Some things CM: Normandy Will Need To Be Able To Do

[Michael Dorosh]

A hypothetical question:

It seems to me that the overall issue is that there's a dichotomy between top-down design to get "the feel" of combat command and a bottom-up design which attempts to organically grow realism out of the modelling of low-level units and personal in a realistic way ( we can argue whether or not they actually achieve this realism at this low level but I'd rather assume that they do so we can have the hypothetical discussion about which would actually be successful ).

CM:BO and BB were attempts to create a company and Bn-level game in which you felt a little like a Bn-commander and in which you ordered units around the map just like a tabletop wargamer. Very definitely a game and very much one which attracted old tabletop and miniature gamers ( and grognards... who were quite happy to accept CMBO and BBs failings given the promise that they would be incrementally fixed ).


CM:SF appears to be more of a 1:1 details up model which in focussing on the 1:1 and minor detail issues ended up losing the overall effect which CM:BO/BB gave.


My question is this: Do you think that most old-time ( CM:BO/BB ) players would actually like a CM:SF no matter how well it was implemented? I think that, perhaps, the issue isn't so much how flawed CM:SF was out of the gate but just that CM:SF is a design which lost what drew the old-times ( CM:BO/BB ) to CM in the first place.

Comments?

But CM:SF doesn't actually focus on minor detail issues, and that's the problem. It gives the appearance of doing so by making a couple of gestures in that direction but that's it. There is the "buddy aid" for example - troops stop to give assistance to wounded soldiers. But all that does is give some visual prettiness; none of it is under the player's control, and it has little real impact on the game, and what impact it does have is mostly indecipherable to the player in actual data terms.

Most of the game is like that.

The infantry move about in different formations. But the player doesn't select any of them, nor are they customizable. So they stack up to clear rooms, for example, but the player has no control over how they do it or why. If I want to clear a room with the LMGs going in first, middle, or last, I have no way to change that - and I'm not advocating that I should, just stating the fact.

It looks like the player has more to do in 1:1, and it feels like there is more going on but in reality the premise is the same - moving squad-sized "blobs" from place to place.The difference is that in CM:BO et al, the player used his imagination to fill in the blanks. If a squad took 10 casualties in assaulting a house, he could write it off to the open ground he just ran across and didn't think much about it. He knew that the squad depiction on screen was a major abstraction. In CM:SF, I see the soldiers depicting each individual and naturally expect them to be able to tell me exactly what is going on.

 

[dalem]

Another description of the scale mismatch and lack of DFE, I think.

-dale

 

[Redwolf]

Michael already made a great reply with the main point: much or most of the finer graded modeling is not to the player's benefit. It's only an exercise in TacAI programming, effectively, a price payed in exchange for better looks.

I'll make some additional comments below but that doesn't mean I find the following points more important than the main one.

A hypothetical question:

It seems to me that the overall issue is that there's a dichotomy between top-down design to get "the feel" of combat command and a bottom-up design which attempts to organically grow realism out of the modelling of low-level units and personal in a realistic way ( we can argue whether or not they actually achieve this realism at this low level but I'd rather assume that they do so we can have the hypothetical discussion about which would actually be successful ).

CM:BO and BB were attempts to create a company and Bn-level game in which you felt a little like a Bn-commander and in which you ordered units around the map just like a tabletop wargamer. Very definitely a game and very much one which attracted old tabletop and miniature gamers ( and grognards... who were quite happy to accept CMBO and BBs failings given the promise that they would be incrementally fixed ).

I'm not sure this was an objective as such. The resources (programming and computer speed) for a finer graded model were not available.

CM:SF appears to be more of a 1:1 details up model which in focussing on the 1:1 and minor detail issues ended up losing the overall effect which CM:BO/BB gave.

Again, I don't think this is deliberate.

People know that I'm strong fan of the "point-like" infantry and tanks of CMx1, but I feel justified in saying this:

What you see is just an effect of losing control. The new requirement to place soldiers individually, explicitly, and to make them perform the actions of individually modeled single soldiers in a graphically convincing way has taken over development. Getting this to work introduces so many constraints that gameplay is now "ruled" by the outcome of the 1:1 modeling.

Some gameplay control can probably be wrestled back. We'll see how this goes.

My question is this: Do you think that most old-time ( CM:BO/BB ) players would actually like a CM:SF no matter how well it was implemented? I think that, perhaps, the issue isn't so much how flawed CM:SF was out of the gate but just that CM:SF is a design which lost what drew the old-times ( CM:BO/BB ) to CM in the first place.

Comments?

You mean CMx2, not CM:SF, right?

 

[Nemo121]

Well, I wasn't actually wanting people to refer to CM:SF ( as that has so polarised opinions that I'm not sure it can be the basis for a rational discussion )....

