Could this be the first pic of CMx2 WWII ? [Edit] Yes!

[Dietrich]

because of the God problem (i.e. player is God, he knows everything) there is still Area Fire...

*** POTENTIAL SPOILER(S)! ***

In playing "Debouch to Disaster" (from USMC Semper Fi, Syria! campaign), I had an HQ team spotting from the crest of a hill and an LAV positioned hull-down about 20m to the team's right. The HQ team had spotted a BMP, whereas the LAV, though it had LOS to the area where the BMP was, had not spotted the BMP itself. I figured the CO would send one of his immediate subordinates (or even his XO) over to the LAV to climb onto the turret and tell the LAV's commander, "Hey, there's a BMP right by the east wall of the northern building of that farm down there -- the captain figures if you fire a couple dozen shells into the little courtyard there, you're bound to hit it", so I had the LAV area-fire on that spot for several seconds and sure enough, the HQ team's spotting confirmed that the LAV knocked it out. *shrug*

*** END SPOILER(S) ***

When almost every unit has a radio, I don't think it's gamey for a well-positioned but lightly armed unit (like a sniper team) to call something with not-so-good LOS but heavier weaponry, along the lines of*:

Sniper: "Kilo One-One, this is Super Four-Two; come in, over."

LAV: "Super Four-Two, this is Kilo One-One; I read you."

Sniper: "Spotted an RPG team on the upper floor of the building at your eleven o'clock."

LAV: "[long pause] The two-story building with the balcony and three windows on the upper floor?"

Sniper: "Affirmative. [pause] Recommend you recon by fire; that'll probably send 'em packing."

LAV: "Acknowledged, Super Four-Two...."

Sometimes, though, the C2 indicator shows that a given unit (this is especially true of dismounts) does not have radio contact with friendly units; in such a case, telling a friendly unit to fire on an area where the out-of-contact unit had spotted an enemy unit would not be realistic.

* I have little idea how BLUFOR units communicate with each other via radio; I can only assume that movies like Black Hawk Down are at least basically accurate in that regard.

[McIvan]

Units in contact with each other do filter the info around already, that's kind of the point and is meant to abstractly simulate the actions you describe.

The God element is the player area firing before their unit gets the nod from the spotter or otherwise spots the target themselves, just as you did with the LAV vs the BMP.

Difficult to legislate against the God element and still have a game that's fun to play.

[Battlefront.com]

As McIvan said, there isn't much we can do about the God problem (or "Player as God", as we commonly call it here). Relative Spotting, especially in Iron Man Mode, can interfere with a smooth understanding of the battlefield, thus reducing the chances the player will notice something with his God like powers, but it doesn't do all that much (especially in WeGo).

The real "solution" is to have a player command only a single unit. But that would suck The next best "solution" is what we call CoPlay, which is many players per side with specific units under their command. This means a player in charge of a Platoon literally won't know any information his units don't see unless it's passed onto him. And since we can control how the information is passed (unless players work around this with phones/email/online-chat) the experience will be quite different than the game now. But CoPlay is still a few more games down the road for us.

Steve

[McIvan]

We are definitely going to put back in "detailed hit text", as well as some sort of indication of "hit chance %". Ironically, it's harder to do the latter in CMx2 than it was in CMx1. Because CMx1 was an abstract system it was far easier to give some sort of relevant chance of a hit because that was, in fact, what the system was concerned about... chances. In CMx2 it's not nearly as simplistic as this because it has to do with how the actual shot path intersects the vehicle in question. Basically, to have "hit chance %" Charles has to code up something similar to CMx1. Odd, eh?

Steve

Great stuff.....very pleased to hear it

So what you're pointing out is that hit probability depends on the motions of the target, which of course can and will change by the time the ordnance is fired and arrives, even if something as minor as turning the turret.

But presumably the chance if your vehicle fired at that precise instant at the target's precise then position, inclination and speed could be calculated and displayed as a rough guide. Although I suppose even then flight time of the round would render the percentage moot. Would still be useful though.

[Battlefront.com]

Oh, I forgot to say that "hit chance %" and "kill chance %". They are both more difficult to do in CMx2, but the kill chance is even worse because it's not so easy to come up with rough averages like we did in CMx1. Not so bad for WW2 than modern, but it's still tough.

