of fragmentation density, ... All that is supposed to work in real-time, on available hardware platforms, scalable, should include up to several hundred 3D characters per scene in glorious visual splendor with constantly 120 frames per second, cost next to nothing, and work on a tablet PC just as well as on an eight processor monster with two of the latest double-GPU cards in a crossfire/SLI configuration, right?

In short, I suspect that nobody in this thread (including me) has a remote idea about the monumental dimensions of such an attempt.

As a developer, I have to concentrate on what's feasible with the given resources and market conditions. There is no massively rich sugar daddy to finance a 500 person team for fifteen years, and I suspect there never will be. Given these real-world limitations, I think we've done about as good as was to be expected with SB Pro (and we're not done yet, mind you).

Ahh.. Thank you Ssnake for blessing us with your wisdom . Anyways, I think it should just focus on one tank, and simulate that tank very well. Other vehicles should be simulated, but not crewable to put the focus on the one vehicle that the sim is supposed to simulate.

I'd say at least 2 tanks. So you can have PvP games ;-)

Anyway, that'b be a high risk decision...what tank(s) would have the most market appeal?

I don't have a problem with a survey sim that doesn't model everything 100% accurately.

Again, Strike Fighters for tanks.

1) Why would other tanks distract? You as the player could simply choose to ignore them, it seems...?
2) I think you know Steel Beasts Pro PE... how could we possibly increase the fidelity of our work if you look at the Leopard series, the M1A1, the CV90 series, the Pizarro, the Centauro? I'm not sure what realistic expectations could be - beyond the levels that we achieved there. I mean, would you also want a fully featured driving simulation (and what for)?

Better looks - sure.
But as far as realism of procedures and realism of results are concerned I have a hard time imagining how much "more perfect" do you want it to be, given the constraints of what can actually simulated on a PC in a desktop/home environment. The next step would be hardware - control handle replicas, control panels, in short: Partial or full instrumentation. But that's by definition outside of the scope of a simulation software.

I hope that I don't come across as an arrogant SOB who's telling you that his product is perfect and can't be improved. It certainly isn't. But I don't see the vehicle/fire control simulation as the weakest link in the chain. Where I know that there's room for a higher degree of realism, I also know that it must inevitably lead to an unwieldy user interface which in turn raises concerns about the "realism of results". The harder a vehicle becomes to manage through a plethora of insignificant control elements (when in real life you'd have a crew of four, with eight arms and forty fingers to handle the situation), the longer becomes the player's OODA loop and it could very well result in less situational awareness and ability to react adequately in a given situation than it would be in real life. There's more to simulations than just switchology.

Also, it is the issue of a tank being part of a combined arms team with units of complementary capabilities. In order to achieve good results this "periphery" of supporting elements (or some other elements on the sharp end that the player would need to support) must be modeled too, and must be made adequately capable. It is not a surprise that Eagle Dynamics is now adding ground maneuver elements to their study simulations of the A-10 and gunship helicopters. The simulation's fidelity no longer depends on an improvement of airplane switchology, but about putting the entire thing into a bigger context.

Needless to say that the overall simulation complexity will grow very fast, the more combined arms elements are being added, with a serious threat to overwhelm the player with complexity. At the same time, artificial "intelligence" is nowhere near the capabilities of what would be needed for an approach of largely autonomous, computer-controlled formations of ground vehicles. Progress is being made, yes, but the desired scope of the simulated environment as articulated by customers still seems to outgrow the advances made in the "AI" field.