What I was interested in was the overarching question of whether people wanted a more CM:BO-like experience which really seems to replicate miniature gaming (IMO ) or ( whether by chance or design... and I happen to think it was largely by design given that way back when I was privy to some of the design discussions which went into it in the Alpha phase and there was a pretty strong emphasis on them not wanting to get players bogged down in details and number-crunching while wanting the engine to have that level of detail "under the hood" ) whether ( if it could be implemented fully in a way that it wasn't with CM:SF ) they'd actually like a 1:1 representation which attempted to introduce veracity with bottom-up modelling?

Leave aside that BF.C made both games and assume that this has nothing to do with BF.C but is purely a question of theory of game design.

 

[Michael Dorosh]

Well, I wasn't actually wanting people to refer to CM:SF ( as that has so polarised opinions that I'm not sure it can be the basis for a rational discussion )....

What I was interested in was the overarching question of whether people wanted a more CM:BO-like experience which really seems to replicate miniature gaming (IMO ) or ( whether by chance or design... and I happen to think it was largely by design given that way back when I was privy to some of the design discussions which went into it in the Alpha phase and there was a pretty strong emphasis on them not wanting to get players bogged down in details and number-crunching while wanting the engine to have that level of detail "under the hood" ) whether ( if it could be implemented fully in a way that it wasn't with CM:SF ) they'd actually like a 1:1 representation which attempted to introduce veracity with bottom-up modelling?

Leave aside that BF.C made both games and assume that this has nothing to do with BF.C but is purely a question of theory of game design.

Hard to leave it aside, though - CM:SF and CM:BO are essentially, at their heart, the same game. Company-level, squad-based tactical combat in a 3-D environment with a real-time/turn-based player interface of some kind. What made CM:BO successful was that the level of abstraction was just right for that type of game. In fact, there was no reason to believe that 1:1 representation was in any way necessary to improve "the game". Yes, there were calls to improve it from a vocal minority who wanted more to look at in their screenshots. I do not know of anyone who made a valid challenge from a gameplay standpoint - that honestly made a case that a company-level, squad-based game must somehow depict individual soldiers and their ballistics, morale, fatigue, etc. (man-for-man, as opposed to squad-for-squad) in order to be a better game/simulation. I would be at a loss to understand how anyone could possibly do so now.

 

[Michael Dorosh]

It's also a bit odd really now that I think about what dalem really means by 'scale mismatch' - they wanted to get a clean interface for the Real Time in CM:SF - a simplified command system. So they dropped all kinds of options and abilities from the command menu and left out others, from scaling walls to taking prisoners, etc. Which is all ok, if you're going for a simple game - but on the other end of the scale, you're weighing down the data available to the player with individual soldier statistics on morale, fatigue, even rank, none of which have any real direct bearing on the game, since one can't calculate overall firepower data from them the way one could in, say, Squad Leader or even CM:BO. And tracking individual bullets from individual guns (they claim it does this) is nice, but there is no way to display any of this during or after the game, so it's wasted effort that could have been resolved by a CRT, really.

It would be the same as if I took all the Multi-Man counters out of Advanced Squad Leader and replaced them with Single Man Counters instead. I'd have special rules for the SMCs though - they couldn't split off and move independently - they had to move as squad-stacks or not at all. They couldn't fire independently, it was all at the same target as the rest of their squad, or not at all. So why would I have them on the board? I would have the advantage of being able to break or eliminate men one at a time rather than in groups, but that would be about the only advantage; I would still have to abstract what happened when a soldier was wounded - he'd be removed from the map altogether (no medical aid in ASL).

I think any sane ASL player would reject such rules in favour of just keeping the MMC on the board.

 

[Nemo121]

Ok but imagine a CM:SF in which those details weren't abstracted and the 1:1 model veracity didn't replace the ability to give squads orders at the squad level?

It seems to me that no-one would have minded the 1:1 modelling if the complexity of orders and the squad-level of command was maintained.


After all if you have the same orders and give them at a squad level isn't it an improvement if you have the visual model at 1:1 ( not necessary but good for immersion ).

 

[Michael Dorosh]

Ok but imagine a CM:SF in which those details weren't abstracted and the 1:1 model veracity didn't replace the ability to give squads orders at the squad level?

It seems to me that no-one would have minded the 1:1 modelling if the complexity of orders and the squad-level of command was maintained.


After all if you have the same orders and give them at a squad level isn't it an improvement if you have the visual model at 1:1 ( not necessary but good for immersion ).

It worked for Close Combat insofar as CC was apparently very popular (it spawned enough sequels, anyway, and had both commercial and critical success). It was also "burdened" with a lot of detail at the individual level among the squads, but that was actually a selling point of the first games - "wow, look, we model soldier MORALE!" It worked, because, as you point out, there was a maintenance of squad-level of command (even if individuals did rout off on occasion, which was annoying though realistic).