A modern example is something like an ATGM round having a 100% chance of killing a tank with a flank shot if it hits straight armor, but having 0% chance if it hits ERA. Which one do you show the player? Averaging doesn't seem like a good idea since now the player thinks he has a 50/50 chance of a kill, which really isn't the case at all. Especially if the ERA is only covering 15% of the total surface area. Even a weighted average isn't really that useful. So it's not very straight forward.

McIvan,

So what you're pointing out is that hit probability depends on the motions of the target, which of course can and will change by the time the ordnance is fired and arrives, even if something as minor as turning the turret.

Correct. Not only were things more simplistic in CMx1, but they were always being shown with the game in a static state because there was no RealTime mode. So when you lined up a shot in CMx1 you thought "sure, it might change due to conditions" but in reality people relied on that number to gauge if it was worth taking a shot. And every time they come back to that unit (on that turn, of course) the number doesn't change. In RealTime it will likely be different each and every time you look at that unit.

But presumably the chance if your vehicle fired at that precise instant at the target's precise then position, inclination and speed could be calculated and displayed as a rough guide. Although I suppose even then flight time of the round would render the percentage moot. Would still be useful though.

Correct, it wouldn't be that useful. It could also be a distraction/burden when playing in RealTime since having the Target cursor up means the system has to run these calculations in realtime to be accurate. I don't know what that would do to the framerate. On top of that, the player could become distracted by sitting there focused on one unit waiting for the best value before clicking on the target. But that's really bad since that value only applies if your gunner fires that milisecond, which isn't likely to happen. So the values are really too variable to display to the player the way CMx1 did.

Our solution is to display a fuzzy approximation. You click on a unit, select Target, hold it over an enemy unit, then get something like "Chance of Hit - Good" and "Chance of Kill - Poor". Far more useful because the ranges for each fuzzy value will be large enough that small changes won't likely change the display. It is also directly communicating to the player that these are "guesses" and not something intensively calculated. Which is actually more realistic.

Steve

[akd]

Well, I personally don't think that "chance to hit/chance to kill" values are appropiate information to feed to player, other than perhaps at a lower difficulty level. They strike me as gaminess not necessary in the current CM engine.

Anyways, knowing the limitations of your equipment and the characteristics of your opponent's is part of the challenge. Tell your AT guns to open fire beyond effective range and suffer the consequences. Tell your Shermans to go head to head with a Tiger and see what the outcome is without aid of a tooltip to warn you of your own poor judgement beforehand. Only unfair if you don't make the pertinent information available to the player in the game or manual.

[McIvan]

If your crew would have a rough idea on their chances of hitting, then I don't see that the player shouldn't know.

Re the chances of killing, you make an interesting point. I'm not terribly attached to kill chance in CMBB/CMAK and usually ignore it anyway. As you say the crew might not know that much about their chances.....how many times would the average tank crew have bounced rounds off any German tank, let alone all of the various types.

[James Crowley]

As McIvan said, there isn't much we can do about the God problem (or "Player as God", as we commonly call it here). Relative Spotting, especially in Iron Man Mode, can interfere with a smooth understanding of the battlefield, thus reducing the chances the player will notice something with his God like powers, but it doesn't do all that much (especially in WeGo).

The real "solution" is to have a player command only a single unit. But that would suck The next best "solution" is what we call CoPlay, which is many players per side with specific units under their command.

Steve

Co-play, IMHO, is an alternative method of play rather than a solution. I have enough difficulty organising and integrating a game with a single player, taking work and family into consideration; four, five, six or more players would be a once-a-year event, at best.

But why should commanding one unit be the only other solution? Why not a top-down command level solution, whereby you only command (and receive intel from) units that are in-command?

I have, but don't actively play, CMSF (not interested in modern/hypothetical) but from what I have seen and read of playing real-time, there are many similarites between that and a top-down solution anyway.

