Mapmaking Fundamentals - Spawn Design

[WindHunter]

Mapmaking Fundamentals - Spawn Design

This was originally written as an internal design document meant to guide our current development of new campaign maps. It has been shared here for interested players and as an aid for aspiring community mapmakers. 

Introduction

BattleForge campaign maps have a particular feel to them, and this feel is a fundamental component of the unique experience which is playing BattleForge. The goal of this design guide is to use existing campaign maps, particularly the best campaign maps, to understand and to categorize the distinct features of the spawns in BattleForge’s campaign, such that we can better design future campaign maps. This guide is therefore intended to categorize the design choices of the original devs and to act as a reference for future development for both community and official map development.

Legend

Given the nature of map design, this guide uses a high density of visual examples to illustrate map design principles and standard practices. Across all examples, a standard legend is used.

• Purple – is used to highlight NPC camp formations. Purple text is used to describe the various dynamics occurring in the enemy camp.
• Light Blue – is used to highlight spawn locations and any special interactions which these spawn locations enable.
• Orange – is used to highlight non-standard flanking routes available to players. In the majority of cases, orange designates available paths for flying units, but it can also represent potential cliffing opportunities, particularly for artillery units such as Firedancer and Firestalker.
• Red – is primarily used to highlight player attack routes. It is secondarily used to highlight key features of camps which affect said attack routes.
• Black – is used to highlight terrain features.
• Green – is used to highlight map objectives and any locations that might be directly connected to those objectives.

Table of Contents

1. Spawn Types
2. Defensive Maps (Defending Hope)
3. Hybrid (Attack/Defense) Maps
   A. Nightmare's End Analysis
   B. The Guns of Lyr Analysis
4. Escort Maps
   A. The Treasure Fleet Analysis
5. Conquest Maps
   A. Ocean
   B. Mo

Spawn Types

Enemy spawns come in four primary types: one-time, timed, conditional, and permanent. These spawn types are not exclusive, in fact it is normal for maps to have multiple or even all four types at once, and often specific spawn interactions constitute multiple types simultaneously.

1. One-time Spawns: These spawns are typically scripted based on the fulfillment of a map objective, such as the Raven fleet’s approach in Ocean when the player clears enough of the center island, or are based on proximity events, such as the many “ambush” events in maps like Encounters with Twilight. Regardless of how they trigger, these spawns only happen once.

2. Timed Spawns: What is meant by timed spawns are those spawns which are primarily triggered based on the passage of time in-game and not player actions or map conditions. These types of spawns are quite rare on their own but are frequently combined with other types of spawns. For example, the permanent spawns on Convoy and Nightmare Shard steadily scale in strength based on the amount of time which has passed. A time component is almost always included in the case of permanent spawns to pressure the player or to keep pace with the player’s growing strength. The attack waves on Siege of Hope and Blight are both conditional timed spawns. If the player fails to kill the spawn buildings in time, a large one-time attack wave is released after a specific period of time. Conditional spawns often contain an alternative time provision. The spawn actives after X objective is fulfilled, or if Y time has passed.

3. Conditional Spawns: Conditional spawns are tied to map triggers which change on their own or due to required player actions. They can be temporary or permanent. Spawn buildings are the primary example of conditional spawners. The spawns are directly tied to the building’s continued existence; destroy the building, and the spawns stop. Many spawners also have additional conditions which must be fulfilled before they begin to spawn units at all. Some spawners only replenish units tied to specific camps, while others only launch attack waves once the player has reached a certain location or has activated a specific trigger condition. In random PvE, only the spawners directly adjacent to the player’s location initially launch attack waves, but if the player attacks a camp and kills an entity within that camp, all adjacent camps also begin spawning attack waves from that point forward. Sunbridge and Slavemaster are examples of conditional spawns tied to player action. In both cases, while the spawns are permanent, the player has the agency to direct the spawn paths via a switch and a fire emitter respectively. In the case of Sunbridge, once the Amii Power Shrines cease to exist, the gate-based spawn waves also cease.  

4. Permanent Spawns: Once begun, permanent spawns never stop. This type of spawn is a common feature in defensive maps such as Defending Hope and Guns of Lyr. Non-defensive maps do not typically have unconditional permanent spawns, and if they do, they are typically limited in either strength or location. Mo is an example of a non-defensive map with permanent spawns.

Defensive Maps

Defensive maps are those maps where the player’s sole primary goal is the defense of an objective or a set of objectives. Any expansion made by the player is meant to better facilitate his defense of the existing objective and not because he needs to conquer additional areas of the map to complete other required objectives. Defense of Hope is the only current official campaign map which is a pure defensive map (Ascension map 1 is the next closest example). Despite only having one such map as an example, we need not fear because Defending Hope is a superbly designed defensive map that contains all the necessary components to teach us how a map designer should design a defensive map’s spawn mechanics.

