DD21: Pushing Back the Fog of War

[ Edited Saturday, April 1st to correct IE incompatibility ]

While I continue catching up from GDC, and fighting back the jet lag, I thought I would dredge up something from the archives share this gem from the past. Dated 1998, a rumination on the Fog of War and how one pushes it back.

A number of these ideas have appeared in games since then, but there are a few in here that I haven’t seen done yet. I like the map reveal ideas in Physical Recon, especially if they were applied to a smaller scale tactical group in modern warfare, and I think there are some good ideas in the Field Intel section.

Enjoy!

---

Pushing back the Fog of War

The role of reconnaissance and espionage in RTS games

In the current crop of RTS games, all units are intelligence gatherers, simply pushing back the fog of war. Minor enhancements might include the ability to cloak or sneak. The most ambitious implementation in my experience has been the diplomat in Civilization, but the structure of reconnaissance been simplistic at best.

I propose a more ambitious, more realistic, (in feel if not in fact) and much more interesting solution. The proposal that follows is structured around a modern military scenario, but the principles involved could easily be adapted to any scenario.

All aspects of strategic information within the game I will label as Battlefield Intelligence. Battlefield Intelligence is then broken down into two categories, Physical Reconnaissance and Field Intelligence.

Physical Reconnaissance

Physical recon is broken down into two categories, Ground Recon and Aerial Recon.

Ground Reconnaissance

Is limited to line-of-sight, and is skewed in the direction of movement during travel. Instead of

Fog of War, the battlefield map is first revealed as a simple diagrammatic map. (Fig. 1)

Figure 1

As troops move over and through, the map reveals its more realistic contours. (Fig. 2)

Figure 2 (rough)

Any troop movement over land will reveal its physical properties, and will have a good chance of revealing moving or un-entrenched enemy units, chances diminishing with distance and based on size and characteristics of enemy units. I.e., easy to miss scattered infantry at the edge of the sight sphere, harder to miss a large, noisy tank., harder still to miss (most) aerial units. Effectiveness drops while troops are in motion.

With combat experience, chances of enemy detection improve for basic ground units.

More advanced units might include Scouts and Observation Posts.

A Scout would have increased sight-range (slightly better while moving, much better when stationary) and better peripheral detection, with better chances of detecting camped-down enemy troops, ambushes and snipers, as well as moving and un-entrenched enemy units. Scouts can be put in open-top vehicles, jeeps/humvees etc, but effectiveness should drop appropriately, say 20 to 30%, as opposed to a 10-15% drop while moving on foot. The scout is lightly armed and armored, but can hide more quickly and effectively than common ground troops.

Observation Posts would offer both increased sight-range and a higher vantage point to extend line-of-sight. Radar upgrade would offer telemetric observation of aerial units at great distance. They would also serve as a center for interception of radio communication and enable more secure point-to-point communication (see Field Intelligence).

Aerial Reconnaissance

All flying units will reveal some level of map detail, based on height, speed and observational design. If outfitted to reveal a high enough level of detail, basic units will have a good chance of detecting almost all buildings, roads and air fields, large ground units and congregations of smaller units. Movement of ground units increases their chance of detection. Aerial units engaged in combat will reveal little useful ground information. Because of the nature of modern air combat, combat experience does not necessarily lead to improved chances of air unit detection, as most of that type of detection is electronic in nature. Combat experience would help improve the chances of weeding out false contacts and potential friendly fire situation (which might be likely to crop up in team play where different nationalities team together).

As with ground units, visual reveal is skewed forward . Radar contact would be 360 degrees. The forward skew is more pronounced on ground reveal, visual aerial contact is forward biased, approximately 150 degrees (Figure 5). Individual craft design can of course modify these parameters, for instance, the addition of a rear-gunner or ground-facing camera.

Note: Units identified by non-visual means (Radar/RF/Infrared) should be identified on screen as colored outlines or wireframes, possible lower res at bottom of tech tree, gaining detail as tree advances. Units detected remotely and later confirmed visually would show normally with the colored outline overlaid.

Advanced units might include Spy Planes, AWACs, Low Altitude Recon, Helicopters, and Satellites.