Take a company level action in current 'God' mode. In real time, as player, you set up your plan, issue your orders and then concentrate on the priority hot-spots. You can't be everywhere at once so you trust your units to follow your orders, at least until contact with the enemy, which is usually where the plan falls apart. Then you jump from place to place, trying to coordinate your forces and meet your goals. Inevitably, some of your units will receive scant attention and have to look after themselves, following the orders you have previously given and with, presumably, the AI kicking in and at least looking after their immediate preservation.

In a top-down command scenario you give orders to all the forces in command (presumably all, at the start) etc, etc. At some point, you lose contact with, say, 1 Platoon, so its constituent parts look after themselves, via the AI again.

The big difference is that, under 'God' mode, you can still co-ordinate all of your units, no matter what their situation, towards fulfilment of the overall plan. You know where they all are (or can fairly easily find them), their status and their appreciation of the enemy.

In top-down, you have no idea of any of those things and cannot get that information until you re-establish contact again, if ever. You will have to adjust you plans accordingly, possibly commit your reserves and always be worried that the platoon, or squad or fireteam, has been wiped-out. It would, I think, make initial planning far more important and would, or should, force players to concentrate on keeping their forces in contact and, therefore, in command.

That, again only IMHO, does not suck but is more realistic and at least as challenging and fun as the current 'God' mode.

[Lethaface]

James Crowley, sounds like a nice idea. Perhaps it is possible to have an 'uber-iron' mode that would make it impossible to give orders to squads outside c2. That way only those who want to experience this, will have to.

One point though; If you would be playing the RED side, this is almost impossible. Especially for unconventional forces and to a lesser extent reserve/infantry. RED is supposed to fight without command, it would be a bit dull to look at an empty map, wait for 60 minutes and then see 'total defeat'.

Perhaps a compromise would be possible; units are visible but orders cannot be given? This still would be a problem for red Unconventional units though, but they could be simulated having mobile phone c2 with standard delay.

[Drusus]

I think a small pause for area fire would be good. Say +5 seconds if there is a question mark at the location, 15-30 seconds else. +5 seconds if the unit is in connection to it's HQ, 15-30 seconds else. That would give a minimum of 10 seconds to wait before fire starts and a maximum of minute. Maybe add in some experience modifiers etc.

At the moment in real time play the Area Fire of God is killing realism. You can even kill unspotted enemy vehicles with area fire!

[Other Means]

I think a small pause for area fire would be good. Say +5 seconds if there is a question mark at the location, 15-30 seconds else. +5 seconds if the unit is in connection to it's HQ, 15-30 seconds else. That would give a minimum of 10 seconds to wait before fire starts and a maximum of minute. Maybe add in some experience modifiers etc.

At the moment in real time play the Area Fire of God is killing realism. You can even kill unspotted enemy vehicles with area fire!

I agree with this 100%. I'll be presenting by theses later

[Drusus]

Just realized, this would allow easy indirect fire for on map mortars. Just don't check for LoS when area firing and it is done!

[Michael Emrys]

IMO, James Crowley has hit the nail right on the noggin. The catch might be whether BFC can program such a system and I don't have any answer for that.

Michael

[Cpl Steiner]

+1 for James Crowley's idea.

Imagine you're a squad leader and you get a garbled message over the TAC Net like "enemy in trees south of village; fire on the tree-line". First off, you will be thinking, "does he mean that tree-line, or this one?" Next, you will be scanning anything that vaguely answers the description "tree-line south of village" for enemies in the hopes you can positively identify one and thus know for certain where the fire is to be directed. No doubt after 5 or 10 seconds of this you will be getting another garbled message like "for God sake why aren't you firing?", at which point you will probably just pick a target at random and open up on it. Lots of delays!

[McIvan]

I don't think it would work, due to the limitations of the AI, with all due respect to the programmers. A realistic AI is, as far as I know, just too hard at the present state of technology. So your troops that went out of effective command would just be a source of frustration rather than fun.

Each to his own though, you guys might love it. I don't think I would enjoy it much however. I like to be in command.

I quite like Drusus' idea as a sort of half way fudge...although I'd put no delay if you are area firing in the vicinity of a question mark or a fortification as that's an obvious use for area fire. Slap a delay on for just area firing into the middle of nowhere/some random house....that could be reasonably effective.

