Buttoned and radioless AFVs

[nijis]

I posted this on the Russian infantry thread before, but it didn't get a response, so here goes again.

Contains Yelnia spoilers.

Lest I be accused of whinging, I'd like to start off by saying that the demo's great, the scenarios show off the changes in the system, the graphics are beautiful and I've learned to quit worrying and love the shockwave-free bomb, or shell as the case may be.

My one reservation is that radioless, buttoned-up tanks still seem to have an awfully easy time acquiring targets. I've had the T-34s in the Yelnia scenario motor up and down the line of foxholes, happily acquiring and dispatching infantry to their rear and flanks as soon as they popped up to throw a grenade bundle or whatnot. This means that Soviet heavies can't just break through the German lines (realistic), but can also go about methodically destroying all the enemy infantry on the battlefield (maybe not so realistic, although I can hardly claim to be an Eastern Front grog).

The tanks did admittedly have the advantage of an infantry battalion a few hundred meters to their rear borg-spotting for them. Probably until relative spotting can be introduced it will be impossible to get the infantry/armor balance of power just right; some scenarios will favor the infantry and others the vehicles. However, I did want to hear if any other players had the same feeling, if it came up on the Beta forum, etc.

I had one idea involving command delays for buttoned, radioless AFVs -- not only should they be stiff, but they should be random. Say, 65 seconds tick down, and each subsequent second there's a one in something chance that the tank finally kicks into action. This makes it much more difficult to coordinate moves. Maybe this should only apply if the buttoned tank is out of visual contact with its commander, and can't see the flags or flares come out of the hatches. Or maybe it's not codeable, or not realistic, or not worth an overworked BFC's while, or some combination of the three.

[ September 10, 2002, 12:35 PM: Message edited by: nijis ]

[Battlefront.com]

Nijis,

This is definitely a Borg problem. The heart of what you are talking about is the balance between what an individual unit should be able to see vs. what the game system says it can see. Penalize/reward one too much and the other becomes unrealistic. We have spent a lot of time trying to acheive the optimal balance. Until Realtive Spotting comes into being, we don't think we can make any more improvements.

Currently, buttoned up tanks without cupolas are nearly blind. Their reaction times suck, even with decent crews. Especially if the target is not in its frontal arc. If we penalized it any more, we would see a mountain of complaints about tanks being unrealistically useless, which would then cause an imbalance in the other direction.

Radios/C&C have *nothing* to do with spotting, so they are irrelevant. It is impossible to do anything with radios in this regard until Relative Spotting is introduced. Radios/C&C do increase reaction time, but not by all that much for decent crews. Vehicles, especially tanks, were supposed to be able to act on their own. In fact, during heated tank battles it is clear that most tanks were "on their own" in the tactical sense.

Steve

[MrSpkr]

Originally posted by Battlefront.com:

The heart of what you are talking about is the balance between what an individual unit should be able to see vs. what the game system says it can see. Penalize/reward one too much and the other becomes unrealistic. We have spent a lot of time trying to acheive the optimal balance. Until Realtive Spotting comes into being, we don't think we can make any more improvements.

Steve, is that going to be one of the major foci for the new game engine?

Steve

[MikeyD]

That's an aspect of gameplay I hadn't thought much of, balancing a realistic AI against the 'eye-of-God' human player hovering above the battlefield.

People often complain about how the AI sucks to one degree or another (a view I do not agree with, by the way) but a human can direct their buttoned radioless tanks to the far end of the the board to counter a threat that they couldn't possibly be aware of in real life. The AI's fighting with one arm tied behind its back, metaphorically-speaking.

So the only way to REALLY represent blind buttoned tanks is for the human player's armor to refuse outright to obey a command that was obviously based on 'eye-of-God' knowledge. By that I mean a straight allied AI vs axis AI battle with minimum human input. Hmmm... doesn't sound like much of a challenge that way.

