Game Design: Combat

The web release is going well with a lot fewer bugs or horrible balance issues than I was anticipating so I've been trying to take it a bit easier this month and recover a little bit before diving into Early Access. There will be a few more updates for the Web-Version mostly fixing up any remaining bugs and tweaking the balance a bit but, surprisingly, I think its actually mostly complete. In the meantime I've been going through and compiling some of my notes on various aspects of the games design and figured I'd share them here for those who are interested.

THE FOUNDATIONAL DESIGN TENSION:
At the core of RFIVs combat is a fundamental design tension between two opposing gameplay styles. On the one hand the game is intended to be highly tactical and reward clever decision making, game knowledge, and careful planning. This side of the design draws heavily on traditional rogue-like design as well as strategy and tactics games in general. Of equal important and opposing this perspective is the idea that the game should be fast, action packed and full of crunchy, satisfying explosions. This side of the design points draws on my love of old school 'Boomer Shooters' with all their high mobility, circle strafing and general mayhem.

The goal of the design is to maximize both of these seemingly opposed design principles and create something that is not just a middle ground between slow, thoughtful play, and rapid, crunchy mayhem. If taken to the extreme we could say that a maximally tactical game would have very long turn lengths, say for example 60s. Taken to the other extreme we could say that a maximally action based game would have very short turn lengths, say for example 1s. The solution to having both is not to strike the middle and have turns of 30s. Rather the idea is to have the vast majority of turns be very fast, rapid fire, simple decisions with quick animations so that the player can just click click click. Every once in a while something happens that suddenly causes the player to stop, take his hands off the controls, think, and then very carefully and slowly play out the next several turns.

By designing the game in such a way that the combat tends to swing back and forth between these two states we can have a game that plays very rapidly like an action game, while still maintaining its tactical focus. Because the 'fast play' consists of many turns of short length and the 'slow play' consists of few turns of long length the total time spent in each 'mode' over the course of a run should end up being roughly equal. So the player may play 60 turns very quickly in 60 seconds and then take another 60 seconds to very carefully play the next 6 turns. This gives us the best of both worlds while also introducing a very nice core game loop that naturally cycles back and forth between both styles of gameplay.

This idea is further expanded to the strategy layer of the game in that, after every couple of cycles of this fast/slow loop the player will pick up some piece of equipment, get a new talent or attribute point, find a Tome-of-Knowledge or else encounter something that requires them to make one or more long term character building decision. In these cases the player may take a very long time to make a decision that actually only takes a single turn in game. So that's how the core game loop is designed. Rather than trying to make every moment of game play be fast, tactical and strategic at the same time we simply loop through a cycle in which each part has a different emphasis.

DEFINING TACTICAL COMPLEXITY:
I think the easiest way to think about tactical complexity is to consider the total number of different actions the player must make over the course of a winning run. These are not as simple as basic actions like movement or attacking but rather include the purpose of the action so we would consider movement to kite away as separate from movement to get into position to attack. The more different actions the player must make, the more tactical the game and ultimately the higher the skill ceiling. As the game becomes more tactical the player must know, understand and effectively use a wider and wider range of actions.

This is the primary reason why I'm so eager to nerf or remove any ability or action that is effective in way to many situation. The more heavily the player can rely on a single action for success the less other actions he needs to use and, by the definition above, I would consider this to be reducing the amount of tactics in the game. This is also why I consider it important to introduce enemies that act as counters to otherwise effective actions since this forces the player to adapt some different play style, take different actions and ultimately increase the tactical scope of the game.

MOBILITY:
- Both the player and the monsters have more movement abilities that allow them to quickly and more drastically change their position on the Game-Board.

- By moving more drastically, the Game-Board can take on many more configurations leading to a wider range of tactics. Without large movement abilities, combat will almost always devolve into a line of enemies coming at the player from one direction and the player kiting backwards to maintain distance and avoid being overrun.

- By moving more quickly, the position of units on the Game-Board can change more rapidly. The situation can change more significantly turn by turn. Ex. a number of melee units can suddenly lunge into melee range, a priority target the player has been flanking can suddenly blink out of range.