[panzermartin]

Having units switching hands from human to AI can be frustrating with the lack of intelligence and initiative from the CPU side. Imo the future lies in COplay, not in wasting work hours making an AI that can never be good enough. It's not that hard as people make it sound here. Il-2 has 1000 players online at a single time, with hundreds of virtual squadrons, that can achieve amazing levels of coordination via comms. If BFC invests a bit more in online play, with lobbies, ladders, perhaps a meta campaign later, then we can have a lively online community and mulit games wont need a pre month arrangment.

[James Crowley]

Having units switching hands from human to AI can be frustrating with the lack of intelligence and initiative from the CPU side. Imo the future lies in COplay, not in wasting work hours making an AI that can never be good enough. It's not that hard as people make it sound here. Il-2 has 1000 players online at a single time, with hundreds of virtual squadrons, that can achieve amazing levels of coordination via comms. If BFC invests a bit more in online play, with lobbies, ladders, perhaps a meta campaign later, then we can have a lively online community and mulit games wont need a pre month arrangment.

If BF go down the MMO route I, for one, will be heading rapidly in the other direction. I'm fairly picky as to who I share my limited leisure time with; the idea of playing on-line with 1000 potential muppets makes my blood freeze!

As for the AI, I respectfully disagree. Matrix games' Conquest of the Aegean and Panzer Command:Kharhov both have pretty powerful AIs, both on the attack and defence. CMx1's AI was pretty decent anyway and could (and still does for me) provide a viable if not brilliant opponent.

And if no one bothers with AI then it will never improve.

In any event, for what I am suggesting, you do not need super AI. For the most part you only need it to do what it is already doing when it controls the enemy force i.e. follow you waypoint route until it contacts the opposition, fight and, if successful, carry on to the final destination.

What some folk object to is not being able to control all of their units all of the time; being able to micro manage everything so as to co-ordinate their perfect plan without, probably, having had much of a plan in the first place. Makes for a good video game but it is a tad light on the realities of the battlefield.

As someone else suggested, this level of C&C could be optional, like extreme FoW in CMx1 and Elite in CMx2. That way everyone gets to suckle the teet.

[Drusus]

It would be cool if you could slug it out AI vs AI. You would make the first AI plan and your opponent would make the other AI plan. No micromanagement in this one Could be a nice meta game.

[JonS]

In a top-down command scenario you give orders to all the forces in command (presumably all, at the start) etc, etc. At some point, you lose contact with, say, 1 Platoon, so its constituent parts look after themselves, via the AI again.

[snip

In top-down, you have no idea of any of those things and cannot get that information until you re-establish contact again, if ever. You will have to adjust you plans accordingly, possibly commit your reserves and always be worried that the platoon, or squad or fireteam, has been wiped-out. It would, I think, make initial planning far more important and would, or should, force players to concentrate on keeping their forces in contact and, therefore, in command.

That, again only IMHO, does not suck but is more realistic and at least as challenging and fun as the current 'God' mode.

Except that mission-command is moving in the opposite direction. Commanders no longer expect to be 'in command' at all times (if they ever did), and will continue with the overall plan based on the orders they have in front of them and opportunities present on the battlefield.

You, on the other hand, want everyone to shut up shop and turn turtle (under command of the AI) when they lose command and run out of set orders, if I understand you correctly.

That isn't, IMO, a good model of the way command relationships are supposed to work.

The issue is, I think, too much co-ordination between forces that shouldn't be very co-ordinated (because they're out of command, etc). IMO, randomly assigned command delays are a better model for that than just having units go doggo. The delays need to be random (within ranges based on particular parameters - training, experience, quality, time out of command, quality of command link, etc) since otehrwise it's too easy to factor them in to the player's godlike planning and effectively 'plan out' any delay.

So;

* good quality, well trained troops with robust command links will have essentially no delay.

* good quality, well trained troops with no command links will have some delay, but it will be manageable (ie, they can be assumed to be continuing with prior orders)

* poor qaulity troops out of command ... forget it. They will essentially stay put and only respond to direct stimulus from the enemy, unless and until you get them back into command.