[ September 10, 2002, 01:01 PM: Message edited by: MikeyD ]

[Battlefront.com]

Steve,

Oh boy... that is a big question I think the overriding goal is to simply recode everything to be more flexible. This is something that can only be done AFTER you have learned exactly what it is the code needs to do. CMBO taught us a lot about that, but you would be surprised how much we continued to learn after making CMBB. We think that two versions are really needed before one can make a significant improvement in the underlying game code.

But if I had to just choose one specific feature, to hold up above all other improvements, I would say it has got to be Relative Spotting. Not only does this fix most of the serious Borg issues (not all, since the player is a part of the problem) but it allows us to do many other fantastic things that simply aren't possible. For example:

1. Having a radio network. This allows spotting information to be passed from formation to formation (according to the network's structure!) without them actually having to spot things for themselves. This would make a platoon of radio equipped tanks that much more superior to a platoon of radioless ones, for example.

2. Better simulation of artillery. Currently we have the FOs in the game as a sort of (at times) abstraction of the ability to call down artillery. With Realtive Spotting and a radio/com network, we can have a far more realistic situation where artillery can be called down by different units depending on national doctrine, communications links, etc.

3. Multi-Multi player. While I think MMP would work great with CMBB, think of the possibilities of commanding a Company and only seeing on the map what your guys have seen or has been communicated to you through the communications network! I predict many nails will be chewed down to the point of pain

4. More realsitic fire control. Currently a platoon of tanks WAAAAAAAAAAAAY on the other side of the map can fire at anything they can shoot at, even if in reality they had no realistic way of spotting the unit in the first place. Restricting fire to only those units that you can actually see will make things realistically easier/harder for the attacker/defender depending on the specific circumstances. Instead of a dozen infantry squads opening up on some happless Flamethrower dudes, perhaps only one or two will fire away. This has a lot of implications for the Flamethrower team as well as keeping the other 10 infantry squads which you might rather have kept quiet.

The list could go on for a bit more, but I think you get the point. Relative Spotting is not just about more realistic spotting, but about a fundamental change in the way CM plays. We think it will not only be far more realistic (not perfect mind you!), but also a ton more fun!

Steve

[nijis]

I wasn't trying to suggest that the radios had anything to do with spotting, but with reaction -- that buttoned tanks without radios (and cupolas, now that you mention it) should have very little chance of reacting to a spotted target that's to their flank or rear. Given the amount of time that's clearly gone into the game, however, I'm certainly willing to believe that the armor/infantry balance is the best it can be within the limits of borg spotting. Certainly looking forward to all the cool relative spotting features.

[ September 10, 2002, 01:07 PM: Message edited by: nijis ]

[aka_tom_w]

Thanks for all the posts lately Steve (BFC-Steve)

This must bode well for the release of CMBB as by now things at your end may have slowed down now so you have a chance to do things like fix the front door and get around to doing chores and laundry

(I'm just guessing)

anyway

if anyone else here is interested in reading about Relative Spotting this thread will serve as a good reminder (or introduction) to the concept. There are 8 pages of posts and some of it is somewhat confrontational, but it never turned into a real flame war, (close perhaps) but just a few differing opinions as to how it should work and what it should accomplish.

here's the thread

this is page 7 of 8 with Steve comments in it:

http://www.battlefront.com/cgi-bin/bbs/ultimatebb.cgi?ubb=get_topic;f=1;t=024461;p=7

page 8 was the final page:

http://www.battlefront.com/cgi-bin/bbs/ultimatebb.cgi?ubb=get_topic;f=1;t=024461;p=8

This is probably my favourite aspect of the future potential/possibility of the re-write. Mostly because I think it will be very challeging and complicated to get it to work well so it will be both realistic AND fun to play (with EFOW of course )

-tom w

[ September 10, 2002, 01:23 PM: Message edited by: aka_tom_w ]

[MrSpkr]

Originally posted by Battlefront.com:

Steve,

Cool stuff about Relative Spotting

That is good to hear.