Defending Hope - Spawn Wave Interactions

1. Well Defined Defensive Perimeter – The area which the player must defend is visually distinct from the surrounding area, being protected by naturally existing terrain and buildable walls. This defensive area has overhang spots, occupied initially by Defense Towers, which can attack in a wider arc, as well as plenty of space to build defenses behind the walls. Areas from which enemies normally do not approach are blocked by Lyrish houses, adding an aesthetic flair while also preventing players from building otherwise inefficient defenses.

2. Alternating Attack Waves – Spawn waves on Defending Hope alternative between approaching from the north and from the south side of the city. This allows the player to opt for micro-managing archer units on both the east and west sides, rotating archers between the northern or southern wall depending on which direction is currently under attack. A skilled player can then save up power more quickly to make the transition to T2 or T3.

3. Spread Out, Multi-directional Attacks – While the walls on the east and west sides are close enough that archers can be rotated between them in the early game, they are far enough apart that the player is required to invest in defenses for both the north and south walls on each side, or to invest in a permanently mobile ranged component which can be rotated as needed. This multi-dimensional nature of the defense is important because it keeps the player engaged while also reducing the power of spells. If the units all approached from one-direction, the player would only need to invest in one set of defenses while also being able to destroy most incoming waves with spells, quickly leading to a feeling of stagnation (An example of this stagnation would be position 4 on Bad Harvest, which is initially very challenging, but once defenses are established, becomes dry and boring). With four directions to defend, permanent defenses are encouraged at each location while spells function as a means to plug gaps that form in the player’s perimeter, as their charge limit and cooldown prevent continual usage.

4. Optional Non-fortifiable Expansion Zones – Outside of the singular exception of aiding Rogan Kayle to enter the city, the player need never wander outside of the city’s defined defensive perimeter. Yet, if he choices to, he is rewarded with the possibility of many Power Wells and access to T4. Notably, these areas of expansion are shorn of any means of fortification while also being under constant attack. These means that the player must calculate the potential risk of both venturing forth from the city and investing in additional hard-to-defend assets, but if he proves capable of doing so, he earns himself a permanent benefit in terms of additional power and access to higher tier cards.

5. Permanent & Conditional Spawners – Instead of only spawning enemies from caves, the original designer chose to introduce 4 spawn buildings which the player can destroy. This gives the player more to do, while also forcing him to balance the risk/reward between investing into an expeditionary force that takes power away from the defenses and the benefit of permanently reducing the enemy’s offensive pressure. The buildings themselves have significant defending forces and are XL-spawners with 3200 life points, meaning that the player must do more than just run around erupting them to death.

Hybrid (Attack/Defense) Maps

Hybrid maps are those maps which mix player responsibilities between attacking and defending. These maps require the player, and any potential teammates, to move forward and conquer new areas of the map while defending map objectives (not Power Wells & Orbs) in other areas.

Hybrid Maps: Nightmare's End

Nightmare’s End - Player Defense Points & Enemy Spawn Locations

The issue with the spawn waves on Nightmare’s End is not so much that spawns can attack every player location on the map, though this is indeed an issue, but that none of the places the player is required to defend are capable of being fortified outside of the initial starting location (which eventually is no longer attacked as the player’s T3 and T4 orbs redirect enemy waves to themselves). Players are required to try and cram defensive structures and units into small areas where buildings often block each other from attacking. These same buildings cannot be built into a coherent frontline due to a lack of space, meaning that even if the buildings placed in the back can attack, they are blocked from reaching the ranged units targeting the building in the front. Additionally, there are no walls or other mechanisms by which the player can slow the enemy’s advance, meaning the frontline units or buildings must also tank the incoming wave. Combined with being the largest map in the entire game by far, making reinforcing areas with existing units nearly impossible, and the fact that enemy waves respawn nearly instantly in close proximity to their intended targets, the spawns on Nightmare’s End are truly a nightmare. Consider then, that after establishing all these defensive networks, the player is suddenly told to abandon them and defend yet another non-fortifiable location from a 5-minute-long onslaught. The fact that this last defense is in no way foreshadowed except for a small wall south of the Forge Shard, is a fitting illustration to how poorly thought through and poorly implemented the spawn mechanics are on this map.

A final point worth mentioning about this map is that the trigger conditions for spawning new attack waves and increasing the strength of existing attack waves are poorly defined. For example, one side of the map allows the player to build their T3 without destroying the concomitant Amii Power Shrine within the same camp. This in turn prevents the map from spawning the intended attack waves to attack the player's T3 location, because the trigger for spawning these waves is not the player building their T3, but the Power Shrine being destroyed. Setting aside whether or not the T3 should even be attacked in this manner, the key point here is that the trigger conditions for spawn waves and map events need to be carefully defined to avoid player exploitation. This same issue occurs on Behind Enemy Lines, where the trigger for increasing attack wave difficulty is tied to destroying the Twilight spawn buildings in the T2 and T3 camps. Said buildings are placed in such a way that players can build their T2 and T3 without ever destroying them, thereby trivializing the map in the process. In both cases, simply tying the wave increases to the player achieving a higher tier would have fixed the problem. 

