Skip to content

Fog of War

Strategic information hiding

Fog of war hides what the player's own forces cannot see. Computer wargames had it from their first years, and designers argued about it in print through the 1980s; Empire painted the unexplored world black, the real-time strategy games of the 1990s made the black map standard, and Warcraft II added a second fog that closes back over ground your units have left.

On a board wargame both players can see every counter. A computer can keep a secret, and the earliest computer wargames used that to do what cardboard could not: show each side only what its own troops could see. Wargamers called it the fog of war, borrowing the military term for uncertainty in battle, and for most of the 1980s they argued in print about how much of it a game should have. Then Empire turned it into a black map that scouting peeled back, and the real-time strategy games of the 1990s put that black map in front of millions of players.

Fast facts

  • The idea: the player sees only what their own units can see; everything else is hidden, or shown as it was last seen.
  • Early examples: Avalon Hill’s Tanktics showed “the locations of any enemy tanks which they can see” (reviewed 1982); Galaxy’s designer called its secret orders a “simulation of ‘the fog of war’” in the same issue.
  • The argument: “one of the most popular catch-phrases of computer wargaming” (Computer Gaming World, 1985); in 1988 Chris Crawford wrote “dump fog of war”, and Dave Arneson called it “one of the biggest ‘plus’ factors”.
  • The black map: Empire (Interstel, 1987) starts with the world “a tapestry of black”, nine squares known out of 5,684.
  • The real-time split: most real-time strategy games had a black shroud that exploration removed; Warcraft II (1995) added an optional fog that hid enemy units again once yours moved away.

The first computer wargames

The second issue of Computer Gaming World, in early 1982, shows both uses of the idea side by side. Bob Proctor’s review of Tanktics, Avalon Hill’s tank battle on the Eastern Front, described a game played on a cardboard map with a computer as referee and opponent. “At the start of your turn, the computer displays the positions of all of your tanks and the locations of any enemy tanks which they can see.” Sighting was a matter of chance beyond the next hex, and Proctor’s one complaint about the system was that the game kept no memory of it: “it sometimes happens that you can see a tank on one turn but can’t see it on the next, even though it continues to fire at you!” Tanktics was Chris Crawford’s design, which he had first sold himself at the end of 1978.

In the same issue Tom Cleaver, the designer of Galaxy, a space-conquest game for up to 20 players that Avalon Hill was distributing, explained his own version. “The orders which you key into the computer are invisible, so no one can tell what you are doing. This simulation of ‘the fog of war’ makes it possible to launch surprise attacks on your opponents.”

Two years later Karl Wiegers explained the term to Antic’s Atari owners in a survey of computer wargames: “A computer is also an ideal tool for introducing the ‘fog of war’ into a war game. This term refers to the element of uncertainty and randomness that is a constant aspect of real combat and that adds spice and realism to a simulation.” Reviews began to single it out. In Avalon Hill’s Gulf Strike, Computer Gaming World reported in January 1985, enemy unit types were shown “only when a friendly unit is ‘adjacent’. This simulates the ‘fog of war’ and prevents you from knowing what the enemy units are until you have recon’ed them.” In SSI’s Vietnam game Nam, Ahoy! wrote in 1986, “Hidden movement, which keeps enemy units off the board until they move or engage in fighting, produces the authentic feel of ‘the fog of war.’”

A wargamers’ argument

By September 1985 it was, in Jay Selover’s words in Computer Gaming World, one of “the most popular catch-phrases of computer wargaming”. His article sorted the games on sale into three kinds. A few ignored hidden information altogether, “a game in which all units of both sides are open to free inspection”, which “wastes one of the greatest assets of the computer”. A few made battlefield intelligence “a focal design point”: he named SSI’s Napoleon’s Campaigns 1813–1815 and Carrier Force, Simulations Canada’s Grey Seas, Grey Skies and Fall Gelb, and SSG’s Carriers at War. Most fell between.