I am interested to see how you get around the player Borg sighting dilemma. The arguments over this issue will make for very good reading.

Steve

[demoss]

I am interested to see how you get around the player Borg sighting dilemma. The arguments over this issue will make for very good reading.

Perhaps the during-game views are what makes it up the chain of command to the CO?

Of course, this begs for an after-game full replay with complete spotting....

[xerxes]

I'm just happy to hear that multi-player is a top-10 item for the rewrite. Add in arty and relative spotting and I think you've hit on the key elements. Personally, having ever last detail of armor weak points, penetration, etc, is much less important than multiplayer, arty and spotting.

[Battlefront.com]

nijis,

I wasn't trying to suggest that the radios had anything to do with spotting, but with reaction -- that buttoned tanks without radios (and cupolas, now that you mention it) should have very little chance of reacting to a spotted target that's to their flank or rear.

The radio has nothing to do with the internal tank crew's ability, or lack of ability, to react to stuff around them in CM at all. And there is no way to account for that. So it would be incorrect to penalize a radio tank and a non radio tank differently. In reality the radioless tank should be at a disadvantage because his buddies can't radio warnings. But since we have Absolute Spotting, this point is totally irrelevant now. That is what I meant by the presence of a radio being irrelevant. Being buttoned up, no cupola, facing, and crew experience are the things that most influence how well the unit will do.

Steve

[dalem]

Originally posted by MikeyD:

That's an aspect of gameplay I hadn't thought much of, balancing a realistic AI against the 'eye-of-God' human player hovering above the battlefield.

Mikey-

It's a huge issue across all of the wargaming hobby; hexboard, miniature, paper, and computer. In real life, obviously, commanders have orders, and commanders have limited to zero knowledge of anything going on outside their particular "scope". In most wargames, however, we can easily assess the entire battle situation and decide what units need to go to what places to be most effective, or least ineffective. Even in games with some sort of hidden movement mechanic, once units are spotted and revealed on the board/table/screen, it is well-nigh impossible to "un-spot" them again - even a good mechanic for that event becomes overburdened when you have more than a few pieces to keep track of.

In board and tabletop games I've played, there are some really good attempts to cut down on the "god's eye view" the gamer generally has. There are event randomizers (cards, tables, chits, or timers) which prevent you, as a player, from always moving the units you want when you want, or shooting at what you want to shoot at. My favorite is a simple expedient of forcing your units to obey the last order they were given, i.e. move here, hold there, form here, until a roll on a rather unforgiving chart (this is a tabletop game I'm thinking of here) allows you to give that unit an order change. So you may see a battalion getting cut to pieces on the left side of the line, but if your other battalions are headed to the right, you have to keep moving them that way until they get their order change, etc.

There are many ways to gimmick a game to tone down the players' control, but I think the best solution will necessarily have to wait for the computer - in this case, an attempt by BFC (or indeed some other gaming company) to devise a spotting methodology that is at once relative and playable.

I look forward to it, but I don't envy them their efforts.

-dale

[ September 10, 2002, 05:46 PM: Message edited by: dalem ]

[gunnergoz]

I'm not a programmer, but I'm still curious as to the sticky issues of programming a relative spotting AI.

I assume that the current engine is set up so each unit has a data flag for visible/not visible, correct?

If so, is the challenge then to program a data base that will give a given A side unit a data flag for visible/not visible as to EACH B SIDE ENEMY UNIT?

Without necessarily getting too technical (unless you'd like to ), can any BTS programming guru clarify the basic issues and challenges?

And no, I don't seek proprietary information...

[Battlefront.com]

gunnergoz,

Yes, you are correct. Currently a unit is either spotted (to some degree) or not. The "flag" for this is retained by the data assigned to the unit itself. As you guessed, Relative Spotting means the reverse will have to happen. Each unit no longer knows if it is spotted or not, but rather what units it has spotted (to some degree).

