Steel Beasts Complexity

[stormrider_sp]

If I could choose where to improve complexity it would be in the infantry, the thermal imagery simulation, artillery and company+ battle management.

[Hoover]

Like many of you I love sims that have depth and are immersive. A good sims draws me to the terminal like a needle is drawn to a magnet. When I have to go sometime, I’m sure, that I spent too much time simulating the reality then being there. Maybe I will regret it but otherwise I just do what I really enjoy.

• “…more than three diff…”
  • using yes, mastered to a certain degree only
• “...how long…”
  • not too long if you already know one or two
• “...difference between a Retreat…”
  • knew it, forgot it, I have to look it up again
• “… game among many...”
  • SB is a gem, it is something special as the community is. Steelbeasts.com is a habitat for endangered simmers
• "...struggling with..." (not struggling, just annotating)
  •  having many CPU cores but only one or two in use
  •  UI. Laminar Research had a legacy UI until X-Plane 11. They are a small developer team too. LR created a beautiful modern and easy to use UI for X-Plane 11 and are extremely successful with it. In this forum I often hear the argument that the military customer does not place importance on things like UI and other eye candy things. Maybe this is true for the current generation of deciders. Could be that future people in authority who grew up with X-Box-, PlayStation- and PC-games will think different
  •  graphics, great improvements are made as the nice things thread shows. I hope I do not come across ungrateful but the insatiable user hoover always wants more 😎. For example indirect illumination💡. If I’m right, SB still uses the DirectX 9 framework. If you permit I will again refer to the Laminar Research team. At one point in time LR hit the wall with their legacy Open GL framework. Inexplicable stutters…, not designed to use multicore CPUs, low framerates and difficult to debug. LR decided to go the Vulcan (Metal) way.
    https://developer.x-plane.com/2016/03/what-vulkan-means-to-developers/
    https://developer.x-plane.com/2019/04/vulkan-and-metal-it-runs/
    https://developer.x-plane.com/2016/11/physically-based-rendering-is-always-on/

[Bond_Villian]

7 hours ago, thewood said:

 

1) Clickable messages to jump quickly to the unit in the message.

2) A list of friendly units I can jump to quickly and gives me a quick status on ammo, damage, fuel, etc.

3) A more straight-forward boolean system, with optional full scripting with something like Lua.

4) An overhead view that lest you pull back further

5) An optional jump to unit firing or under fire. 

I agree that some changes to the GUI that help manage large forces/let you see whats important to you (your own units only, for example) would be most welcome. There is actually a shortcut for 'jump to next engaged unit' already. I use it a lot!

[Gibsonm]

15 hours ago, tungstenfall said:

Fire control systems are definitely not the most difficult part of Steel Beasts. I'm not even sure how one could even remotely consider them difficult. Take pretty much any dedicated DCS module and only there it can start getting complicated. And I really hope SB Pro PE will take that route as well at some point, bringing more detail even to the drivers stations and fully clickable cockpits to everyone.

 

The biggest difficulty of SBPPE comes from the need to manage a huge amount of assets, often all by one person and at the same time, all the while using the real world tactics and procedures (which of course you need to learn first). So, it's indeed basically a CMANO (or BCT Commander for that matter) with the possibility to occupy one of the units. And the strategic plus the micromanagement part might not be to everyone's liking, especially considering that you need to hand-hold the AI most of the time because of their tendency to do things their own way to the point of you feeling like cloning yourself and occupying every single station in every single vehicle. Which, I guess, is basically what happens in a class. It's not very well optimized for single player use, unless you are playing the most basic scenarios.

 

Well I guess "Huge" depends on the size of the scenario.

 

Select a Brigade attack as a single player and you should expect to be overwhelmed.

 

FWIW in our classrooms anyway its rarely used as a 1 person, one vehicle product. Often a person control and Troop / Platoon, usually they control a Combat Team (3 - 4 Platoons).

 

The strength of the AI, in IMHO, is that it averages out RL performance (despite movies, etc. everybody in every crew position does not have masses of SA). Often units do get hung up in a forest, etc.

 

However using the right combination of route/ formation, speed and spacing (where the learning investment comes in) will avoid this - if you trust the AI. I find must users have "issues" when they decide to "self drive" one vehicle and therefore break the AI routing for the others.