Selover was sharpest about programming shortcuts. Avalon Hill’s Close Assault kept enemy squads hidden “until one of your units actually ‘sights’ them”, but once seen, “they remain on the map for the rest of the game. It does not matter how far away they move or how many buildings are between your units and them, they always remain in clear view!” He put that down to “a limitation inherent in the programming approach used in Close Assault which made it too difficult to rehide exposed squads”. Hiding something again, once revealed, was the hard part.

In April 1988 the magazine ran a round robin of designers, prompted by a letter suggesting that “extreme limited intelligence contributes to a computer phobia” among wargamers. The replies went both ways. Crawford owned up to “an unflattering priority”: a reviewer in the same magazine, he recalled, had “roasted” Tanktics for a fog of war system that spoiled the game for him. His conclusion was blunt: “Most people don’t buy wargames to study military history, they buy them to have fun… So give wargamers what they want and dump fog of war.” Stephen Newberg of Simulations Canada answered complaints about plans going wrong with “Tough!” Dave Arneson, the co-creator of Dungeons & Dragons and a historical wargamer before that, took the other side: “Creating an effective Fog of War is one of the biggest ‘plus’ factors in computer simulations today. Not knowing what is going on is certainly the most common factor found in wars throughout history.” His remedy for the frustration was a debriefing: “A program should have an ‘After Action Report.’ This would allow players to see what happened that he couldn’t see during the game.”

British reviewers expected the same. When CDS’s Tank Attack turned out to be a board game whose computer kept no map and hid nothing, Your Sinclair’s reviewer complained in May 1989: “The computer should be used to simulate the ‘fog of war.’ You should only get to see enemy units when your” own troops could see them.

Empire: the black map

Empire is where the fog became a picture. Walter Bright wrote it in the 1970s, “while he attended Caltech”, Amazing Computing said, and rewrote it for the PDP-11 minicomputer, where it spread. Interstel advertised the “improved authorized edition”, by Bright and Mark Baldwin and distributed by Electronic Arts, as available for the IBM PC, Atari ST and Amiga in October 1987.

Todd Heimarck’s review in COMPUTE!’s Atari ST in February 1988 describes the opening: “At the beginning of a game, most of the world is terra incognita, a tapestry of black. Your first priority is to send out scouting parties to discover some of the 50-100 cities. In turn one, you own a city. The map tells you the status of your city and its neighbors — only 9 of the 5,684 squares.” A world-map window showed “the entire known world (with unexplored territory in black)”, and “the role of the maverick explorer” was “the key to the early rounds”. Because the program could generate a new world each game, Computer Play noted, even an experienced player started blind, which “presents a ‘fog of war’ absent in the more traditional and historical simulations”. That fog, it concluded, made Empire “a game worthy of inclusion in the wargamer’s library”.

Real-time strategy

Westwood’s Dune II (1992) carried the black map into real time, and within two years it was the reference point. Previewing an overhead action game in 1994, Amiga Power explained that “you can only see areas of the map you’ve explored, so like Dune 2; the screen’s black until you explore everything.”

Not everyone liked it. Martin E. Cirulis’s review of Command & Conquer in Computer Gaming World in December 1995 called the “blacked-out terrain that must be explored by units in order to be revealed” “the crudest form of Fog of War”: appropriate for Civilization or Warcraft, he thought, but it “has no place in any game with a Sci-Fi premise”. “I’d rather be able to see all of the terrain in question and have enemy units appear only when they get close enough.” A caption put it more briefly: “you’ll die several times in each scenario just trying to unveil the terrain.” The magazine was still grumbling a year later about “ludicrous fog of war rules: despite satellite technology, you couldn’t see a tree a hundred yards away until you sent a scout to recon.” When the game reached the Nintendo 64 in 1999, Nintendo Power explained the same black map to a new audience: “The fog of war is portrayed as a black cloud… As your troops move into an area, the fog will lift and new territory will be revealed.”

