CMSF irregular thoughts

[LongLeftFlank]

Noted. I'd expect cut off SAA units to quickly start looking like Uncons anyway, mounting SW under tarps on pickups, making lots of IED/VBIEDs, hiding their few remaining AFVs in barns, etc. (frankly, the latter are targets/bomb magnets more than weapons in such conditions). Can't help you much with QB strictures tbh.... 

[SimpleSimon]

One very interesting use of irregulars is to pair them up with Syrian Army regulars on the same map and try to work out those dynamics. How should one use them then? As a trip wire against an unexpected attack? As hastati to wear down the enemy before he confronts your principes? Interspersed with Regular units so they can "corset stiffen" otherwise weak  units? As Auxilia to reinforce a failing defense? As a last ditch holdout with your HQ units? 

This is why I play. 

Edited October 24, 2018 by SimpleSimon

[domfluff]

This is where the Fighters start looking worse, I think - on typical settings they'll have worse equipment than Syrian regulars, but higher motivation. They'd be useful in a (costly) assault, perhaps, or as a last stand, but otherwise they're something of a third wheel - they're better equipped and trained than combatants, but worse than the army - and usually one or the other is better for a given task.

With sufficient civilian density, the Combatants have tremendous utility as spies and scouts, at least until the shooting starts. Blending into the population and tracking Blufor closely. They'll still not match Syrian regulars for firepower, but that's not always important.

Spies, IEDs and VBIEDs are disruptive technology, naturally - these can do things the regular forces on either side simply can't.

[Lethaface]

Just flavor wise can already be enough justification for including irregulars with SAA (and IIRC plenty CMSF1 scenario's do). BTW some fighters can have very interesting equipment (IEDs, AT-14, RPG-29, etc) in small units. They also have access to heavy weapons, crew-carried or on a vehicle. Realistically some fighters (lets say chechens) could be quite aptly trained and provide higher motivated troops compared to the standard army.

A mixed irregular force with combatant sentries, (VB)IEDs, fighters waiting in ambush and a mobile counter attack force of Syrian mech coupled with some technicals would be a quite interesting force to fight with. Indeed more suited for scenario's then quickbattles.

We will know more when we see the quickbattle force selection details. If the prices are somewhat asymetrical balanced one could make quite an apt force from just 'unconventional' forces. Especially on defense. Given that a meeting engagement with technicals vs Abrams isn't going to be that interesting, but the same goes for T-72's vs Abrams ;-). 
And still. I don't know the price for an irregular AT-14 team, but if you can buy 4 for the price of an Abrams... things could be quite interesting ;-).

 

Edited October 24, 2018 by Lethaface

[domfluff]

Doing some more stealth testing of different Civilian levels, and the distance that combatants are spotted at. Assumption was that this would be some kind of linear relationship - perhaps something like 1 square for Very Heavy, 2 squares for Heavy, etc.



This was the distance on "Very Heavy", which matches the previous tests.

... this was *also* the distance on every other  test I ran, including civilians set to "None" (the screenshot is actually from the "None" tests)

Now, the interesting bit was that I noticed this unit had been generated with Crack level experience. I tried a Conscript (set everything back to Typical) on the same "None" setting and they got spotted straight away.

The same Conscript infantry on Very Heavy civilian density made it to the below:



This does mean this stealth is more complex than the linear relationship I expected. Needs more testing, obviously, but it's a start.

Other things found - Deploying an MG is a "giveaway" action - this is yet another reason why RPG's are really useful to have around.

[Combatintman]

13 hours ago, domfluff said:

The importance of Infowar elements in the above makes me wonder if there's something clever you can do with Spotting objectives.

I doubt it because the spotting objective is like destroy objectives. The target unit is either seen (you get points) or not seen (you don't get points). The only way I've seen them satisfactorily implemented has been in recce scenarios. One of the missions in the British campaign set on an airfield was all about spotting the enemy versus engaging the enemy.

If you were to do a mini campaign then yes I think there is scope for spotting objectives, otherwise I don't think there's much mileage in them.

[c3k]

I'd love to see RedFor victory conditions get a huge boost for captured BluFor. 