 

As for time spent in different views, we tend to go with:

 

Gunner / Driver - 100% 3D world, 0% Map (you don't have one in RL)

Crew commander (includes Troop / Platoon Commander) - 75% 3D world, 25% Map.

Combat Team Commander - 40% 3D world, 60% Map.

Battle Group CO - 25% 3D world, 75% Map.

Brigade Staff, etc. - 10% 3D world, 90% Map.

 

As you go up that tree, your focus shifts from "personally" operating vehicles to relying on the AI.

 

Edited August 28, 2019 by Gibsonm

[Eicio]

Well I'm pretty new in SB pro pe, I mean I had two one month licenses about a year ago but I ended up not having the time to go deep into it or at least not as would have wanted to, each time.

 

So I recently bought the usb license to be sure to have the time (this time). However, that's where I see the line between "game and hobby" I'm not new to simulation since the "realism feelings" is what I seek, even if I don't have any experience what so ever with the military.

It's just that  "classic" games, meant to be played by many (kids included), are IMHO somehow "dull", you spawn you shoot you die and so on...

 

In a simulation what you do really matters, you have that feeling of "being here" and you have to learn to do the most basics things. That's why I spent a lot of time in DCS already and why I'll do the same on SB.

 

So far, for me, learning new fire control systems are not the harshest thing. I can't say that I really master anything but, for an exemple, some days ago I jumped into a BMP-2 and I went to the wiki to know how it works, how to use the ATGMS and in no time I could do the basics with it.

 

For me the true "big step" will be to understand the "tactic" aspect of the game; how the interface in the map really works, how to lead your forces and what tactics to use, how to set up support sections, how and when to use it... Or even how to order infantry to assault a position without being butchered.

[Hedgehog]

On 8/27/2019 at 10:07 AM, Ssnake said:

Who of you can claim with confidence to having mastered more than three different fire control system families?

How long did it take you?

Do you know the difference between a Retreat route and a Retreat condition?

Is Steel Beasts a computer game among many for you, or is it a hobby of its own?

 

What are the things that you personally are struggling with?

Pre A2 M1s

Leopards

CR2

All T series

CV9040

Marder

TOWs

 

5/10mins?

 

Yes, Retreat Route = where the unit goes (or is supposed to)

Retreat Condition = When to embark on a retreat route.

(Always leave one uncondition retreat route away from a waypoint with tactics other than Hold or Stay, other wise the waypoint tactic will default to Hold)

 

It's a Hobby all of its own.

 

Computerised FCS that try and do everything for you.

Particularly the RWS things, click the wrong button and it goes haywire.

And making videos on positivly antiquated hardware.

 

[Ssnake]

12 minutes ago, Hedgehog said:

Retreat Condition = When to embark on a retreat route.

A-HA!

Gotcha. That's not what it is. It's the condition under which to stop advancing and then to return to the previous waypoint, irrespective ot the path's chosen movement tactics.

[Hedgehog]

2 minutes ago, Ssnake said:

A-HA!

Gotcha. That's not what it is. It's the condition under which to stop advancing and then to return to the previous waypoint, irrespective ot the path's chosen movement tactics.

That's "retreat back to starting waypoint if..."

 

What would you call manually set embark conditions on a retreat route? (E.G. Embark if...this unit is under indirect fire)

As to me that's a "retreat condition".

 

[thewood]

21 minutes ago, Hedgehog said:

That's "retreat back to starting waypoint if..."

 

What would you call manually set embark conditions on a retreat route? (E.G. Embark if...this unit is under indirect fire)

As to me that's a "retreat condition".

 

That's how I interpreted it also.

[thewood]

3 hours ago, Bond_Villian said:

There is actually a shortcut for 'jump to next engaged unit' already. I use it a lot

I use it a lot too, but I have to be paying attention to the messages and if something happens in between, it makes it very hard to figure out who's in trouble and where.  Some of this I chalk up to realistic fog of war, but when I am testing a scenario during development, I want every tool available to quickly move to specific units and see what's going on.  Maybe limit my suggestions to scenario testing is one method.  To me, the critical complexity reducer for someone playing at that CO-level is being able to click on a message and get some info and insight on the reporting unit.