Nightmare’s End - Southwest Quadrant

Hybrid Maps: The Guns of Lyr

The Guns of Lyr - Defending Player Progress & Attacking Player Spawn Waves

In my opinion, Guns of Lyr is an overly complex, badly designed, and badly balanced map that players have figured out how to exploit in a way that actually makes it enjoyable to play. An entire document could be written solely on how poorly thought through and executed the map's mechanics are in the final product we see in-game. While such a document could perhaps prove useful to the team, we will satisfy ourselves here in pointing out three issues with the map related to how its spawns function. The first is related to what triggers the infected Twilight camps to start sending out attack waves against the player in the "attack" position on each side. These spawns are actually triggered not by the actions of the attacking player, but the defending player. The map suggests to the player, with its set of retreating walls, that the initially defending player is supposed to forfeit their starting base and replace it with the wells and monuments found on the route of the Kobold Engineer which they are told to clear. Simultaneously, the respective side's attacking player is supposed to clear the 3 Twilight camps before transitioning into helping the defender. The issue is that the 3 Twilight camps do not actually trigger attack waves based on interactions with themselves or how much time has elapsed in the map, the camps only replenish defending units lost in attacks on the camp. Instead, the attack waves of these camps are triggered by the defending player attacking the units in camp 1 marked on the minimap above. This in turn leads to a situation where knowledgeable defending players simply do not expand past T2, but wait for the attacking player to reach T4 and trigger the final spawn wave early to end the map. Regardless of intention, it simply does not make sense that the trigger condition for spawning attack waves from these infected camps is tied to attacking that specific location on the map. A valuable lesson can be gleaned from this mistake, which is that spawn triggers for spawn waves need to be carefully defined so as not reward players who exploit the map while punishing those players who engage with the map in the intended manner. 

The second issue we will discuss in regards to Guns of Lyr is the ability for the player to avoid the map's mechanics and trigger the final spawn wave before actually achieving the necessary map objectives. Map designers need to be cognizant of the ability of players to damage locations behind "impassable" terrain features and take the necessary steps to safeguard these locations or punish the player (in a logical way) for doing so. In the case of Guns of Lyr, the original devs could have responded to the issue in several different ways. An easy method would have been to simply not spawn the final Twilight Manifestation building until the gate was opened, or to shield the building until said condition was met. A more creative approach would have been to immediately trigger other final spawn waves from any remaining Twilight Infestations on the map, making the condition of having to destroy these infestations meaningful, lest these other spawns bypass player defenses and kill Rogan. 

The Guns of Lyr - Northern Defenses

The third and final issue to discuss is how the map's terrain stifles the ability of the primary Twilight attack waves to be meaningful threats by forcing them through single path chokepoints. The map is clearly designed in such a way that the original devs assumed player defenses would steadily move backwards in the face of increasingly difficult attack waves. The map even destroys the original walls in an attempt to force the player to abandon their already constructed fortifications. If players acted in this way, it would cause early waves to attack a single wall, while later waves would eventually attack two and then three walls, spreading out enemies and player defenses over a wider area. The issue is that players do not act in the intended way and are not incentivized by the map to do so. It is always better to defend a single location versus multiple, and it is a hard pill to swallow to voluntarily sacrifice both existing defenses and your starting location, thereby accepting a permanent weakening via lost power and charges. Besides these reasons, allowing enemies to spread out, when you can instead force them through a narrow chokepoint would just be silly. The funnel at the top and bottom of Guns of Lyr clumps together enemies for players to then smash with powerful area-of-effect abilities and spells such as Worldbreaker Gun, Cluster Explosion, and Frenetic Assault. At the same time, the closeness of player defenses means that a single Protector's Seal or Revenge can cover the entire defense, which would not be possible if players voluntarily spread themselves between 3 separate lanes. For the map's defensive gameplay to succeed, it would have to be either more forceful in removing players from the initial chokepoint via harsh penalties or more rewarding in terms of benefits accrued. Another better option would be for the map to spread enemies out initially instead of shepherding them through a perfect kill box. If the devs wanted players to have a last stand location, it would have been better to allow them to set one up directly surrounding Rogan's Stronghold in the center of the map. This would make the map reminiscent of a medieval fortress, with large hard to defend edges that eventually coalesce into a single fortified keep at the city's center. Long story short, do not funnel enemy attack waves through a single narrow chokepoint, spread them out of a wider area or create multiple points of attack to force players to do something more than spam spells and launch Heavy Snowballs. 