This will, unfortunately, increase the amount of data that needs to be updated, retained, and utilized. We have no idea what this will do to file size, memory requirements, or game play speed. However, the impact should be very small and the benefit fundamentally huge. Plus, we are talking about 2 years away for this to be a customer reality. We think today's hardware can hack this easily enough, so future hardware will be even less impressed by the extra demands

Steve

[Pvt. Ryan]

Originally posted by Battlefront.com:

we are talking about 2 years away for this

Well, there it is. A release date for CM3!!!

[gunnergoz]

Thanks for the answer, Steve.

It then occurs to me that some programming and processing efficiencies might be gained with relative spotting, because the computer will only need to calculate potential fire exchanges between units known to be sighted.

For instance, with relative spotting, if A is seen only by X and B is seen only by Y, we only have to calculate two possible shots: one each for A to X and for B to Y. With absolute spotting, the computer -- in addition to the above -- also has to examine the possible shots between A and Y, then for B and X as well, because it must assume that A may see Y and B may see X.

To my untutored mind, this looks like 2 times the calculation work for absolute sighting.

In some ways it strikes me that you have a more complicated database with relative spotting, but the greater number of computational cycles to relolve a given turn might lie with absolute spotting.

Or am I totally off base with this intuitive view?

(PS you other programmers out there can chime in any time, BTW)

[aka_tom_w]

Thanks Steve

GREAT thread

needs a Bump

Steve says that CMII the rewrite is only "two years away" (I'm still betting on the early summer of 2005, but thats just my Doubting Thomas ™ nature )

thanks

-tom w

[aka_tom_w]

thanks Steve

[aka_tom_w]

With regard to the other thread this one might be interesting "Easy way to fix borg spotting"

http://www.battlefront.com/cgi-bin/bbs/ultimatebb.cgi?ubb=get_topic;f=23;t=000718

I don't think there is ANY "easy" way to fix Borg Spotting or Steve and Charles (and or some opinionated Combat Mission fanatic) would have figured it out by now IMHO

The engice re-write with Relative Spotting programed in from the get-go is the only real solution. (again IMHO)

-tom w

[ September 15, 2002, 07:46 PM: Message edited by: aka_tom_w ]

[Firefly]

Originally posted by gunnergoz:

Thanks for the answer, Steve.

For instance, with relative spotting, if A is seen only by X and B is seen only by Y, we only have to calculate two possible shots: one each for A to X and for B to Y. With absolute spotting, the computer -- in addition to the above -- also has to examine the possible shots between A and Y, then for B and X as well, because it must assume that A may see Y and B may see X.

To my untutored mind, this looks like 2 times the calculation work for absolute sighting.

In some ways it strikes me that you have a more complicated database with relative spotting, but the greater number of computational cycles to relolve a given turn might lie with absolute spotting.

Or am I totally off base with this intuitive view?

(PS you other programmers out there can chime in any time, BTW)

I'm pretty certain that what Steve means here, is that currently each unit has a flag set to spotted/unspotted which needs to be constantly checked as the turn progresses. In relative spotting the unit would require a flag for each unit that it can possibly spot. So that if the enemy force had a maximum of, say, 64 units, each unit would require 64 flags to be constantly checked in the course of a turn, which would be 64 times the work that the checking routine currently performs.

Of course I could be totally wrong, I often am .

[ September 15, 2002, 09:16 PM: Message edited by: Firefly ]

[moneymaxx]

Originally posted by dalem:

... My favorite is a simple expedient of forcing your units to obey the last order they were given, i.e. move here, hold there, form here, until a roll on a rather unforgiving chart (this is a tabletop game I'm thinking of here) allows you to give that unit an order change. So you may see a battalion getting cut to pieces on the left side of the line, but if your other battalions are headed to the right, you have to keep moving them that way until they get their order change, etc.

