Jump to content


Photo

Steel Beasts Pro P E, Version 4.1

July 2019

  • Please log in to reply
105 replies to this topic

#101 Chris Werb

Chris Werb

    In Zod We Trust

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 23,891 posts
  • Gender:Male
  • Location:Orkney, Scotland, UK
  • Interests:But it's got electrolytes! They're what plants crave!

Posted 13 July 2019 - 0639 AM

I never doubted you thought that. :)

 

I live in relatively open terrain, but even here, relatively small undulations can provide many opportunities for cover and concealment*. Even with the current terrain, obviously a lot of play is about positioning yourself to maximise the usefulness of your field of fire whilst denying the enemy the ability to concentrate direct fire on you. Here you could see fro miles and miles in lots of places by cresting a hill, but you had better be quick about it because you would have everything but the kitchen sink thrown at you in short order. It also means that tanks with flat shooting guns and primitive/non-existent FCS like the T-62 become a force to be reckoned with vs far later and more sophisticated vehicles, particularly were they can enfilade or ambush. One thing I am looking forward to very much is that the terrain will make infantry much more survivable, particularly if player controlled - you will be able to readily duck behind solid cover from which you can sneak/run to an alternate position before indirect fire arrives. Yes, timed airburst and, to a lesser extent AHEAD type ammo options will make this less viable, but in rolling terrain it may not be easy for your opponent to find something to range on to get the desired effect with something like 3P 40mm, even with the linear depth burst option selected (which like the other modes is brilliantly modelled in SB). Of course the introduction of active protection systems in game will make it harder for them to inflict casualties on enemy armour. I expect this to expose weaknesses in Western doctrine such as over-reliance on ATGW.  Snake are you modelling IR jammers and automatically triggered obscurant systems too?

 

*In scenarios without persistent enemy drone presence.


  • 0

#102 Ssnake

Ssnake

    Virtual Shiva Beast

  • Members
  • PipPip
  • 6,419 posts
  • Gender:Male
  • Location:Hannover, Germany
  • Interests:Contemporary armor - tactics and technology

Posted 13 July 2019 - 0937 AM

Only partially so. There's still so much to do, everywhere.

 

 

I'd like to devote our entire time to improving the underlying modelling as well as the simulation part. But I have to earn money to afford doing so. So I need to sell some of the development time to military customers for customization projects (which in turn drives a part of the direction that we're taking, in general), and I need to devote a considerable amount of time to adapt the product to the changing hard- and software environment (new processor capabilities, new Windows versions), and to let the team do everything they can to boost our own productivity, mainly by removing, one by one, obstacles of code maintainability.

 

And that, basically, explains 95% of what we're doing.


  • 0

#103 DB

DB

    Crew

  • Members
  • PipPip
  • 11,113 posts
  • Gender:Male
  • Location:Hertfordshire, England

Posted 13 July 2019 - 1002 AM

There is a definite parallel with weather prediction models, where grid size and initial condition variations can have large effects on results. In fact the professional seaweed fondlers are known to run perturbed variants of simulations to gain a measure of the "chaos-ness" of the inputs. If the minor perturbations yield similar results then they increase the confidence of the outputs.

I do static reliability models using fault trees for a living. I don't believe the input data and I'm dubious about the order of magnitude of the results we get. I do believe them when they say the design isn't good enough, though.

For the specific issue of the terrain model, I will suggest that once the grid spacing drops below a certain fraction of the object you're attempting to conceal, then you have likely reached a suitable threshold.

Large grid spacing approximates a low pass filter, tending to smooth features. When the spacing is smaller than the vehicle, or infantry object, then to a first approximation the object will already be hidden and more resolution merely adds texture.
  • 0

#104 Chris Werb

Chris Werb

    In Zod We Trust

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 23,891 posts
  • Gender:Male
  • Location:Orkney, Scotland, UK
  • Interests:But it's got electrolytes! They're what plants crave!

Posted 13 July 2019 - 1242 PM

And it's also about about cover. Regardless of whether you can see a tank or not, it is going to be more likely to be hull down, or partially so, relative to enemy AT systems in the new terrain, so there will be a bias toward turret hits which, as Ssnake mentioned somewhere earlier, is borne out by real-world operational analysis. As a player, this is a really big deal for me.


  • 0

#105 Ssnake

Ssnake

    Virtual Shiva Beast

  • Members
  • PipPip
  • 6,419 posts
  • Gender:Male
  • Location:Hannover, Germany
  • Interests:Contemporary armor - tactics and technology

Posted 13 July 2019 - 1524 PM

Of course it still depends on the terrain itself. There are, after all, flat stretches of terrain where an increase of resolution only adds to data size but not the quality of results. But then there are also areas like the sad dunes of western Jutland which have a tremendous impact. The next big issue is then to adapt AI behavior to such a changed environment, and to take advantage of micro variations in the terrain. We haven't even started with that, yet.


  • 0

#106 Ssnake

Ssnake

    Virtual Shiva Beast

  • Members
  • PipPip
  • 6,419 posts
  • Gender:Male
  • Location:Hannover, Germany
  • Interests:Contemporary armor - tactics and technology

Posted 13 July 2019 - 2040 PM

Also not an official video, but still...:

 

 

 

Not the greatest framerate from an old machine running a pretty large battle, but still very much playable I'd say.


  • 0




2 user(s) are reading this topic

1 members, 1 guests, 0 anonymous users


    Stefan Fredriksson