Escort Maps

Escort maps are those maps whose primary objective revolves around protecting a non-player-controlled unit as it moves from one location to another. 

Escort Maps: The Treasure Fleet

The Treasure Fleet is the purest form of escort map in the sense that the player has both a direct control of the area around the target being protected, unlike Convoy, and that the target moves forward before the player could have reasonably cleared the route of enemies in any permanent sense. King of the Giants is technically an escort map as well, but it does not feel like one for the vast majority of the map. This is because Rogan arrives far after the player begins conquering the map, so it is less escorting and more clearing a path, and because enemies do not threaten Rogan on the way if the player clears out the pre-existing enemies. These same conditions are not met on Treasure Fleet, and the player is consequently tasked with defending a wagon from the Treasure Fleet as it travels through enemy territory under constant assault. While The Treasure Fleet has a number of issues (which will be discussed later), such as monotonous gameplay, low replayability, and a fixed timer, these issues are largely unrelated to the spawn mechanics of the map. 

Both the spawn and camp designs of Treasure Fleet are well-done. As the Treasure Wagons move forward, the path itself is relatively unimpeded by walls or permanent enemy encampments, the exception being the Twilight encampment just before the target location which the player can then conquer for himself. Instead, incoming attack waves come from a mix of permanent caves and destroyable spawn buildings built adjacent to the wagon's pathway. The player can choose to only act as an escort for the wagon, or to invest resources into destroying the spawn buildings on the side, making future wagons safer but reducing the resources which can be used to defend the wagon currently en route. At the same time, there are 3 permanent spawn caves, ensuring that players cannot completely eradicate the enemy's capacity to attack. The area around these caves is open, allowing the player space to build defensives without impeding the forward progress of the wagons. There are also a number of optional areas into which the player can expand for additional resources. These side objectives are largely safe from attack, keeping the player's focus on the wagon travel route. The continual focus of the enemy attack waves on the wagon path, along with existence of some of the player's resources on the path itself, is an example of how map and spawn design go hand in glove. The player's own resources are only attacked where it makes sense for an enemy obsessed only with the Treasure Wagons to be attacking. 

Despite a well designed layout and well scripted spawns, The Treasure Fleet suffers from several balance problems and core design issues. I think these are worth pointing out because it is important to understand how multiple small issues can drag down an otherwise well designed map. The first major issue is that the map has a fixed time limit which cannot be meaningfully changed outside of early game exploits via killing one of the wagons. While this is fine for a few playthroughs, it severely damages the map's replayability, which is the second major issue. A good player quickly learns how to exploit a map's weaknesses to achieve objectives in creative manners and this accumulated map knowledge is typically rewarded with an easier playthrough, or better, a faster completion time. What happens with veterans in The Treasure Fleet, assuming it is played normally without the early wagon kill, is that veterans quickly and efficiently destroy enemy spawn buildings, grab early wells for fast power generation, and neutralize the spawn caves whose attack waves are easily destroyed due to not being guarded by map terrain or reinforcements. By the time the last 1-2 wagons spawn, the player has already beaten the map and merely needs to repair the odd tower while they scroll social media waiting for the wagon to meander its way to the finish line.

In other maps, the player's high level of proficiency would lead to a faster conclusion to the map itself, providing a meaningful way for a player to measure their own improvement and to show off. While you could post a replay showing you cleared the map of enemies faster than other people, the fact is that this is not an easily quantifiable measure of skill because it does not show mastery of the map as such, in the same way that faster map completion times do inherently. Also, no one is going to watch the replay.  This issue would have been fixable through mechanics such as the ability to reroute a wagon via a faster but more difficult route, or the ability to release multiple wagons at once but by doing so, the player triggers much harder attack waves as a result. Such mechanics would have required much more complex map scripts to achieve and given the time necessary to achieve this, and the fact that the original devs, based on the abilities of the Twilight Edition cards, seemed to believe PvP would be the more popular mode, it is not that surprising The Treasure Fleet was not given the love necessary to generate long-term replayability. 

Conquest Maps

Conquest maps are those maps whose primary objective is to clear the map of enemies, potentially culminating in a boss fight which triggers the map's end. Lodgement maps, as discussed in the companion camp design guide, are a type of conquest map.  