 

To show how I do it now during scenario editing, I create a spreadsheet of all units.  I keep track of ammo and damage on it.  I also keep track of a few of the critical waypoints and triggers.  Its the only way I have found to keep things organized.  That's the complexity mitigation I work with.  It always seems crazy to use manual spreadsheets with a computer game, but it keeps me sane, since I don't have a staff.

[Gibsonm]

54 minutes ago, thewood said:

To show how I do it now during scenario editing, I create a spreadsheet of all units.  I keep track of ammo and damage on it.  I also keep track of a few of the critical waypoints and triggers.  Its the only way I have found to keep things organized.  That's the complexity mitigation I work with.  It always seems crazy to use manual spreadsheets with a computer game, but it keeps me sane, since I don't have a staff.

 

Are you aware that it generates a "report" file?

 

By way of an example, here is the report file from last week's TGIF:

 

Domfessel 2013-4159.sce_16_08-24-19_12_23_15.htm

 

If you scroll down you'll see it tracks ammunition usage, damage, etc.

 

Might save you some of that record keeping.

 

Edited August 28, 2019 by Gibsonm

[thewood]

Yeah, I only saw that recently.  Only yesterday looked at it in detail and realized that.

[thewood]

btw, thats a good example of the complexity.  You can find almost all the info you need.  Its just a matter of literally finding it. 

 

Its like the "jump to fired on unit" hotkey.  It theoretically scratches the itch as a jump key, but you have to hit it quickly and just at the right time.   Again, just a simple click on the message window would  reduce complexity as a CO immensely.

Edited August 28, 2019 by thewood

[Bond_Villian]

3 minutes ago, thewood said:

Its like the "jump to fired on unit" hotkey.  It theoretically scratches the itch as a jump key, but you have to hit it quickly and just at the right time.   Again, just a simple click on the message window would  reduce complexity as a CO immensely.

This is something that has been mentioned in the 'wishlist' a number of times

[Ssnake]

1 hour ago, Hedgehog said:

What would you call manually set embark conditions on a retreat route?

An embark condition for a retreat route. That's how I'd call it.

[Graycap]

On 8/27/2019 at 11:07 AM, Ssnake said:

I'm wondering how much more complex the Personal Edition of SB Pro should become

 

Very interesting topic. 

Never reply to a question with a question. That's true, anyway... More complex for the programmer or for the player? The more complex the programming the more simple is for the gamer to have situational awareness (I mean the sensation of being there and to understand what to do and why).

 

I've been a player (solo) for quite a time. As a simulation gamer I was coming from Falcon4 and its incarnations. My biggest problem at the start was the missing "picture" (a world "living" around you) and the missing commander. If I had been playing as a platoon leader in complex scenario (for the need to have a world around me) I never received "orders" from above. And the radio net was completely saturated with repetitive messages spoken by the same voice coming from all units in the map. I doubt that a platoon leader should know anything above company level. As a solo player did I need to worry about all that units under fire? I never heard on the net orders coming from AI to the units under fire (hopefully with a different voice) giving me the sensation that there is a "commander".

 

What about transforming all the conditions in "orders" issued by an AI commander and maybe adding an explanation to the issued order ? That's what I mean with complexity for the programmers and playability (immersion) for the user 

 

 

 

[thewood]

2 minutes ago, Bond_Villian said:

This is something that has been mentioned in the 'wishlist' a number of times

Its some simple UI changes like that that would reduce complexity for someone entering the game and wanting to just watch what's going on.  I learned SB from finding SP scenarios that had plans for both sides and then watching them play out.  I then went and chose one side to make plans for and watched how my plans played out.  I kept adding to what I did until I got to the point where I understand the planning process very well.

 

But a couple things make it much more complex than it needs to be to learn by doing the above:

 

1) The difficulty in moving to where the action is quickly to see what happened

2) Lack of granularity in the AAR.  It shows hits and, if the AAR gods shining on you, some non-hit events.  But it shows nothing for misses, etc.

3) I would have said tracking unit status, but now that see the detail in the report, my only wish is integrating that into when I click on a unit to get the same info quickly.

