WarCraft III Maps by Antistone
AoS Map Ideas
Notes for various modifications to the typical AoS template.
Update: Some time after this page was originally written, I began work on a map entitled "Fields of Stained Snow" that incorporates many of the ideas described below. You can download the current beta here.
It seems (at least to me) that in a typical AoS map, the focus of the map is on hero vs. hero combat, and all of the bases, towers, and computer-controlled units just serve as an interesting backdrop. You kill computer-controlled units, because it gives you resources (generally XP and gold), but you can leave them alone and they'll pretty much cancel each other out without any serious risk of upsetting the game. You destroy enemy buildings, because you need to do it eventually to win the game, but you don't have any serious chance of winning unless your heroes can consistently beat the enemy heroes. The computer units and buildings are very basic, with few (if any) interesting qualities, while the heroes are myriad, diverse, and have access to a bunch of skills, items, or other upgrades. In short, the typical AoS is basically just a glorified hero team arena.
Personally, I'm fond of games with some kind of strategic depth, and I think there's a lot of potential going to waste here. Thus, the following is a series of ideas and concepts for modified AoS maps designed to shift the focus away from hero duels and emphasize territory control, strategic maneuvering, and the large-scale clash of armies, while preserving (and maybe even enhancing) the feeling of being a proud hero, leading an army into enemy lands.
Some of these ideas are clearly designed to work together, but this is far from an all-or-nothing deal. Many of the ideas are separable, and could be implemented independently of each other. Certain ideas are even incompatible (or seem so). This is a collection of ideas, not a unified design.
Rather than the typical advance towers placed in each lane, which don't really do anything except prevent advancing into the enemy base too early, scatter key structures around the map (with lanes going through them) that can be captured and controlled by one team or the other. Various means of "capturing" these structures could be used, but the most logical one would probably be to require that you destroy the enemy structure, and then wait while a new one is built under your control (meanwhile defending it from being destroyed in turn). Many useful buildings are possible; for example:
These should be designed to give long-term benefits if control is maintained, not quick advantages. A team which pushes forward into enemy territory is expending resources (probably the health and mana of its heroes, if nothing else) and lengthening its lines, making it temporarily vulnerable to retaliation, so as long as controlling the points of interest grants a long-term strategic advantage, rather than an immediate one, there is an important strategic trade-off involved in taking them, and the enemy team has a window of opportunity to regain lost ground before they are put at too great a disadvantage.
In order to help prevent a stalemate where the two teams keep trading the points of interest back and forth, the maximum health of each structure could be reduced every time it changes hands, so that it becomes progressively harder to keep control of any point of interest unless you can utterly overwhelm the area and push the fighting line far away from it.
Rather than presenting simple, straight paths between bases, the paths twist, turn, and overlap one another (effective "overlapping" may be accomplished by judicious use of way gates). Some paths have their length concentrated closer to one base than the other (it will be necessary to use rotational symmetry, instead of reflective symmetry), so that the natural front line / midpoint of each path isn't along the center divide of the map, but it will be natural to have friendly units along one path going right past enemy units on an adjacent path.
Then, heroes and units can launch ranged bombardments across lanes where they come into close contact (preferably with differing elevations, giving one side or the other a tactical advantage). Thus, possibilities are created for "flanking" maneuvers. Additionally, heroes may be allowed to travel between lanes at selected locations, thereby enabling surgical strikes into enemy territory, or inversely, quick escapes.
Of course, one reason this hasn't been done is most likely because no one wants to go to the trouble of forcing the computer units through so many way points to keep them on the correct, convoluted path. But this concern is easily addressed by some clever pathing. Give heroes a collision size of 31, and other units a collision size of 32 or greater, so that heroes can fit through spaces that other units cannot. Place lines of flags along the edges of lanes where they are otherwise open, leaving just enough space between them for heroes to fit through. Do this even inside each team's base, with flag lines connecting to each team's main structure, so that it is reachable from all lanes, but no lanes have openings into each other large enough for non-hero units to use. Now you don't need any way points at all--all computer units can be blindly ordered straight to the middle of the enemy base, and all of them will stay perfectly within their own lanes.
Of course, similar results might be obtained by completely isolating the paths and giving heroes some form of flight or teleportation, but those abilities would have much more significant tactical uses.