Conquest Maps: Ocean

The majority of gameplay in Ocean is fighting on different islands made up of different combinations of center and line-based camps. These sections of the map are largely uninteresting for our purposes here. Besides island hopping, Ocean has an additional mechanic, which is that Raven Battleships spawn in from 8 different locations around the edge of the map, and if these Battleships succeed in assembling a fleet of 5 ships, the player loses. This mechanic is a good example of creative spawn scripting, which is well integrated into the map and its lore, and which adds a dynamism to the map that helps make each playthrough a bit different. I would also note that I personally think semi-randomized spawn events of this kind make for a better map experience because they make the map feel more alive. In general, I think Ocean does a very good job of making the map feel like a real location and not just a videogame map to play in. While some islands feel silly, such as the tiny ones populated by ground troops, others like the Skyrake island do a good job of both sticking to the standard enemy faction of Bandits, while also feeling like a place where Skyrakes might actually populate and breed. This feeling of life is helped by a large number of one-time map events, such as the bird attack at the beginning, the mines which spawn on the western isle, and the ability to earn Skyrakes for yourself when you conquer their island. This is a good reminder that one-time spawn events play an essential part in making the level feel alive and lived-in. The existence of the Deathray summoner, which seems a bit random when reflected upon, but which is an iconic part of the map, is another reminder that sometimes it is okay to add a twist to the map. 

After the player reaches the main island, a second Raven ship mechanic activates which spawns Raven ships along the entire edge of the map hellbent on killing the trapped white juggernaut. This is an example of a bad spawn mechanic. The massive invasion comes out of nowhere, is never foreshadowed, and it is poorly executed. Most of the islands and base locations which get attacked by the incoming ships lack the ability to respond to attacks due to the twin factors of little space to build and the fact that Raven Battleships have a 50m range and can fire from fog of war. While this is functionally the boss battle of the map and therefore it is not necessarily wrong that the player's bases are put under real pressure, the issue is that the map itself is not built to facilitate such a fight. In the first part of the map the arrival of each ship is foreshadowed and the player can prepare by building towers and moving units, but the same is in no way true here. The confusing part is that this omnidirectional attack is then followed up by a single direction attack from Blight which seems much less scary in comparison. The final Raven fleet fight would likely be much better if the fleet came from 1-2 directions, instead of all directions, but that the chosen locations were still semi-randomized so that the player could not preemptively defend against it. I personally would have made the existence of the incoming fleet be foreshadowed similar to the singular Raven ships, while giving the player a 1 minute timer to rapidly prepare some defenses in the direction which got chosen. I then would have made Blight appear with ships from all directions to provide a better sense of hopelessness and impending doom, thereby making sense of Mo giving into QueekQueek's blackmail due to a real worry for his life. 

Conquest Maps: Mo

Mo is simultaneously a well designed and a poorly designed map, which is why it alone has been referenced in both the spawn and camp design guides. The general idea of the map, the player getting to control a boss-like NPC unit which he needs to safely guide to the end, is simply a great map concept. The power fantasy of playing as Mo and singlehandedly smashing through Bandit defenses, helps to give the player a real sense of why Mo is treated as such a threat in universe. In terms of sections, Mo is split into three distinct parts. Each of these three parts have substantially different spawn patterns and each have a lot to teach us about good versus bad spawn design.

In the first section, Mo smashes through half-camp after half-camp until reaching the player's T3. While passing through this section, Mo activates 3 permanent spawn caves which continue to vomit flying units for the entire rest of the map. These hostile fliers can kill every well and orb from the player's starter base to his T3. After reaching the player's T3, Mo turns north to fight a mini-boss in the form of Banzai Lord. Banzai Lord spawns Banzai Birds, which Mo can either kill or play hide-and-seek with on his journey northward. The layout of this section facilitates both playstyles, giving the player several safe zones where they can stop and heal their ailing juggernaut. Once the Banzai Lord is dead, the Banzai Birds stop spawning and this section remains permanently safe. The third section, which was discussed in detail in the camp design guide, is then mostly self-contained and functions as a lodgement map. The lodgement section makes use of conditionally permanent spawn waves. Enemies spawn from caves, the spawners are therefore indestructible, but once the condition is met (death of Raven Command Walker or Bandit morale reaching 0) the spawns permanently cease. If the spawn locations themselves were moved and the waves spawned less frequently, to allow players to avoid and hide from patrols while otherwise slowly gaining ground, the third section would be very well designed overall. While the third section's spawns feel less organic than the second section's, the concept of Bandit morale is very sound and lacking only in the fact it is hampered by the issues mentioned above. Speaking of the second section, little needs to be said. The whole area is done so well it is a clear example for how Skylords Reborn map designers should design such areas in their own maps. 

