What's missing from wargames in general?

[Redwolf]

Hi, Broken,

Originally posted by Broken:

Originally posted by redwolf:

</font><blockquote>quote:</font><hr />

Two reasons:

1) too much effort if you see a unit with some orders, change your mind what it should do and then have to go and delete orders from some other units to free up "slots".

I see your point. However, I don't like systems that favor micromanagement, where a player can issue an endless stream of orders in five minutes of game-time. Perhaps a solution is to invoke a "command delay penalty" when a player issues a large number of new orders.

</font>

[flamingknives]

How about limiting the SOPs available only to certain experience levels, or having certain SOPs locked for certain troops.

[PanzerBonq]

What about wounded? In all of these games they are taken into account as losses in general and once as the soldier is hit he disappears from the game world. What about field hospitals or places where wounded are taken (in the small scale battles), which usually was not far away from the frontline.

I guess i want more logistics...

And sprites of the dead in CM, becouse the battlefield looks to sterile.

And one more word about units orders and time penalty. My thought was that unit orders not always depicts Commanders instructions, but simulates (for example) squad leader inteligence who must obey the general order for reaching destination. His commander (i think) does not tell his subordinates to first run to a treeline, then sneak for awhile, then run again etc. etc.

But i might be mistaken...

[Conscript Bagger]

Originally posted by redwolf:

I think the best way to describe what I want is that I want to limit the amount of micromanagement that affects behaviour in the next turns. But I would allow a player to plan far into the future.

So to speak, the number of orders and SOP changes you can set is unlimited and free, but most of them don't apply to actions in the next turn.

What I want is to limit the amount of tinkering that the player can spend on a single turn. It serves no purpose and wastes both player's time.

What I don't want to limit is the amount of pre-planning you can 1) plan 2) already enter into the game engine at any time you want. This doesn't affect how much time the player spents on all his turns on average, but it allows them to take a quiet evening to invest in a game and then take it easy the next turns.

I would draw a distinction between the amount of tinkering a player can do and the amount of tinkering a player must do in each turn. The latter is tedious like you said, and good SOPs and AI will be necessary to avoid it. But if you remove the player's capability to affect the next turn, then the time scale is too fine. Why have one-hour turns if it always takes three hours for your orders to go into effect?

Maybe a variable time scale is needed, where the amount of time represented in a turn is different depending on the situation? And/or an impulse system, where high-initiative units (veterans, recon, HQ) can receive (and execute) new orders every turn, while conscripts, disorganized units, or those engaged in combat get fewer opportunities.

[Redwolf]

Originally posted by Impudent Warwick:

I would draw a distinction between the amount of tinkering a player can do and the amount of tinkering a player must do in each turn. The latter is tedious like you said, and good SOPs and AI will be necessary to avoid it. But if you remove the player's capability to affect the next turn, then the time scale is too fine. Why have one-hour turns if it always takes three hours for your orders to go into effect?

Maybe a variable time scale is needed, where the amount of time represented in a turn is different depending on the situation? And/or an impulse system, where high-initiative units (veterans, recon, HQ) can receive (and execute) new orders every turn, while conscripts, disorganized units, or those engaged in combat get fewer opportunities. [/QB]

Both of these paragraphs assume a realistic and somewhat optimal working MicroAI. That's not gonna happen. A MicroAI better than human fiddling is basically impossible. Thus, if you don't put limits and force the player to choose what he wants to control tightly, then you will almost always see people spend gazillions of hours planning a turn in competitive play.

What I want is a strict limit of what you can do without compromising the realism of the game. I want an "efficient" wargame, a game where you get a realistic and thrilling experience out of as much planning as you want but a limited amount of clicking on the computer. Take CM and Close Combat 2 for example. I think few people doubt that CM is the better game in most respects, but CC2 is certainly much more efficient. You can play the same size battle about 5 times faster, for what is less than a factor of 5 realism drop (but certainly a factor of 10^8282 graphics drop ).

The variable time scheme has the problem that if one side's actions trigger a pause the other side might be tipped off that something is going to happen by the fact that time stops (or turn is short).

And as I said, I don't think generally "plain" delays like CM has (along with POA2 and Airborne Assault) will ever be a satisfactory thing to have, UNLESS you are actually building a command game where you can ONLY do what your highest officer can do. But seeing what each unit is seeing instantly but delaying orders to them is a realism break in itself.

Furthermore, this prevents you from solving the delay problem by what I call "post-event evaluation", you could do that in a true command game.