I don´t want to change the subject of this thread, but it is connected to a question I always had about CM.

In CMBO I often deleted movement paths for units I had given in previous turns because in the action phase a new threat, spotted by another unit, appeared (e.g. an AT gun or another tank). So I could prevent e.g. a light tank from beeing pulverized by a tiger.

I myself consider this a little bit gamey, but can´t stop to use this tactic .

Would it be possible/a good idea/accepted bc CM players to implement a reaction time in which the unit follow its last orders until it obeys the new ones? I think deleting a movement path is the equivalent to a stop order and should be penalized with a time delay. Especially to prevent "telepathic" behaviour of buttoned tanks without radio which react to a threat spotted by another unit.

[Wol]

The best system around the 'eyeballs in the sky' and 'Herr General, this discreet minute is over, you may instantly issue orders to every unit in your batallion (personally, and to each fire team with complete knowledge of the geography and placements of friendly forces and knowledge of every enemy unit known by any friendly unit' that i ahve found is in WRG WWII 1925-1950 (1988 edition) rules. Here you issue orders appropriate to units and subunits, conduct them in appropriate tactical modes (dependent on training and experience), and need to re-acquire units each bound.

In respect of most other current tabletop games, CMBO/CMBB is hopelessly inferior in terms of orders and c3 generally. Equally ludicrous is the "Fritz, KVs reported on report line alpha, we must commit the reserves. But no Hwerr General, the turn is not yet up!"

Never mind the "no need to use nominal tactics, orders/formations/consider personallities of sub-unit commander (imagine if Sgt Rock launched an attack with his platoon all by himself, or Capt Bowley turned your advance into a probe, or pretended he never got the order until prompted or relieved or you visit him, or Lt Pontiatowski orders a retreat all by himself because he hears of a tank in his rear and he aint gonna be cut off no siree)

And people worry about the slope on that Panther's armour!

But as this is a very pretty game, I could not care two hoots! No really. Relative spotting will be nice, but I get a lot of fun out of personally commanding every fire team oin the batallion/company whatever.unit I actually like playing as a god. I just do not think that we should get too worked up about realism

[MikeyD]

Wol writes:

"In respect of most other current tabletop games, CMBO/CMBB is hopelessly inferior in terms of orders and c3 generally."

Man, that must be one freakin' amazing tabletop you got there!

[dalem]

Originally posted by moneymaxx:

I don´t want to change the subject of this thread, but it is connected to a question I always had about CM.

In CMBO I often deleted movement paths for units I had given in previous turns because in the action phase a new threat, spotted by another unit, appeared (e.g. an AT gun or another tank). So I could prevent e.g. a light tank from beeing pulverized by a tiger.

I myself consider this a little bit gamey, but can´t stop to use this tactic .

Would it be possible/a good idea/accepted bc CM players to implement a reaction time in which the unit follow its last orders until it obeys the new ones? I think deleting a movement path is the equivalent to a stop order and should be penalized with a time delay. Especially to prevent "telepathic" behaviour of buttoned tanks without radio which react to a threat spotted by another unit.

moneymaxx-

I think the way that CM tries to address the issue of "instant command" is by having the "wego" system and the 60 second execution phase. Yes, you can give detailed orders every 60 seconds, which can be a little shakey, but then your units are technically out of your control for the following 60 seconds, and the TAC AI is making decisions. It's a compromise, but then, so much of what any game does is a compromise, right? And command execution delays are present. Plenty of times I've had a tank, radio or no radio, sit there being targetted or even fired at, and not respond.

-dale

[MikeyD]

That's a thought. Change the 'We Go" time from one minute to... three? That would certainly take a lot of 'gamey' issues out of the player's hands, but imagine the squeals of pain as you're watching the last 3rd of the turn unfold without you being able to intervene.

The only way I can think of to counteract often-criticised 'eye-of-God' gameplay is to take the human out of the loop as much as is practicable! But how far do you go before it stops being fun?