The real issue with Mo's spawn design is the first section. The spawns are clearly intended to place continued pressure on the player as a kind of failure condition. This was likely done because Mo can simply be held back and he is therefore never in real danger of death. If the player were to do that, the player would never be able to fail the map. In light of this, the original devs added the need for the player to continually defend his T1-T3. The issue is just how ham-fisted the whole thing feels in terms of implementation. Does it really make sense that there would be an infinite spawn of only Skyrakes and Windhunters? In terms of map art, it looks like the player disturbed some bird nests and it therefore makes sense that some flying units would spawn in response, but why do they eventually turn into Tortuguns? Overall, the whole bird spawn mechanic just feels like a band-aid to a mistake made in map design which likely could have been fixed in another way. If you look at the picture included above, you will notice that the first and second sections of the map have the same mountain + green field look. Given their connection in aesthetics, it would have been better to connect them in terms of spawn mechanics as well. If I were to remake the map, I would have tied the spawns in the first section to killing the Banzai Lord in the second section. This gives a reason for why the birds were attacking to begin with, because they belonged to the Banzai Lord. The fact that the player would no longer be able to lose permanently after that point is fine, there are plenty of other maps where this is true, and technically they can still lose by letting Mo die which is likely the more common failure condition anyway. The key lesson to takeaway here is that true permanent unconditional spawns should be very rare in maps; it is almost always better to implement conditional permanent spawns when you feel like you need to pressure the player in this way.

[WindHunter]

Mapmaking Fundamentals - Spawn Design (Cont.)

Practical Considerations

In the course of testing and balancing new maps, a number of important questions have arisen regarding practical considerations which must be taken into account. Here we will examine the various practical concerns which have appeared during the development process regarding both spawn and camp design. 

Small vs. Large Spawn Buildings

Each faction should have a small and a large spawn building. As a standard practice, the small spawner should have 1200 life points and the large spawner 3200 life points. There are a few exceptions to this rule, such as the Fire and Nature gateways in Titans, which have 2000 life points, and Lost Temple, which only has 1000 life points. In the case of Lost Souls, a large spawn building is currently in the process of being developed, after which the small spawner Lost Temple will likely see its health pool standardized. 

There are a few ways to utilize the different spawn buildings. The most straightforward method is to use small spawners for T1 and T2 camps, and large spawners for T3 and T4 camps. This creates a sense of progression and ensures that later camps are not easily cheesed through Eruption spam. On this topic, it is important to position primary spawn buildings within camps such that the player cannot easily destroy them without interacting with the camp itself. Spawn buildings, particularly the primary ones, should be the most protected things within standard camp arrangements. Another way to use the spawners is to differentiate them by purpose, such as objective spawners being of one type while standard camp spawners are of another type, though it is also common to use a spawner as a camp and an objective spawner simultaneously. Some maps alternative between small and large spawners as a way to differentiate between the type of units being spawned. A good example of this would be Encounters with Twilight. The various lodgement areas of Encounters are punctuated by small spawners which only spawn weaker units, while the main spawner is a large spawner which generates the main threat in each respective area.

Beyond lodgement areas where the aforementioned type of design is common, some individual camps include both types of spawners with a similar goal, small spawners for weaker units and large spawners for stronger units. Higher tier camps with increased levels of complexity can create a sense of strategic depth and player progression by having a large spawner nestled in the back, with various minor objectives, such as small spawners and artillery buildings, placed in easier to access locations for the player to focus on and destroy. This allows the player to move forward and destroy an objective, thereby permanently weakening the camp's defenses, without requiring that he destroy the camp outright all at once. An example of this might be that in a Twilight map, the primary large spawner generates Abominations and Evil Eyes, while a small spawner closer to the player's expected point of entry generates Whisperers and Mindbenders. While the majority of damage comes from the large spawn units, the destruction of the small spawn would be a major win for the player because it would substantially reduce the camp's CC capacities. 

Time to Respawn & Player Downtime

Imagine playing a Twilight map and moving forward into a well fortified camp. You destroy the frontline of Vilebloods, move forward to wipe out the archer line, and then just as you are about to kill the spawner a second wave of Vilebloods spawn and destroy your army. By the time you return, the entire camp has respawned and because none of it was towers or support structures, you essentially achieved nothing. What I just described is how a normal player experiences trying to destroy the Shadow camp on Nightmare Shard. In the face of this frustration, the player often chooses to learn how to cheese the map, or they decide to suicide their army for the spawn building from the beginning so they can feel like they actually achieved something. What this example illustrates is the problem of camp respawns and how they relate to camp design. 

A typical spawner in BattleForge has a time to respawn of 15 seconds. This is universal and encompasses nearly all maps and spawners in the game. It should be fairly obvious that the lack of granularity in what is an essential component of map design is a fundamental issue for the balance of both individual camps and maps as a whole. 