[Erwin]

2 hours ago, domfluff said:

This does mean this stealth is more complex than the linear relationship I expected. Needs more testing, obviously, but it's a start.

Well...  definitely doesn't seem like "realistic" spotting if an enemy can get that close b4 being spotted on open terrain.  Am rather surprised.

[domfluff]

I used to think so, but I've come around - what's being simulated is the fact that these guys can be indistinguishable from civilians, which is obviously easier in a crowd of civilians, but should be possible without.

Even with no civilians, if the Crack irregulars above concealed all their weapons and dressed appropriately - are you going to open up and mow them all down? That seems unwise at best or a war crime at worst.

Again, something like this:



Is more than possible to do with the above mechanics, but would be difficult to do in another manner.

Even civilians carrying AK-47's openly wouldn't necessarily mean they were enemy combatants, which makes this kind of thing tricky.

I do want to do some more comprehensive tests around this, since I expect it's probably key to running these effectively.


It's an abstraction, sure, and one of the key conflicts with CMx2 is that the line between explicit, WYSIWYG modelling and abstracted mechanics is sometimes blurry, and the lack of information about the underlying mechanics can make parsing that more difficult.

[Combatintman]

9 hours ago, Erwin said:

Well...  definitely doesn't seem like "realistic" spotting if an enemy can get that close b4 being spotted on open terrain.  Am rather surprised.

You're ignoring the abstraction and intent inherent in the civilian density setting. If we imagine the HMMWV to be mobile VCP then it becomes more realistic. I admit it isn't perfect because it would be difficult for any insurgent to get a long barrelled weapon that close to a VCP in open terrain but that is the nature of the abstraction. Conversely it is entirely realistic for someone with a suicide vest, pistol, knife or hand grenade as thousands of cemeteries will testify.

[MOS:96B2P]

15 hours ago, domfluff said:

... this was *also* the distance on every other  test I ran, including civilians set to "None" (the screenshot is actually from the "None" tests)

This does mean this stealth is more complex than the linear relationship I expected. Needs more testing, obviously, but it's a start.

Other things found - Deploying an MG is a "giveaway" action - this is yet another reason why RPG's are really useful to have around.

I'm trying to make myself wait for CMSF2 before I start testing the civilian density setting because I don't trust that the settings will remain identical between the two titles.  However reading this very interesting thread is making it difficult to wait........................ .

I'm probably not understanding the screenshots.  To me it looks like both the crack combatant team and the conscript combatant team are both two action spots in front of the Humvee?  But from reading the post I think the crack team got closer before being spotted (which would make sense) than the conscript team?  

The "giveaway" action is an interesting find and realistic.  

I think there may be another complication to player vs AI scenario design.  The stealthiness of an unconventional unit is, in part, based on how the unit moves.  A Quick command is more likely to get a team spotted than a Move command.  I think Move is the most stealthy in relation to civilian density.  Slow (crawling) is considered suspicious.  With PBEM this works.  The player orders the team to Move towards the VCP.  However when the player is against the AI the AI plans are created by the scenario designer in the editor.  In the editor there are five movement type orders the designer can give to the AI: Max Assault, Assault, Advance, Quick and Dash.  All five of these commands involve some type of running.  So, it seems, the AI can never use the stealthiest of the movement command Move.  The AI will always be running up to the VCP I think............... hope I'm missing something.............  

@Combatintman do you have any insight on how an AI controlled team can approach a VCP without running at it? 

          

[Erwin]

4 hours ago, Combatintman said:

You're ignoring the abstraction and intent inherent in the civilian density setting. If we imagine the HMMWV to be mobile VCP then it becomes more realistic. I admit it isn't perfect because it would be difficult for any insurgent to get a long barrelled weapon that close to a VCP in open terrain but that is the nature of the abstraction. Conversely it is entirely realistic for someone with a suicide vest, pistol, knife or hand grenade as thousands of cemeteries will testify.

It's an issue of game abstraction.  Ideally the game would enable civvies with hidden guns or suicide vests to get closer.  In the pic it's obvious everyone is armed to the teeth and so looks silly.