Spy Planes are high-altitude aerial reconnaissance units. Flying high and fast, they reveal a swath of the map in accurate diagrammatic form. Upgrades allow unit to “snapshot” a section of the map in higher resolution, geo-physically accurate, perhaps black and white. No movement or updating.

AWACs are aerial and primarily electronic version of the Observation Post. Minimal ground reveal, except in the RF spectrum.

Low Altitude Recon is the infantries best friend. Get detailed, perhaps magnified snapshot of buildings and encampments, including an explicit count of visible troops, ground units, and a cursory building analysis.

Helicopters as hovering units, will be able to discern more ground detail than most general purpose air units.

Satellites are a late game addition, providing periodic, accurate updates of the entire diagrammatic map. They can also be called upon to provide higher res snapshots (see spy planes).

Some Final Thoughts on Physical Recon

It should be noted that as actual terrain is revealed it may differ in significant degree from the diagrammatic version. The degree of variance would be an embedded map-specific variable. At its lowest level of variance human artifacts, roads, buildings, etc, may shift. New roads may have been built, old roads grown over, for example. At higher levels, human artifacts my be absent in the diagrammatic map, and even geological features may be displaced. As detailed above, some forms of aerial recon, and as I will get into, some types of field intelligence will allow for more accurate updating of the diagrammatic map as the game progresses.

To reflect changing battlefield conditions, areas that have note been actively explored for a long period of time will begin to dim, then revert to diagrammatic form.

Samples of some imagined reveal methodologies are shown below in figures 3-5.

Figure 3

Figure 4

Figure 5

Field Intelligence

Intelligence in general operates only on the schematic level, updating the diagrammatic map and providing enemy “unit markers,” iconic representations of projected troop locations and movement (imagine air traffic control labels or Risk playing pieces). Unit markers can appear in non-diagrammatic, even occupied areas, indicating either faulty intelligence or still undetected units.

Intelligence also provides information regarding troop movements, strength, weapons , weapons production, production sites (buildings, as seen from the outside are somewhat generic, a factory is not too different from a warehouse or ammo dump), resources (current and projected), weapons technology and moral.

Field Intelligence is broken down into two sections, Intercepted Intelligence and Gathered Intelligence.

Intercepted intelligence

Is derived from intercepted battlefield communication, captured enemy troops, captured buildings and team chat.

For instance, if the enemy captures a weapons factory, they may be able to learn how to build one of your advanced weapons, or learn the location of troops carrying weapons developed there (Develop “Shredding” early!).

Hand in hand with intercepted intel is the development of Cryptography/Code -Breaking. This will limit the amount of useful intel that is actually acquired with intercepts. It also has the happy side effect of being able to plant red-herrings, deliberately uncoded information designed to mislead the enemy.

As is normal in warfare, the lower level the communication, the less likely it is to be coded. Therefore observation post near combat situations that have developed Intercept ability may be able to gain some useful information. This may include the location of hidden troops in the vicinity, or knowledge that reinforcements are being sent to the area (remote troops ordered to move within a certain radius of combat would have a chance of triggering the reinforcement message), including general unit type (ground or air).

Gathered Intelligence

This is espionage. I foresee two major sources of gathered intelligence, spies you send out, and enemy units that you turn.

Your spies can infiltrate a complex, attempt to turn enemy units, and perform sabotage. In addition to any one of these assignments, they can be given an objective (chosen from a list). These objectives might include steal a weapon design, a codebook, assess nuclear capability, etc.

The spy is of course most vulnerable when in the process of infiltrating/communicating with the enemy for the first time, and in fact this is how most spies will meet their demise. Once your spy is in place, however, s/he can only be detected when under your direct control (movement, updating orders, downloading gathered information). Every time you initiate communication, there is a chance that s/he will be detected (perhaps a color change, or audible signal, that will give away the unit if the enemy happen to be looking at the unit at the time). If the enemy has the unit selected (I.e., under surveillance) s/he will be detected.

Once detected, s/he will be flagged. At this point the enemy may kill the spy, leave them alive to feed red herrings (dangerous, as the spy will continue to report back all available intel, not just what is fed), or the enemy may try to turn your spy.