Time to Respawn Considerations:
1. Individual unit strength - If the individual units are weaker, the camp might be balanced around more frequent respawns. This can create a pleasure experience of fighting through waves of the enemy. If the camp's units are relatively strong compared to the player, such as our example of the player fighting Vileblood's with T1, the respawn timer should be longer because each unit that respawns is a substantially bigger threat.
2. Unit to building ratio - If the camp is entirely made up of units, thereby providing no ability for the player to permanently degrade its strength except through killing the spawner, it might be appropriate to exclude some key units from respawning at all or to increase the overall time to respawn. If a camp's strength is mostly concentrated in buildings, a faster respawn timer for units can be used because only a small percentage of the camp's total strength will be reviving. 
3. Player Tier - In lower tiers, it is harder for players to reach and destroy spawn buildings due to a lack of available tools. In higher tiers, players have numerous options for disabling units and destroying key buildings, including spawners. This suggests that longer respawn timers are better for lower tier camps, because in these camps players usually have to fight through a significant portion of the camp to be able to begin damaging the spawn building. The same is not true for higher tier camps and therefore the respawn timers can be shorter. In general, the higher the tier, the more complex a camp can be in its design, and faster respawn timers are a component of camp complexity. 
4. Distance to Spawner - In a simplified form: time to respawn + travel distance from spawn = player downtime. The farther the distance to the spawner, the longer the player has to recuperate. This is particularly important for defensive scenarios to allow time for healing, respawning, and repairing player defenses. Closer spawn locations will often correlate to longer respawn timers, and vice versa, but the map designer should really balance downtimes based on wall and building repairs and adjust according to the intended feeling of pressure. 

Group vs. Individual Spawning

The general rule is that attack waves should be spawned as a group while units within a camp should spawn individually. By placing attack waves into groups, it means that the player does not have to deal with a constant trickle of units which either preclude any repairs because of a lack of downtime or fail to cause sufficient pressure due to lacking the critical mass needed to challenge player defenses. On the other hand, group spawning often leads to clever players trapping a few units to prevent the group as a whole from respawning. This will be discussed more below. Individual spawns in camps are necessary to allow the camp the dynamism to respond to player attacks. Group respawning, the timer for which only begins after the death of the last member of the group, would mean that the camp will likely never respawn any units before the player can destroy the spawner, but if the group can respawn, the entire camp, or at least a major sub-group, would respawn all at once. Neither of these options are desirable, so camps should use individual spawning.  

Preventing Spawn Trapping

The majority of respawns in the game are based on timers that only begin after the unit which will be replaced has died. The standard timer throughout the game is 15 seconds. As mentioned above, some spawns are group and others are individual, with in-camp respawning typically being individual-based and attack wave respawning typically being group-based. Given that group-based spawns will not respawn until the entire group (or a particular percentage of the group) is dead, this leads to a situation where the player can "trap" units within the attack wave and prevent further respawns from occurring, functionally turning off the defensive aspect of the map. While this mechanic has been normalized on most existing maps and therefore will likely remain unchanged, the same need not be true for future maps. The best way to get around spawn trapping is to make the respawning of attack waves multi-conditioned. For example, if an attack wave takes 30 seconds to reach its destination and lives on average 30 seconds once it has reached its intended location, the average respawn time for that given wave would be once every 75 seconds (15 second respawn + 30 sec travel time + 30 sec fighting to death time). This means, when accounting for slow decks and the player being overrun, it might be appropriate to make the attack respawn on death of the group OR if 120 seconds have passed, whichever is shorter. That way if the wave gets trapped, a second attack wave still spawns after 120 seconds regardless.

Another key point to consider with spawn trapping is abuse of waypoint markers. When attack waves progress towards their target location, they act by moving from one waypoint to another, wait for the entire group to arrive, and then proceed to a third waypoint. If a player is able to block one member of the group from successfully progressing to the next waypoint, the entire group will fail to progress. This is what allows the MotK spawn trap trick on Nightmare's End. When a ranged unit enters the spellbane aura, it immediately retreats in an attempt to move far enough away to attack. If the unit's attack range is less than or equal to the spellbane aura radius, it ends up in a loop where it continually moves in and out of the aura. If the waypoint location is within the spellbane aura, such that the other units can progress to it, but not so close that they will aggro on the source of the aura, the looping ranged units will fail to reach the waypoint marker and therefore lead to the entire group remaining permanently stuck (it should be noted that not all ranged units act like this. The flying units on Mo continue to patrol back and forth despite any spellbane aura, so their scripting ought to be studied to learn how to achieve a similar result). There are a few ways to avoid this. One is to include at least one long range unit within the attack wave, which can then destroy the source of the spellbane. A second, if the spawn trap is discovered pre-release, is to add a patrol along the path where the spellbane will be placed to destroy it and free the normal attack wave. A third is to look at maps like Mo and figure out how to circumvent the issue and allow the units to patrol regardless. 