[domfluff]

1 hour ago, MOS:96B2P said:

I'm probably not understanding the screenshots.  To me it looks like both the crack combatant team and the conscript combatant team are both two action spots in front of the Humvee?  But from reading the post I think the crack team got closer before being spotted (which would make sense) than the conscript team?          

Sorry, I'll try to re-word:

I went into it assuming that the Civilian density would be the main thing controlling spotting distance. Whilst this is true, this is not the only factor.

A Crack combatant team Move-ing on "None" density can walk up to two spots distance (i.e., 1x 8m square in between them).

A conscript combatant team on "None" cannot - they get spotted much earlier.

The same conscript team on "Very Heavy" *can* walk up to two spots away.

Need to do more tests to puzzle out the details, but factors have to include: 

- Civilian density setting
- Unit experience
- Actions they take (these were Moves, running, deploying weapons, crawling all give you away).

It's also not clear if terrain is a factor - there are no buidings in the above scenario, so that can't be a prereq for Civilian density, and no paved tiles. I wouldn't be surprised if roads were though.

Edited October 25, 2018 by domfluff

[domfluff]

1 hour ago, MOS:96B2P said:

@Combatintman do you have any insight on how an AI controlled team can approach a VCP without running at it?           

I can't find a way to do that - that might well mean that the only way for the AI to take advantage of Uncon stealth is to use "Ambush 75m"-type orders, and to try to fight from static positions, or move between them only through covered routes, out of LOS.

 

[MOS:96B2P]

30 minutes ago, domfluff said:

A Crack combatant team Move-ing on "None" density can walk up to two spots distance (i.e., 1x 8m square in between them).

A conscript combatant team on "None" cannot - they get spotted much earlier.

The same conscript team on "Very Heavy" *can* walk up to two spots away.  

Ah, very good. 

8 minutes ago, domfluff said:

the only way for the AI to take advantage of Uncon stealth is to use "Ambush 75m"-type orders, and to try to fight from static positions, or move between them only through covered routes, out of LOS.  

Hmm, this is what I'm afraid of also......................  Well, in single player, if the player was controlling the unconventional forces and the AI was controlling the conventional forces I guess it is still workable.       

[Combatintman]

9 hours ago, MOS:96B2P said:

I'm trying to make myself wait for CMSF2 before I start testing the civilian density setting because I don't trust that the settings will remain identical between the two titles.  However reading this very interesting thread is making it difficult to wait........................ .

I'm probably not understanding the screenshots.  To me it looks like both the crack combatant team and the conscript combatant team are both two action spots in front of the Humvee?  But from reading the post I think the crack team got closer before being spotted (which would make sense) than the conscript team?  

The "giveaway" action is an interesting find and realistic.  

I think there may be another complication to player vs AI scenario design.  The stealthiness of an unconventional unit is, in part, based on how the unit moves.  A Quick command is more likely to get a team spotted than a Move command.  I think Move is the most stealthy in relation to civilian density.  Slow (crawling) is considered suspicious.  With PBEM this works.  The player orders the team to Move towards the VCP.  However when the player is against the AI the AI plans are created by the scenario designer in the editor.  In the editor there are five movement type orders the designer can give to the AI: Max Assault, Assault, Advance, Quick and Dash.  All five of these commands involve some type of running.  So, it seems, the AI can never use the stealthiest of the movement command Move.  The AI will always be running up to the VCP I think............... hope I'm missing something.............  

@Combatintman do you have any insight on how an AI controlled team can approach a VCP without running at it? 

          

In a word ... no. To be honest it isn't something I've experimented with much. I think @domfluffis probably on the money with ways of skinning the cat. Other things would be to spawn reinforcements right up close to the targeted unit but of course that is nigh on impossible to pull off for non-static targets because the scenario designer has to assess with near certainty where a moving target is going to be at a particular time. Building on @domfluff's ideas, I guess you could use a trigger to move a suicide bomber from behind a building, into a building bordering a route that the target is likely to drive down. Clearly the urban environment is going to be more friendly to such AI tricks than rural environments but that is no bad thing as it is pretty reflective of reality.

[MOS:96B2P]

14 minutes ago, Combatintman said:

In a word ... no. To be honest it isn't something I've experimented with much. I think @domfluffis probably on the money with ways of skinning the cat. 