Rather than having a few avenues of attack between the bases that are entirely separate, create a network of branching and interconnected paths. Thus, computer units moving along a path will periodically reach some sort of intersection that presents a choice of direction.
Players can control the distribution of their forces by using dummy abilities on "signpost" units located at each fork (this power could be given to the whole team or one designated player). Different options allow the player to specify that all the friendly computer units reaching that intersection should proceed in a certain direction, or that they should split up between several available options.
Players are only allowed to specify directions that get their units closer to the enemy base, to ensure that players can't create a cycle that will allow them to build up forces and unleash them all at once, but several branches might converge at a single intersection. Thus, players can separate and merge their forces, concentrating their numbers in particular areas while withdrawing support from others, thereby directing the flow of the battle.
It seems to be very rare for computer-controlled units to use any abilities, even simple auto-cast spells like Heal or Slow. Creating diverse computer units with different abilities could make the map more interesting, not only by increasing the variety of effects in battle, but by creating tactical reasons to focus on particular targets, potentially increasing the variety of battle tactics available to the players.
Rather than spawning a fixed set of units in each lane on each wave, players are given options to control the number and type of units produced (separately for each lane). However, every time a wave spawns, it costs team gold, based on the number and type of units produced. This would most likely need to be countered by some form of regular income, such as from controlling points of interest (see above).
This means players can lower production of units in one lane to boost them in another, or temporarily lower output in all lanes to save up money for a concerted push later on. Thus, players have an opportunity to focus their forces on a target of interest, or try to catch the enemy team off-guard. Of course, it would be necessary to have a maximum rate of production for each lane (12 might be a good limit, since that will allow all units produced in a given lane in a single wave to be ordered to move in formation).
Players could even be given the option to build and sell various unit production structures in each lane, to gain access to different unit-types and increase maximum production. For example, provide space to build 4 production buildings in each lane, with 2 unit-types per building. Players can select the building and change production to any combination of the two unit-types resulting in at most 3 units per wave (that's a total of 9 options per building), or can sell a building to make room for another one when they want to change the composition of their forces.
Controlling unit production could even be the dedicated job of one player, who would be responsible for carefully controlling unit production (and possibly some other management tasks, such as the signposts at lane branches), but would not be given a hero. Of course, the feasibility of this depends on how interested the average player is in taking on this role; you don't want to create a necessary job that no one wants to do.
Heroes are given skills designed to support a sizable force, rather than just to heal and buff themselves or to damage and debuff an enemy hero. In particular, lots of AoE buffs, such as Roar, which will be most effective when used with a lot of allies present.
Whenever a hero dies, it automatically grants a temporary but powerful buff to all nearby friendly units, which shout battle cries such as "I will avenge you!" or "For Frodo!" (or whatever's appropriate in the map). This buff might provide bonuses (for example) to movement speed, attack speed, and health and mana regeneration; the bonuses could possibly vary depending on the slain hero.
This will discourage targeting enemy heroes first during large engagements, since taking down the enemy hero will make the remaining enemy units much stronger; however, a hero who charges in alone, or who sticks around when his backup has mostly been killed, may be killed with impunity. This should help shift the focus away from hero assassination and focus it on the larger battle.
Heroes still experience penalties for dying, which may include a delay before respawning, being moved away from the battle, having low health and mana, and/or temporary debuffs. Thus, the penalties for dying may outweigh the temporary bonuses (unless the battle in which the hero fell was very large or very important). Dying need not be encouraged. However, this will still allow heroes to take greater risks and make strategic sacrifices.
Heroes are given special abilities that can be used to take charge of individual computer-controlled unit, up to a certain maximum (and possibly also abilities to return them to their regular behavior). Computer-units that are made part of a hero's personal "escort" will change color to match the hero, move in formation with (or around) that hero, and attack that hero's target. Thus, each hero is able to create a special strike force they can take with them wherever they go, ensuring that the hero never has to fight alone, and giving the heroes power to direct their team's forces in battle.
It's clearly important to impose a limit on the number of units that can be detained in this way, to prevent a hero from amassing too large a force in one place, but if you want to use the game's built-in unit formation logic, you can't have more than 12 units in formation anyway.
This idea would probably work particularly well with Branching Paths and not so well with Intertwined Paths, since it relies on the escort being able to travel with the hero, and there being some place interesting to go.