[ April 20, 2004, 07:18 PM: Message edited by: redwolf ]

[Conscript Bagger]

I missed the whole CC thing; I think I was wrapped up in TIE Fighter and Aces of the Pacific back in those days... so your comparison is unfortunately lost on me.

Good point about the variable time/warning, though I think most players expect something to happen in their games anyway (but I agree the timing could still hamper a surprise maneuver).

I don't know what to do about the "seeing everything" issue. One "solution" other games have used is to force the player to assume a unit's POV in order to direct its behavior. Some vehicle & flight sims have this, where you jump from one platoon to the next, having direct control only of the particular unit you're "riding" at the time while the AI takes care of the rest. However, I'm not personally enamored of this system to begin with, because the player still ends up doing it all, and I would not want it in a CM-level game, at least not without a more savvy AI.

What do you mean by "post-event evualuation"?

[Redwolf]

Originally posted by Impudent Warwick:

I don't know what to do about the "seeing everything" issue. One "solution" other games have used is to force the player to assume a unit's POV in order to direct its behavior. Some vehicle & flight sims have this, where you jump from one platoon to the next, having direct control only of the particular unit you're "riding" at the time while the AI takes care of the rest. However, I'm not personally enamored of this system to begin with, because the player still ends up doing it all, and I would not want it in a CM-level game, at least not without a more savvy AI.

Oh yeah, that's a catastrophy. It is taking the problem that you can invest major amounts of time for minor gains (but which may be decisive in competitive play) to extremes.

What do you mean by "post-event evualuation"?

I should probably finish the webpage on this, but the here is the quick version.

If you write a MicroAI it will screw up in unrealistic ways, no matter how hard you try.

To take a concrete example, lets look at the turret orientation in CM. How many times do you had the turret not facing the right way and get a tank waxed, and that means not the right way as opposed to reality? Real tank commanders are very good at keeping the turret in the direct of a threat, and on top of that a tank formation coordinates their turret facing to have everything covered.

In CMBO it was purely under control of the MicroAI (which usually chose not to do anything but keep it pointing forward), in CMBB and CMAK you either have the same or direct control via covered arc, but the direct control is very inflexible, just for starters you cannot modify facing after you leave the cover of a house.

Now, the difficulty in writing a MicroAI for turret facing is that it is hard to predict the future. It takes "real" intelligence to match a real tank commander's decisions. It is too hard to program.

But if you don't require your MicroAI to predict the future and take some abstraction, you can do something about it. The game engine can analyse the situation after the event and then come up with a more realistic assumption, at least a statically correct one.

Take for example the situation that a tank leaves the cover of a house to the side and get waxed from the side. In reality the commander would have turned the turret and would have faced the enemy.

In a post-event evaluation you don't try to come up with a good facing in advance, it is too hard.

What you do is you look at all the events without the turret facing and then, afterwards, make a decision whether the turret faced the attacker or not.

The key to realism here is probability tables. For this situation you would say, OK, it's a veteran tank, he is leaving right side cover, the chance that he would have faced this attacker to the side is 82%. If the tank is green the chance would be 46%. Then you evaluate the probability and insert into into combat resolution as if the MicroAI had given that turret facing before the fact.

Player's action play a role. If the player advised the tank to orient it's threat expectations to that side, you would raise these chances that the turret is facing the attacker. If the player gave different, that means actively wrong, advice, you would lower it.

This requires more abstraction than CM has. CM shows the turret and it would be off to time-warp it back to correct the facing. Furthermore, the turn end is a problem. But even an only slightly higher abstracted game like TacOps can do that find (it doesn't, but it could).

Hope that makes it clear.

I don't pretend to be able to write a good MicroAI (although I would spend more time on it than BFC and I would probably allow players to feed in their own).

What I would do is fight back MicroAI via 1) SOPs 2) and/or/not trees to modify SOPs based on events and phaselines in time and space and 3) post-event evaluation to ensure that the overall chances of MicroAI screwup or genius even out to what I define as assertion for the probabilities.

[Conscript Bagger]

Okay, thanks--I think I've mostly gotten what you mean. I suppose it's just hard to conceptualize without seeing a game that puts it all together.

[SlapHappy]

What's REALLY missing from wargames? That's easy! Not enough developers that really understand the slightest thing about real wargames...never played one (probably never will). Their only real concern is kissing the publishing companies posterior and pleasing a bunch of pants-wetting fanboys who's only concern is "will I be able to get it for my X-Box?" JasonC, you're right on the money with your analysis of wargames in general. Just go check out some of the inane posts on some of these other boards if you want to get a feel for what many consider a real "wargame" to be! Some of them are hilarious! If this is an example of the buying public, don't expect too many prizewinners in the future.