Interestingly, the 3D translation of CC was not a success - GI: Combat, for whatever reason, did not do well, Squad Assault enjoyed more success. I think it may have been hampered more, to be honest, by the fact that there was less attention paid to armour in it. The average consumer wants tanks - an unfortunate truth. The CC interface (and consequently, the GIC/EYSA interface that it spawned) was almost too slick. It was so fast and easy to give commands, that infantry combat was almost "boring" by comparison to tank combat. Click a few commands in real time and that's it. You mention complexity of orders, and that's the crux - the classic tradeoff. The 3D versions of Close Combat suffered in that tank combat was never modelled in detail and what was modelled was done poorly I think (dance of death, anyone?) and those shortcomings were magnified in 3D. That magnification also highlighted the fact there was comparatively little for the infantry to do - at least for me.

But in any case, there needs to be some way to get your LMG into the window that you want it. Not necessarily a "micromanage button", it can be as simply as a "heavy cover arc" or something that will tell your machine gunner to set up in the window rather than in the middle of the attic whenever your squad runs up to the 2nd floor of a building. CM:SF and EY:SA and any 3D squad game you care to name will have that problem in common. It's not an inconsiderable one.

CM:BO handled it quite nicely, thank you. And you know, I had no problem in immersing myself in CM:BO. The sound effects (that stock MG42 especially), Tom's winter mods, even the eastern U.S. robin's birdsong, the Zimmerit on those lo-res Kwazydog tanks with the wheels that didn't move - it all pulled me in.

 

[Redwolf]

After all if you have the same orders and give them at a squad level isn't it an improvement if you have the visual model at 1:1 ( not necessary but good for immersion ).

But that's the thing, and that's why I'm so insistent on a more abstract game: it doesn't work that way.

If you explicitly show where each man is you cannot patch it up "after the fact", as you can in a more abstract game where the representation is fuzzy.

%%

This "patching up after the fact" is not as crazy as it sounds. Here is how it works:

• you decide based on historical accounts how many casualties a squad takes in a given sitation, let's say when coming under ambush from two LMGs (or whatever). Let's say that the chance for each man to become a casualty is set to 15%.
• in the abstract game you have a point-like squad representing 12 men (or whatever the squad size is), you evaluate the 15% probability for everybody, then you have your results.
• the squad is just a "blob", not displaying individuals.

But in 1:1, you have to decide where each individual is at what time, and then you go and decide who's a casualty based on where each man is (cover etc.). You have just lost the capability to set this from a higher level to assure a correct total outcome. Now you have to get all details "right" to get the correct total.

But, and that's the catch, if your TacAI has positioned your men in incorrect (as in, not realistic) places, then your results will be unrealistic.

So, by going to a level of individually representing "stuff" that is not under player control you have just nuked your ability to have correct statistical outcome on the level of player-controlled units. This is not a good thing.

If your game allows you to control each individual man, then it would be correct to compute casualties on that level.

I think that "blob-like" and "point-shaped" is the right representation for whatever is the smallest unit the player can control.

I can give more examples like this. Turrets on tanks are an obvious one.

 

[mOBIUS]

In fact, there was no reason to believe that 1:1 representation was in any way necessary to improve "the game".
...
that a company-level, squad-based game must somehow depict individual soldiers and their ballistics, morale, fatigue, etc. (man-for-man, as opposed to squad-for-squad) in order to be a better game/simulation.

Can they correctly model the ballistics of the .30cal. M1 ball for each shot in a RT game? Or they going to do it the dopy way TOW does it and use a dispersion fudge factor?

 

[Geordie]

Personally, what I am after is a game tyhat I can play and have fun with.

I fully understand that some of the guys here delve under the hood a lot more than I ever will and have been enlightened by a lot that you say and reveal.

However, that said, when I get 2 members of my squad shot, I dont really think much more deeply than, 'damn, Ive lost a couple of guys'. I may moan when or if they run out of cover to die, and if they consistenly do this I'll cry foul and say thats a bug.

Now I do understand that SF represents infantry in 1-1 but for certain things still abstarcts them (I forget those certain things right now). Im Ok with that though.

SO what should CM Normandy be able to do?

For me its simple. It should be able to represent battles in ww2 at sub-company level.

What I need for this is simple too.

1. An accurate representation of a typical sub-company group in both infantry and armour.

2. A fairly acurate representation of typical weapon types (Im not interested in obscure French armour used by the Nazis etc).