 

None of these really reduce a lot of SB complexity during execution, but they hugely help in understanding and learning how to manage the complexity.  They also help greatly in scenario development.  I guess my conclusion is that its not the complexity per se that inhibits playing SB, its the tools you would expect to have available for a complex simulation aren't there.  Of course, that is dependent on how you play.  If you just play in-tank, not that big of a deal.  But if you play at a higher level, the complexity of learning grows.

[Hedgehog]

21 minutes ago, Ssnake said:

An embark condition for a retreat route. That's how I'd call it.

Well Names aside, yes I know what a Retreat Condition is.

Been playing this since before the 2.270beta

2.235?

 

But since I never played "Online" with 1.226

I'm still a "Newbie" 🙄

[Ssnake]

17 minutes ago, Graycap said:

...the radio net was completely saturated with repetitive messages spoken by the same voice coming from all units in the map. ... As a solo player did I need to worry about all that units under fire?

Maybe I should suppress my urge to provide tech support, to keep the thread onm track. But, just so you know, units reporting that they are under fire/ask for a permission to retreat are signaling that their implicit embark condition to vacate an assigned battle position is met, but that they aren't given a route to take (IOW, a planning error). I can but recommend reading up Chapter 8 again, and to memorize the tabular overviews about behaviors and implicit embark conditions for the various tactics, as these are the rules by which computer-controlled units will make their decisions.

There is no commander because, in single player mode, you are it.

In network sessions you will notice that the radio chatter probably makes a lot more sense (but then you're dealing with other humans).

[Graycap]

6 minutes ago, Ssnake said:

...

There is no commander because, in single player mode, you are it.

In network sessions you will notice that the radio chatter probably makes a lot more sense (but then you're dealing with other humans).

I understand that. With experience I have understood that the more important phase is before pushing the start button.

 

Maybe I haven't been able to explain myself. What I'm trying to explain is this kind of situation: 

- download a scenario (company level) 

- you choose to play as a TC forgetting about anything else (maybe at the beginning of your learning curve)

The problem, maybe solvable at the cost of more complexity, is: you can't forget all the other units. 

 

I understand perfectly that if I would learn the TC job I should play a one versus one. But this could be so different from "the real thing", meaning a scenario a little more complex, that you would be immediately overwhelmed by the informations when entering this new level. The TC job is also about executing Platoon leader orders.

 

This is the kind of "complexity" I was talking about. A more "commander-like" AI. Not only "switchology". You could render SB PRo PE more complex introducing mechanical complexity (e.g. start the engine button by button or choosing radio frequency and so on ) crew fatigue and a ton of different complex things. My favourite complex thing is having a commander if I should choose to play at an inferior commanding level (without the capacity of switching levels in game upward...). But this was the reason for my question: maybe this topic is only about phisics and systems complexity to be simulated. In this case I apologize for the OT.

 

[tungstenfall]

5 hours ago, Gibsonm said:

 

Well I guess "Huge" depends on the size of the scenario.

 

Select a Brigade attack as a single player and you should expect to be overwhelmed.

 

FWIW in our classrooms anyway its rarely used as a 1 person, one vehicle product. Often a person control and Troop / Platoon, usually they control a Combat Team (3 - 4 Platoons).

 

The strength of the AI, in IMHO, is that it averages out RL performance (despite movies, etc. everybody in every crew position does not have masses of SA). Often units do get hung up in a forest, etc.

 

However using the right combination of route/ formation, speed and spacing (where the learning investment comes in) will avoid this - if you trust the AI. I find must users have "issues" when they decide to "self drive" one vehicle and therefore break the AI routing for the others.

 

As for time spent in different views, we tend to go with:

 

Gunner / Driver - 100% 3D world, 0% Map (you don't have one in RL)

Crew commander (includes Troop / Platoon Commander) - 75% 3D world, 25% Map.

Combat Team Commander - 40% 3D world, 60% Map.

Battle Group CO - 25% 3D world, 75% Map.

Brigade Staff, etc. - 10% 3D world, 90% Map.

 

As you go up that tree, your focus shifts from "personally" operating vehicles to relying on the AI.

 

Of course, that "huge" part mostly refers to the biggest scenarios, like a brigade attack case you mentioned. But we want to be a able to get the best out of all of them, regardless of the size. Or at least this is my wish.

