Background:
I am currently doing an online strategy game and one of its important component is battles. A typical battle will have 2 or more sides, each side able to field a number of players from individual unit commander (swordsmen, archers, etc) to the army commander. Personally I am a big fan of cooperative gameplay thus I would like to have a battle system that rewards the side who has a good team of players instead of the one who has the stronger army.

The battles are timed turn-based affairs - this is to make it much easier to implement technically (browser game) as well as to make it easier on the players itself - with 15 seconds turns where players can give command to their units or sub-units (that are commanded by other players), when time is up all units move simultaneously and resolves any combat. Having multiple commanders also has an impact on various gameplay elements, from command, effectiveness of a unit and Line Of Sight.

Line Of Sight is an interesting gameplay element that I feel has has yet to be fully explored in multiplayer games.

The design (simplified):
- Every commander on the field has a different perspective of the overall battlefield. For example, a unit of spearmen on the front engaging an enemy is probably only be able to see the enemy unit it is engaging as well as a few other units nearby. It may not be able to estimate the strength of the enemy effectively.
- Every commander will see different details of an enemy unit depending on the commander's ability. For example, two allied units may see different details from an enemy unit that is far off. One may see the enemy as a thousand-man strong spearmen unit, while the other may see a 500-man strong infantry unit.
- Overall commanders will have all their sub-units' information have depending on the overall commander's staff ability. (Still thinking this one through but not related to my question)

The problem:
With today's technology (Skype, Messenger, etc) players will be able communicate instantly thus making the whole "inability to get the whole battlefield picture" idea to moot. I have also done some experiments and observations on this idea and the results is kinda divided.

I would like to get some feedback on whether this LOS system is feasible or is it simply a nuisance?

Is there a way to implement a LOS system that achieves what I am looking for - simulating battlefield chaos that rewards players that work together as a team?

Thanks.

I have also done some experiments and observations on this idea and the results is kinda divided.[/quote]
If I may ask, what results were these?

Secondly, I have no technical advice. However, I think that instantaneous chat would actually make your game extremely enjoyable, I don't honestly see a reason to fear this. Some thoughts off the top of my head that may or may not be applicable:

1. Most people are terrible at giving accurate directions and passing on information.
   
2. Most people are also terrible at coordinating together. Games that implement such activities - like those with "raids", etc. - wouldn't be nearly so maddening if coordination alone didn't account for the majority of the difficulty.
   
3. Accurate observation is all about reference points. Good camoflauge works because the reference points for object differentiation are blurred or eliminated completely. Similarly, people get lost in the modern age with perfectly operational GPSes because they lose their geographic reference points and become confused. The easiest way to make information passing between parties difficult is to provide distinct views of a given situation with differing reference points. Separation such as you have gives you a significant amount of this built in; you just need to keep your views from being standardized.
   

Overall commanders will have all their sub-units' information have depending on the overall commander's staff ability. (Still thinking this one through but not related to my question)[/quote]
Frankly, given that this is a computer game - and the entire point of commanders is to command the battlefield from the back - giving them all the information makes sense. However, you can preserve communication/observation errors from the frontlines - e.g. over Skype or whatever - by delaying the update of their information until after a given player's turn has been made or performing an update on a round-by-round basis. This allows commanders with good teams to prevail since, presumably, their teammates/members at the front are much better at discerning accurate information and acting appropriately than their opponents.

Sorry, cut out quite a lot of information to prevent the initial post from becoming too long.

Experiment 1: Give various friends different pieces of information and ask them to share it among themselves. The information can be conflicting and/or unique to each friend. Interestingly, the more they work together the better they get, at the end of the experiment they worked out a way of relaying information (each person reports their information while a chosen person will collect and determine the overall picture) which is quite similar to actual military chain of command.

By the end of the experiment (4 different scenarios), they are able to share their information within 1 min (4 turns in game).

Experiment 2: Give same group of friends (which has their way of relaying information) and give them changing information. This is to test the real battle scenario where you will find 5-20 units moving around the battlefield. At first even their way of relaying information failed, they changed and organise themselves into a pyramid structure (which is again very similar to what a military chain of command is like in real life).

At the beginning of the experiment my friends become more overwhelmed as the number of moving units increases. By the end of the experiment they managed to work out a way to reduce this but are still not able to overcome the problem.

The experiments are very encouraging, my friends worked together as a team to overcome the problem, I can imagine those who don't work together will be completely overwhelmed by the system (which is both good and bad).

Let me go into more details for the designed battle system.

As stated before there can be one to many commanders in an army. Commanders can range from unit commander (commanding a unit), section commander (commanding one or more units of a section of the battlefied e.g. right flank) and overall commander. Each commander has limited number of command points to give units orders, that regenerates over turns. Thus the army with more human commanders and capable ones will have a huge advantage over the opponent.

Thus information restriction is mostly on the frontline unit commanders, whereas the section and army commanders will have a clearer picture of the whole battle thus allowing them to better direct their sub units in the battle.

Very interesting topic. This is only tangentially related, but I have been recently pondering about single-player or 1v1 strategy games where the play hinges on making sense out of limited and partly false information, and/or having a delay in receiving information and getting your orders executed.