3. As realistic terrain and building as is possible, to both enhance the look and give the feel of ww2 NW Europe.

4. If your giving me 1-1 infantry, then I want infantry formations (an old bugbear of mine, I know). One thing that distinguishes the military from a gagle of civvies is just that, they dont gaggle around from point A to point B. The ultimate would be pairs movement, the base would be line, extended line, arrowhead, Diamond etc.

5. Vehicle formations would also be desirable.

6. Terrain that actually provides cover, SF is terrible at this as outside a house Syrians are toast!

7. The ability to play the game (RT or WEGO) just by using the Mouse, like many, I never use Hotkeys and am not interested in them, even hitting the spacebar in SF is a bloody chore.

8. Infantry close armour assault that is not abstracted, BF have already stated this wont be happening and so all we will see yet again is an unrealistic grenade kill in an AFV. Hey if you say its 1-1 Rep, then it better be 1-1 Rep, not abstracted because you cant be arsed to do the animation.

Some of the above is desirable rather than essential but you get the gist....

 

[Redwolf]

Now I do understand that SF represents infantry in 1-1 but for certain things still abstarcts them (I forget those certain things right now). Im Ok with that though.

At least some of the smallarms fire is.

SO what should CM Normandy be able to do?

For me its simple. It should be able to represent battles in ww2 at sub-company level.

What I need for this is simple too.

1. An accurate representation of a typical sub-company group in both infantry and armour.

Yeah, but the less vehicles you have the more severe do you feel bugs when vehicles are put out of action in unrealistic manners.

3. As realistic terrain and building as is possible, to both enhance the look and give the feel of ww2 NW Europe.

That's probably tough, there's just too much variety.

4. If your giving me 1-1 infantry, then I want infantry formations (an old bugbear of mine, I know). One thing that distinguishes the military from a gagle of civvies is just that, they dont gaggle around from point A to point B. The ultimate would be pairs movement, the base would be line, extended line, arrowhead, Diamond etc.

5. Vehicle formations would also be desirable.

Not to mention SOPs.

6. Terrain that actually provides cover, SF is terrible at this as outside a house Syrians are toast!

That might be a biggie. We can't predict what will happen, but you have to keep in mind here that you don't only have to have terrain with realistic cover available (which is both a programming and a scenario design problem).

Now you also have to have the TacAI place your guys in the cover at realistic moment, something that you have no control over.

The point-like CMx1 squads make use of cover easily. You move their point into a certain terrain with certain protection. Done.

We do not know yet what BFC will do about concealment from vegetation, but they can't have a 3D model for each leave, they can't leave out leaves and they can't just have vegetation be like fog like they did in CMx1.

7. The ability to play the game (RT or WEGO) just by using the Mouse, like many, I never use Hotkeys and am not interested in them, even hitting the spacebar in SF is a bloody chore.

8. Infantry close armour assault that is not abstracted, BF have already stated this wont be happening and so all we will see yet again is an unrealistic grenade kill in an AFV. Hey if you say its 1-1 Rep, then it better be 1-1 Rep, not abstracted because you cant be arsed to do the animation.

My money is on that we'll complain about the limitations of squads doing something about tanks. Too tough nut to crack at the same time as all the other problems. As you say, the animations alone are required now but undoable for time reasons.

 

[Rule_303]

You think the Uncanny Valley is deep now, wait until Normandy. It's already mildly annoying to watch scrawny Syrian reservists and Uncons moving fluidly about scanning their zones like highly trained GIs, or swaggering along in MOVE mode like big pump monkey Marines.

Now, just imagine that same body language superimposed on landser and good old Willie and Joe. It will look utterly silly, a universe away from "bent double, like old beggars under sacks" or the tense, purposeful advance of the tiny McGowan silhouettes that gave the Squad Leader infantry counters their character.

Are there more important things to get right in the CMx2 engine? Absolutely -- most glaringly (a) diminishing returns to fire (curing the JasonC "just gain LOS then pull triggers until they die or you do" issue), and (b) artificial clumping of squaddies around Action Spots.

But as people said above, if BFC is committed to 1:1, and to immersion, they also need to make a reasonable effort to get the visuals right or at least not glaringly wrong. Otherwise, they might as well have stayed with the simple CMx1 swivelheads.

 

[dalem]

Ok but imagine a CM:SF in which those details weren't abstracted and the 1:1 model veracity didn't replace the ability to give squads orders at the squad level?

It seems to me that no-one would have minded the 1:1 modelling if the complexity of orders and the squad-level of command was maintained.


After all if you have the same orders and give them at a squad level isn't it an improvement if you have the visual model at 1:1 ( not necessary but good for immersion ).

I would have minded. The ability to do 1:1 for company level didn't exist in 1999 (when the first real BTS/BFC discussion threads on the topic were active) and it doesn't exist now.

Can't be done.

-dale