 

Yes, as I surmised SBPPE is more of a classroom tool, which is no wonder considering what the main product is. And to fully utilize its worth it's best to actually have a classroom or at least several computers bound by LAN, preferably with TCs and gunners occupying the same rooms. Maybe it's a good idea to indicate so explicitly on the main site. Not that it would prevent me from buying a copy anyway, but at least it could help with people not picking the product they might not find very useful for their single player purposes.

 

While I agree that in real world crews have different amounts of SA, levels of fatigue and so on, at least humans don't tend to ram trees where there clearly is a better path right next to them. Truth be told, I found the current path finding much worse than it was in 4.023. Probably I'm not using some special Shift or Ctrl+click combos (to my shame, this is the first time I'm hearing about those), but at least I don't remember having this problem manifest itself to this degree prior to 4.1.

That's not to mention cases like enemy troops occupying buildings staring blankly at my attempts to get an APC next to the said building to shoot them up manually, since the APC's crew also couldn't care less about what's happening. Though, at that particular time maybe I should have had pressed 'F' and that would be it.

 

Well, that table is again, something to be expected. If only I could have the luxury of having a large group of people with me every time I'm trying out a more or less complex scenario (not to mention the luxury of all of them having enough licenses to actually take part in the session)... Alternatively, I could try joining one of the online games published here, but I don't feel like doing that just yet, not to mention I prefer playing with people I know personally (then there are timing issues and so on).

[Ssnake]

2 hours ago, Graycap said:

But this was the reason for my question: maybe this topic is only about phisics and systems complexity to be simulated. In this case I apologize for the OT.

No, there are no such limitations. In fact, your remarks are precisely what I'm after, irrespective of the technical complexity (well, impossibility maybe) associated with it. I agree that the ultimate challenge in SB Pro is to orchestrate the whole forces assigned to you, each unit with its specific set of capabilities, working in mutual support with, ideally, perfect synchronicity.

 

But you realize, of course, that even humans who study it for years can still get it horribly wrong. How could a computer battlegroup commander fare any better?

If there is a way to do that, we haven't found it yet. Even then we need to get the pathfinding right first, as it would be the foundation of all other AI behavior of higher levels.

[thewood]

I have watched a number of players in videos over the years and think there is one thing that is a big strength of SB.  That's triggers and manual activation of them.  You can use them to create divergent or alternative orders that let you as the commander of the entire task force quickly make decisions for AI subordinates.  But its a complication in the planning phase that leads to simplification in the execution phase.

 

Part of the complication might be that if you don't learn how to plan in detail, it will become overwhelming in execution.  

[Graycap]

50 minutes ago, Ssnake said:

No, there are no such limitations. In fact, your remarks are precisely what I'm after....

Thank you for your kind reply.

 

I understand very well the extreme complexity of a "AI battle group commander". 

 

But maybe there are little steps that could be taken in that direction. 

 

Manageable radio networks, different voices with a little of characterization (Red 6 at etc...) orders given at low level (e.g. change formation - if mission designer has so designed; retreat or advance to xy - if a condition is met; which kind of menace the platoon commander or another tank of my platoon has discovered, fall back in formation - if in wrong position.....).

 

The problem to solve is not, in my opinion, to simulate a Brigade commander + his staff, but to facilitate the initial steps and give "sense" even at low level in solo play. If I would be a member of a platoon trying to verify my tactical learning in a little more complex scenario I think it could be more immersive to feel the presence of the team. Leave the concept work to mission designer (paths, conditions, ecc..) but translate them into vocal orders, and maybe create a little of SOPs coded or the possibility to create these by SB PRO PE excellent and technically proficient community.

 

Now I'm sure that I've bothered the forum too much.

 

Thank you for your attention.

[Grenny]

17 minutes ago, Graycap said:

T

 

Manageable radio networks, different voices with a little of characterization (Red 6 at etc...)  => Needs voice Actors and lots of them for a dozend languages....people are expensive

orders given at low level (e.g. change formation - if mission designer has so designed; retreat or advance to xy - if a condition is met; which kind of menace the platoon commander or another tank of my platoon has discovered, fall back in formation - if in wrong position.....).  => many of the things you can already do with the scenario editor or in the game planning phase

 

T