Tier Emphasis & Camp Design by Tier

Nearly all campaign maps will take place over all four tiers, but the emphasis of each map is different. Some maps will have long T1 sections, while others, such as Bad Harvest, will skip T1 entirely. While it might initially seem best to spread the map out equally between all tiers, this is often less ideal than it might initially seem. One of the largest limitations in BattleForge is the 20 card deck limit. By requiring the player to be able to respond to threats equally at all tiers, the designer encourages the player to opt for more generalist deck options. For example, if the map required the player to be able to respond to ranged, melee, and flying units at T3, the player would forced into using cards which can respond to all three threats. If that player were playing Fire, they will likely default to using Magma Hurler with Unity. If he wanted void return and had to defend a wall, he is also going to include Shrine of War and Tower of Flames. At that point, since the player needs to be flexible on all four tiers, he will only have 1 slot left for T3, which he might want to flex into other tiers given Magma Hurler can take care of most threats. If, instead of designing the map to have equal threats at all tiers. the designer instead decided to shorten the T2 and focus on a larger T3, the player will now have more deck slots to specialize versus any challenges the map might include. If there were large amounts of buildings in T3, the player might be able to include Virtuoso or Juggernaut as a response. If there were no flying units, the player could opt for Vulcan or Giant Slayer. If the camps had weak AA or if there were a lot of hard to reach flying units, the player could choose Spitfire. On their own, each of these options is unlikely to be the sole choice of unit for a Fire player, but each can easily become supplemental options when the camps a player faces in a particular tier allow for such specialization and the other tiers do not use up all available deck slots. 

The key takeaways here are that it is often better to focus on 1 or 2 tiers in a given map, so as to allow for more interesting deck building options within those tiers, and that unifying camp and spawn design across a tier can allow for and push players towards less typical options, for example, not including flying units but giving camps strong AA, can open up space for melee units to succeed.

Defender's Advantage

When fighting against camps, the general principle is to balance the camp's composition as equivalent to being 1 tier higher than the player (on expert, advanced can be equal strength). The reason for this is that players have an adaptability that NPCs do not. This is seen both in the fact that players can cast spells while NPCs must rely on unit abilities or inflexible events via map scripts, and that NPCs cannot rebuild their own camps. If the artillery piece keeping the camp together dies, it cannot come back. In contrast, if a player loses a unit or misplaces a tower, they can just summon a new one. So if the entities in an NPC camp are typically one tier stronger than those currently available to the player, what about when it is the player defending and the NPCs attacking? It would be nice if we could provide a simple rule here stating that attacking units are always one or two tiers stronger than the defending player, but it is not that easy. Defending scenarios are simply too varied for a one-size-fits-all approach. Instead, let us consider some factors that might change the player's strength while defending.

Defense Considerations:
1. Location Fortification Potential - Whether or not a player can easily fortify their location substantially changes the relative strength of incoming attack waves. Fortification potential is related to two key factors, (1) Are they buildable walls, and (2) how wide is the enemy's angle of attack? Buildable walls increase the relative defensibility of a location more than any other single factor. Walls add effective life points to the defending towers and units, while also impeding enemy progress, therefore providing the player with more strength depth. Attack waves which might overrun a player in an open field can become a joke to him when he is sitting behind a set of Amii walls. The width of an enemy's attack angle attacks in a similar way. If, similar to Guns of Lyr, enemies attack into a narrow chokepoint, even giant waves become trivial to defend against. At the same time, if the enemies can spread out such that area of effect spells and abilities cannot hit large portions of the attackers at once, the wave's overall threat increases. Therefore area with narrower angles of attack and buildable walls need stronger attack waves, while areas with wider angles of attack and no buildable walls need weaker attack waves.   
2. Length of Defense - The longer a player is allowed to sit in a single location, the larger the attack waves will need to be to dislodge him. While small, low tier waves might threaten a player initially, even waves one or two tiers higher than him will seem insignificant if he has been given a large enough time to prepare. In general, if a player has only been given a short time to begin preparing his defenses, attack waves could be sufficiently strong even at tier and unit count parity. But once he has been allowed to settle in and build a layered defense, attack waves will likely need to be both stronger and more complex to overrun him.
3. Player Tier - Lower tier players have less defensive options than higher tier players. The higher the player's tier, the stronger the attack waves will need to be in comparison to the players on paper strength.  
4. Total Areas to Defend - A key point discussed at length in this guide has been that the fewer areas a player needs to defend simultaneously, the easier it will be for him to succeed. While even the equivalent of a Tier 6 attack wave from a single direction might fail to overwhelm a player with T4 protected by a wall, even T3 units from enough different directions could overrun the same player. Creating multiple areas where a player must defend taxes not only his power pool, his charges, and his cooldowns, it taxes his mind. It becomes exponentially more difficult to respond to major threats that could undermine your defenses, when these threats are coming from different directions, and particularly if they are far enough apart to not be visible on the same screen. Particularly at higher tiers, where the tools available to players are so strong, it is essential to create multiple avenues of attack. At the same time, a map designer must be careful not to overwhelm the player's mental capacity with an excess of attack directions simultaneously.

[Chimaka( ͡° ͜ʖ ͡°)]

Finally... A fine read for hours. Thank you a lot for over analysing this whole topic and chewing it for the community. Will gladly take time to read it, even if i wouldn't use it particularly.