Ah, okay.  It is what it is.  Maybe very heavy civilian density in an urban environment and weather set at thick haze (like a dust storm)      Looking forward to doing some experimenting in the CMSF2 Editor.   

[sburke]

1 hour ago, MOS:96B2P said:

Ah, okay.  It is what it is.  Maybe very heavy civilian density in an urban environment and weather set at thick haze (like a dust storm)      Looking forward to doing some experimenting in the CMSF2 Editor.   

And at night.   A moonless night.  

[Sgt.Squarehead]

Very informative thread, cheers @domfluff.....Every day's a school day as we have all come to learn on our journeys with this game engine. 

[domfluff]

I'm mostly curious to find out if CMSF is actually worth playing multiplayer, if it's possible to play Syrian or especially Unconventional Forces in Quick Battles and have this be anything like a balanced or competitive fight.

Part of that is trying to work out how best to use the tools available, by doctrine or otherwise.

The lack of a points buy system in CMSF 1 has meant that we never really got a chance to see proper quick battles in action - CMSF 2 will still be worth playing, even if it isn't a competitive PBEM experience, but I'm interested in the challenge of working this out.

 

[IICptMillerII]

43 minutes ago, domfluff said:

I'm mostly curious to find out if CMSF is actually worth playing multiplayer, if it's possible to play Syrian or especially Unconventional Forces in Quick Battles and have this be anything like a balanced or competitive fight.

The short answer is yes. 

As an example, if you pit Republican Guard T-72AV's crewed by well trained/led/motivated crews, then you can actually take on US M1A1HC's and Leopard 2A4's head on and win. Syria will be a much more capable nation now thanks to engine and TO&E improvements. They will be more potent in QB's as well due to being able to choose their gear and stats.

However, most scenario's will still realistically depict the Syrian army as it was in 2008, which is to say you will be facing a lot of conscripts/green troops with poor motivation. That said though, there are a number of existing scenario's out there that depict force on force engagements against a more generic OpFor instead of the Syrian army proper, such as @George MC's excellent Forging Steel campaign. Hopefully once SF2 is released we will get more of these force on force engagements from George and other community scenario makers. Tons of new potential in SF2.

[domfluff]

Oh, the head-on conflict isn't quite what I meant - asymmetric scenarios are typically hard to get right, but I'm curious to know if an inferior force, used correctly, can compete in quick battles, with competent players on both sides.

I'm pretty sure it's possible, but I imagine you'd have to work a lot harder to win with an uncon force. That's obviously also something that has less literature backing it up, so a lot of the problems don't have "book solutions" - trying to figure out how best to leverage the disruptive strengths and lessen the obvious weaknesses.

[MikeyD]

I'm reminded of an old comment from the CMBO board, something along the lines of 'newbies tend to play Tigers versus IS-2s while veterans tend to play Italian infantry vs Finns' (or something like that). Its tempting to always include the 'Big Guns' in a fight but the 'Big Guns' tend to skew the results. If a scenario is just Blue guys with gun versus Red guys with guns victory becomes far less certain. Titles like CMBN or CMRT aren't afraid to do infantry-dominant battles. In CMSF you need that same mindset. Not every combat patrol is going to have an M1A2 Abrams acting as close infantry support. Not every Syrian defensive position is going to have a T90 and Kornet ATGM team backing them up.

[Ghost of Charlemagne]

Very interesting thread domfluff

Personally, I tend to use OPFOR ATGMs at ranges under 1000 meters (unless they have the Kornet, then its 1500 meters), as I find that they tend to miss at longer ranges than that. 

What do you guys find to be the optimal engagement range for OPFOR ATGMs? 
And do you manually target enemy vehicles or use target arcs? 

Edited November 12, 2018 by AtheistDane

[domfluff]

Setting a 1,000m arc is a reasonable plan. I rarely directly target them, since relying on the turn rolling over at the right point can be a problem.

I think the real point with ATGM's is to pre-plan the kill zone, the primary position, and a secondary position with covered approach route - ATGM's broadcast their position all over the place, so breaking down and changing position after a single shot seems like the best SOP, especially if that shot is successful.