IMO

[Kwazydog]

As for the AI, I respectfully disagree. Matrix games' Conquest of the Aegean and Panzer Command:Kharhov both have pretty powerful AIs, both on the attack and defence. CMx1's AI was pretty decent anyway and could (and still does for me) provide a viable if not brilliant opponent.

Whilst I personally would like to play such a game, I think you are greatly overestimating the capabilities AI in any game James.

What you are basically asking for units to be self aware of their past, current and potential future situations and to take control of their actions based on that information. Should they continue their current orders or should they aid another unit in trouble? Should they defend their location or should they move to defend against a suspect flank attack? Is there likely to be mines in the area or an ambush waiting if they do? Should they withdraw or hold at any costs or move from their current location as it is a likely target of artillery fire?

Current AI technology is all well and good when its the enemy is being controlled by it as the player usually doesn't gets to see the mistakes and when it does it gets to benefit from them. The problem is that those mistakes will make or break a game as you lose important assets to them. Whilst realistic in theory, it is only so if those decisions were made with a human level of intelligence.

Dan

[akd]

If BF go down the MMO route I, for one, will be heading rapidly in the other direction. I'm fairly picky as to who I share my limited leisure time with; the idea of playing on-line with 1000 potential muppets makes my blood freeze!

He is speaking here of a community with 1000s of player online at peak times. These players might be dogfighting with 1-50 others on a persistent server or perhaps playing a co-op mission with just a handful of people.

[SlowMotion]

... and most people who play those multi player online games probably think they are much more interesting than playing against the AI. There's less coordination, but much more FOW and all sorts of unpredictable things happening. AI units have their predetermined differences according to experience etc. Real players have much more variety in how they use their units.

I don't know how well real time MP games using CMx2 engine could work, but even in turn based modes those games could be a big improvement to current 1vs1 games.

[Other Means]

(This may look familiar to some - it was posted 18/01/08)

Suggestion:

Area fire should have a delay and weight dependent on tactical situation + C&C (up to a maximum value).

Issue to be addressed:

The players ability to support with fire anywhere, given their “God like” ability to be everywhere on the battlefield.

Associated issues:

* Lack of player feedback causing an involved area fire process stemming from the unit spotting system.

* Command delays in movement orders.

I’d like to discuss these separately to not diffuse this discussion.

Improvement in game-play change would bring

* The C&C framework would assume proper, realistic importance.

* The level of planning the player would have to make to play the game well would increase.

* The tactical decisions would mirror those made by a commander in the field in sighting overwatch, ensuring C&C etc.

* Ambushes, hit and run tactics, fighting withdrawals, etc would all be more useful as less instant firepower could be brought against them.

* Unrealistic “lone sniper” recce where the sniper is out of C&C would become less useful.

I feel all these add to the tactical depth of CM - and that’s what I’m playing the game for.

Description

I believe we play CM at 3 distinct levels:

* Squad leader - we tell each squad where to go

* Company commander – we form the plan that the units conform to

* God – we can see everywhere and have access to all info our units have and can give instantaneous orders to them

The player sees all and knows all, so can act as God. So I think it might be an idea to start limiting his powers.

One of the innovations that CM1 brought us was command delays. I really liked these as I think they illustrated the command net and brought a level of planning to CM. I’d like to see them re-introduced but I won’t argue that here.

What I would like through is similar delays when ordering area fire.

In the situation above, the unit with the thick red line has LOS to the enemy unit, but the MG on the right has no knowledge of it. How would it know to fire on it? In the game he can fire straight away, so what does this do to our tactics?

* We can advance a squad anywhere out of C&C and not suffer penalties to the way they integrate with the rest of the company.

* We can leave support weapons anywhere and rely on them being able to instantaneously support an advance.

Our tactics are changed because we suffer no penalties for unrealistic play.

Proposed solution

Model the command net in the calling of area fire, the main way the player acts as the central nexus of information. Give area fire a delay and weighting attached to it dependant on the firing units perception of the enemy positions, either via the C&C network or self spotting (all figures etc TBD):

* If the unit can see the enemy - no delay, maximum fire density

* If the unit has a sound contact - minimal delay, 90% fire density

* If the unit has a tentative contact - medium delay, 70% fire density

* If the unit has no sight of the enemy, the shortest communications delay between any unit that can, via CoC, OR some upper limit (more for WWII), to simulate ad-hoc contact, 40% fire density

* If there is no enemy unit visible to any unit, maximum delay and light “recon by fire” 20% fire density

I feel this addresses the fundamental issue of Player as God - it removes the ability to USE the information gathered by the player due to his playing a game.