Blizzard’s Warcraft II (1995) did what Selover had found missing in 1985: it hid things again. In the first Warcraft the land around the player’s town hall was unexplored at the start and showed on the maps “als große schwarze Fläche”, as a large black area, until units walked into it. In Warcraft II, as IGN Unplugged’s history of the genre put it in 2001, “While most real-time strategy games had already incorporated a black shroud that was only eliminated through exploration, Warcraft II went a step further. The fog of war meant that you could only see enemy units that were within the line of sight of your own units. Now you had to rely on scouting patrols and observation posts to bring you detailed intelligence.” It was an option. A German magazine explained in 1997 that strategists who found nothing too hard could switch it on: even in explored territory they then saw the enemy only around their own units, and the rest of the map and its enemies vanished into fog as soon as their troops moved away. Buildings stayed on the minimap, but only those that had been there on the first scouting trip.

Blizzard’s next game softened the edges. Computer Gaming World’s preview of StarCraft in November 1997 promised that “the fog of war will be blended, to make for a more realistic darkness effect”.

How it works

The Warcraft II model gives every map cell one of three states:

  • Unexplored: never seen, drawn black.
  • Explored: seen before but not now. Terrain is drawn, with buildings as they were when last seen and no units.
  • Visible: inside the sight range of one of the player’s units. Everything is drawn, including enemy units.

A common way to build this is to keep two flags per cell. Each game tick the program clears the “visible” flag on every cell, then walks through the player’s units and sets it on each cell within each unit’s sight radius, setting the “explored” flag at the same time. “Explored” is never cleared. The renderer then draws enemy units only in visible cells, and draws explored-but-not-visible cells from a remembered copy of what was there. That remembered copy is what Warcraft II’s minimap showed: the buildings found on the first scouting trip, not the ones standing now. The other half of Selover’s 1985 complaint, the Close Assault bug, is the case where the program sets “visible” and never clears it.

Whether the computer opponent obeys the same rules is a separate choice, and it is invisible to the player. The wargame designers of the 1980s argued about it as a matter of fairness; the real-time games mostly left it to the reviewers to guess.

Sources

  • Computer Gaming World issue 2, January–February 1982: Tom Cleaver, “Galaxy: Designer’s Notes”, pp. 13–14 (up to 20 players, secret orders, “the fog of war”); Bob Proctor, “Tanktics: Review and Analysis”, pp. 17–20 (enemy tanks shown when sighted, probabilistic sighting, no memory of previous results).
  • Karl Wiegers, “Atari at War”, Antic vol. 2 no. 11, February 1984, p. 44.
  • Computer Gaming World issue 20, January 1985: Gulf Strike review (unit types shown only when a friendly unit is adjacent).
  • Jay Selover, “Fog of War: A Critique of Limited Intelligence in Wargames”, Computer Gaming World issue 23, September 1985, pp. 15–16.
  • Steve Davidson, Nam review, Ahoy! issue 33, September 1986, pp. 53–54.
  • “‘Fog of War’: A Clearer View”, a round robin with Chris Crawford, Stephen Newberg, Roger Damon and Dave Arneson, Computer Gaming World issue 46, April 1988, pp. 24–26.
  • Tank Attack review, Your Sinclair issue 41, May 1989, p. 52.
  • Interstel advertisement for Empire, COMPUTE!’s Atari ST vol. 2 no. 7, October 1987 (authors, platforms, “written by Walter Bright in 1977”).
  • Todd Heimarck, “Empire: Wargame of the Century”, COMPUTE!’s Atari ST vol. 3 no. 9, February 1988, pp. 48–51.
  • Amazing Computing vol. 3 no. 12, December 1988: Empire review (Caltech and the PDP-11).
  • Empire review, Computer Play, November 1988.
  • Amiga Power issue 40, August 1994: news preview of Apache’s untitled overhead action game.
  • Martin E. Cirulis, Command & Conquer review, Computer Gaming World issue 137, December 1995.
  • Computer Gaming World issue 147, October 1996: comment on Command & Conquer’s fog of war.
  • Nintendo Power issue 118, March 1999: Command & Conquer for the Nintendo 64.
  • Bestseller Games (Germany) issue 17, August 1997: Warcraft manual (unexplored land as a black area) and the Warcraft II feature (the optional fog of war and the minimap).
  • IGN Unplugged issue 6, September 2001: history of real-time strategy.
  • StarCraft preview, Computer Gaming World issue 160, November 1997.

See also

Checked. Verified against primary sources and signed off by a person.