If turned, s/he will begin to give erroneous information. The information will be pseudo-random, hopefully believable; over/under estimates of troop strength, false troop movements and build-up. Specific false information can be planted , but detection protocol is under effect when the turned agent is under new master’s control.

Turned units are like normal spies, but more effective. As officers in the enemies chain of command, the are eligible for promotion and transfer, and as they move up the echelon, the quality of their intel increases, as does their chances for sabotage.

An aside about promotion and transfer. This assumes that installations like factories and training facilities need personnel to function. Turn a Lieutenant in a small arms factory, and eventually you may control a Colonel in communications. At certain levels of promotion, the turned agent will be able to control certain troops and production lines, allowing for much mischief, tempered by the threat of discovery.

Team Chat and Double Agents. Scenario: You successfully ensconce a spy in enemy headquarters and begin to intercept Team Chat. The agent, unbeknownst to you, is turned double agent by an enemy spy. There are, to my mind, three ways to deal with this (from a program logic standpoint).

• Cut off all future chat from interception. Easiest method, but prone to spark suspicion.
• Maintain flow of chat, with enemy team members aware that the channel is being monitored, possibly with an option for open and (truly) secure chat. A good middle ground.
• Computer replaces place names, numbers and cardinal directions with pseudo-random replacements.

It seems to me some combination of these three protocols could maintain a good semblance of team chat, while providing the appropriate detrimental information.

Beyond Intelligence: The Agent as Saboteur

• There are several ways to implement agent-based sabotage.

Level I: Infiltrate and Destroy

• A one-shot deal, with immediate payoff, and a swift end to the saboteur.

Level II: Infiltrate and Hamper

• A longer-term endeavor, which may go unnoticed for some time.

Level III: Infiltrate and Redirect Resources

• A longer-term endeavor still, requiring a turned enemy of appropriate rank to cause research to be misdirected, troops to be sent on wild goose chases, and buildings to be erected or torn down arbitrarily.

Level IV: Infiltrate and Suborn

• Same as Level III, but at higher rank yet. Sow sedition, hurt moral, cause units to attack each other.

Notes and Errata

Line of sight: A simple idea with potential CPU impact. Ground, and to a lesser degree Air units have an area of perception that governs how the map is revealed to them. Their line of sight restriction cuts off this spread as it is occluded by terrain feature, making it easier to mount terrain based ambushes, and to hide special facilities. It would also be a cool visual effect, as mountain loom large from the flat diagrammatic map, cutting off further recon.

Point-to-point communication: With the aid of an observation post, or possibly an AWAC, it should be possible to develop PtP communications, a tight-beam protocol that would further hamper enemy attempts to intercept.

False Contacts: Radar ghost images, false echoes, and garbled transponder Ids. These problems would be most evident in early development of Radar, and after the development of effective countermeasures.

Building Appearance: By their nature, some buildings will be easily identifiable. Airplane hangars, for instance. But where possible, the appearance of a building should not give away its function. This provides more of a challenge, and requires the deployment of intelligence units.

Attack by Unseen Enemies: Troops can lay down fire in an area, even if troops are not yet visible. Undetected troops firing on your men should be loosely defined by muzzle flashes and smoke effects, as well as stereo sound cues. Odds of visual detection increase substantially once engagement begins.

Day/Night Cycles: It would be much fun to include Fall-of-Night/Break of Dawn scenarios, where troops basic perception effectiveness changes with time. Fall of night would see the map grow darker, the sight range contract, and other cues take dominance, such as Radar and a hazy, monochromatic night-vision , coupled with increased reliance on sound and weapons flashes to pinpoint hotspots.

Weather: Like day and night, weather should have a profound influence on gameplay. Rain will slow troop movement and interfere with Radar (and Stealth technology!). Clouds will obscure ground units from some aerial detection (high flights and satellites). Fog obscures ground and air reveal, even infrared observation, constricting sight-range.

Climate: Like weather, a profound, but more predictable influence on gameplay. Deserts, rainforest, arctic conditions would all call for specialized strategies. And though not strictly speaking a climate, urban warfare could also be quite entertaining.

Engineers Corps: They build the roads, runways and temporary bridges, look for mines, set up perimeter defenses.