[James Crowley]

Whilst I personally would like to play such a game, I think you are greatly overestimating the capabilities AI in any game James.

What you are basically asking for units to be self aware of their past, current and potential future situations and to take control of their actions based on that information. Should they continue their current orders or should they aid another unit in trouble? Should they defend their location or should they move to defend against a suspect flank attack? Is there likely to be mines in the area or an ambush waiting if they do? Should they withdraw or hold at any costs or move from their current location as it is a likely target of artillery fire?

Current AI technology is all well and good when its the enemy is being controlled by it as the player usually doesn't gets to see the mistakes and when it does it gets to benefit from them. The problem is that those mistakes will make or break a game as you lose important assets to them. Whilst realistic in theory, it is only so if those decisions were made with a human level of intelligence.

Dan

Dan, I don't disagree with you in regard to an AI carrying out all or any of those tasks at the same level as a human; that is generally a given. However, I'm not sure that it would have to, only at the same level as it plays the enemy. Would this break the game, as you suggest?

If you have lost C&C with enough of your important assets then I would tend to think that your cause is lost anyway. Being able to then still micro-manage all of those disconnected units (unrealistically) to achieve a final 'victory' is pretty gamey.

Jon, you have interpreted my notion as operating at the other end of the spectrum i.e. no more set orders so they just pack up and go home. I don't even see this happening when I play against the AI in CMx1. Sure the AI may make lots of basic errors and obviously the 'decisions' it makes are sometimes odd and predictable but it will at least carry on trying to do what it was 'ordered' to do. The fact that it will often get wiped-out is more of a function of the high casualty cut-off level that has been allowed, than an inherent 'failure' of AI.

And all of this pre-supposes that a Human player will always make the right and 'intelligent' choices and decisions; this is often not the case at all. Sometimes the random choices that an AI makes can be as good, or bad, as those made by a human player.

Take the British at Arnhem, for example. Faulty initial planning, collapse of communications, no overall leadership and even elite troops could not pull a victory out of the resulting mess.

I do understand the negativity associated with potentially losing control of perhaps all of your units and the subsequent loss of gaming fun; I just question whether the polar opposite, almost total control and full friendly intel, is the right/best/optimum answer.

This is, IMO, more of an issue for the upcoming WW2 title than perhaps it is for modern CMSF. I really do not see the average squad, or platoon even, cut off from its higher HQ doing much more than trying to get to its assigned target, or digging in for defence and awaiting further orders. My limited understanding of WW2 infantry tactics is that it was dominated by phase lines, frontages and timings.

I'm sure modern squads are trained to be much more self sufficient and pro-active.

Interesting discussion anway.

[Battlefront.com]

Michael Emrys,

The catch might be whether BFC can program such a system and I don't have any answer for that.

Ooo-oooooO!!!! I have an answer for that one...

no

We've discussed the impossibility of creating a wargame where units within your command are one thing, those not in command are left to the AI to handle. The points raised by others here show exactly why it's just not going to happen. The bottom line is that it's a commercially impossible game to create because the level of AI handling people expect is beyond even what a couple million Dollars in AI expert programming could purchase. Well, OK... if we were given a grant for a couple million I'd at least be interested in trying, but anything short of that isn't even worth considering. Not for a tactical wargame anyway.

Plus, it doesn't really address the problem because units under your control are unrealistically coordinated just like they are now

But not to worry... when we get CoPlay into the game it will be an optional way to play, just like RealTime is now. There's absolutely no reason why we can't have CoPlay *and* retain the single player game experience like we have now.

AKD,

Agreed that hit/kill info is somewhat unrealistic to provide the player. We've had that discussion a lot over many years, which is why it was made optional in CMx1. It's the sort of feature that people like or don't like, so it should be optional.

Steve