In almost any AoS map, heroes accumulate gold and XP by killing enemy units. This is an ubiquitous technique, because it's naturally supported by the engine--there's XP and gold for kills in standard melee, and it's very easy to adapt to a custom map. Unfortunately, the system sucks, and I really think it would be best to get over it.
The problem, you ask? There are several. First, it places far too much emphasis on the killing blow. Supposedly allied players are forced to compete with each other to score kills, and it even encourages such perverse behavior as killing your own units to deny XP and gold to the enemy team, which I personally feel should never be a legitimate tactic. Secondly, it allows heroes to get better at killing by doing nothing but killing--a directly self-reinforcing cycle, which promotes runaway leads, which simply isn't good for game balance. Sure, the whole point of fighting for strategic advantage is that you get some advantage out of it, but inserting some other steps into the loop provides an opening for enemy interference and makes game balance more stable. Third, the system is horrendously unfair to any hero trying to fill a defensive or support role, and by forcing all heroes to be offensively-focused it seriously depletes the potential variety of the map.
Fine, you say, but what can it be replaced with? I have several ideas on other ways resources could be distributed, which are outlined below. But, frankly, I think you'd be better off creating an AoS where heroes never level up and can't get items than using the standard bounty system.
Capturable Points of Interest (see above) can bestow resources over time; for example, giving all heroes on a team a fixed amount of gold and/or XP every few seconds while they're under that team's control. This was already discussed, but I thought it needed to be explicitly referenced in this category.
Each team can periodically spawn some champion unit, which would be comparable in power to a hero (or perhaps even stronger), but under computer control, and thus fairly stupid. The team may or may not have some control over the precise time at which the champion charges, and which lane he fights through. Regardless, the champion is powerful enough to seriously upset the balance of power in his lane. The enemy team receives a warning message when the champion is created, allowing them to prepare a concerted defense.
However, when the champion dies, all heroes on the enemy team gain a level (even if they're far away or dead at the time), and possibly some gold. Thus, a champion gives you a chance to push really hard on the enemy team for a while, but when he falls your opponents gain a long-term advantage. Enemy computer-controlled units could possibly also be given a temporary buff to represent the morale boost gained by slaying such a mighty opponent.
The rate at which champions spawn could depend on a variety of things; for example, controlling points of interest might increase the spawn rate for your team--but of course, that would be a double-edged sword. My favorite idea regarding the champions is that, the less territory (possibly measured in points of interest) that your team controls, the faster your champions spawn, as the desperation of your cause inspires them. This means that the losing team gets an opportunity to make a comeback (assuming that having less territory means you're losing), but if the winning team can dispatch the champion without suffering severe losses, they can solidify their winning position with the extra levels. Thus, if one team is winning by a little, the other team has a better chance of reversing the situation, but if one team is solidly dominating the battlefield, their lead increases quickly, so they can finish the game.
At this point, some readers may be wondering how this is different from the bounty system I so loathe. The key difference is that the XP rewarded has nothing to do with who kills the champion, whereas with the traditional bounty system, you need to be near to get XP, and need to score the killing blow to get gold, so the resources given to each player and to each team can vary widely depending on luck and tactics. Additionally, champions are a much more tightly-controlled resource; the designer can make the production of champions (and therefore levels) tied to whatever other game elements he likes, thereby making the reward ultimately dependent on the secondary goals of his choice, rather than simply killing. Unless the champion wins the game for the enemy team, you're really guaranteed the resources when he spawns (even if you need to wait until he dies to collect them), so you get XP by controlling the spawn rate of champions--killing them is a necessary step, but it only affects the timeliness of your XP, not the rate at which you get it.
All players are given resources (gold, XP, and/or whatever else the map uses) as time passes, at an equal rate, completely ignoring what they're doing or how effective they're being. Tthe rate might change as time passes.
Not a very interesting system, I'll grant, but it can be used to ensure that heroes become more powerful as the game goes on without tipping the playing field, and that can be a good thing (1) to give new players a window in which to learn about the game when they mistakes aren't as important, and (2) to help make sure the game lasts the time you want it to, rather than ending early or dragging on.
And, of course, this can always be combined with other systems.
All concepts and ideas on this page may be freely used by another in implementing any WarCraft III map. If you implement any of these ideas in your own map, closely following the descriptions above, please email me so that I can take a look.