- From the players perspective there is simply a much larger tactical decision space since more tiles are suddenly 'in range' for movement at any time. As a characters mobility increases the number of avaiable actions (moves in this case) increases. This is why mobility is so critical to increasing the skill ceiling. Increased mobility doesn't inherently make the player stronger but rather opens up more tactical possibilities, which, only if taken advantage of, result in power. 

AVOIDABLE ATTACKS:
A large part of RFIVs enemy design that is pretty unique is that enemies have a large number of abilities that the player can avoid through simple movement. Things like slow moving projectiles, telegraphed attacks and channeled spells that take effect over several turns are all avoidable by the player by simply moving out of the way. These attacks are generally given very nasty effects or high damage but in a sense the whole point of the design is not what they do when they hit but rather what sort of movement they force the player to make to avoid them.

As mobility grows in important in the design, avoidable attacks become a way to force the player to be mobile, to force him to move into a position he otherwise doesn't want to be in. This creates unpredictability in combat and allows interesting new layouts of the Game-Board. If the player is never 'forced' to move then the number of possible layouts is greatly reduced as the player will always maintain some optimal position. This almost always leads back to a situation with all the enemies on one side and the player on the other, kiting backwards. By introducing this 'forced' movement we can end up with many more interesting situations.

PRIORITY TARGETS:
In RFIV I'm really trying to push the idea that every enemy type is unique and has its own strengths and weaknesses. Since the default fight involves the player against 2 or more enemies there is almost always going to be an enemy that is the highest priority target. This may be because he is the most dangerous to the player or it may be a support type enemy that is buffing or healing his allies. In as many ways as possible I am trying to intentionally introduce enemies that when encountered in a group become the Priority-Target and must be eliminated as soon as possible.

By introducing these Priority-Targets we get a lot more varried tactics in the combat since, based on the enemy group composition, layout on the Game-Board, and abilities available to the player it can take quite a bit of work to actually kill the target. Rather than simply blasting away at whichever enemy is closest the player often needs to plan a few turns ahead, ignore the nearest enemies and somehow maneuver himself into position to kill the Priority-Target who is often hiding in the back.

MORE FREQUENT BUT WEAKER ABILITIES:
This is a big change in the design from RFIII. I've been gradually making abilities weaker while reducing the cooldowns or increasing the rate of resource regeneration so that they can be used more frequently. Note for the purpose of this discussion I consider Speed Point Sprinting to essentially be an ability. I believe this model has several advantages:

- First the player simply does more during a fight. There are more decisions to make and fights generally feel more dynamic and engaging. There is almost never a turn in which the player is not either using an ability or moving into position to use an ability in the next turn or two.

- Since each ability is weaker and is unlikely to finish a fight all by itself, its much more likely that the player will use a wider variety of abilities during a fight. This increases the tactical complexity of the combat as the player is using a much wider range of abilities rather than just 1-shotting things.

- Another major advantage is that weaker abilities reduce the chance that a fight ends right at the start before it has a chance to evolve into a more interesting tactical situation. When a fight first starts we can assume the player is generally fully regenerated and so has access to all of his abilities. By weakening these abilities we assure that even if the player goes full burn in the first few turns he is unlikely to finish the fight. By reducing the CDs we ensure that even though the initial burn does not end the fight the player is able to quickly continue fighting as his abilities begin to cycle.

ENEMY GROUP COMPOSITION:
The default fight in RFIV consists of the player fighting a group of enemies. Much more than RFIII I am trying to design enemies in such away that they synergies when in groups or at least are distinct enough that the player needs to consider the overall composition when planning his tactics. Simply adding individual enemies to the game increases the tactical complexity somewhat, but the real power comes when these enemies are designed with group composition in mind so that each new enemy must be treated differently depending on what sort of group he spawns with. By designing enemies to work together we get a sort of exponential growth in complexity. The same enemy added to 5 very different groups should produce 5 very different combat encounters. Some examples of arch-types to maximize this compositional complexity:

• Supports: any monster that heals or buffs his allies. Typically a very high priority target. Buffs are really important as they can  change the way another monster behaves. Most obvious example being a hasted vs non-hasted ogre.
• Nukers: any monster that has the ability to dump a ton of damage into the player. This could be ranged or melee. The player really does not want to get hit by these guys but depending on if the damage can be avoided they may not always be a high priority target. The player does however have to pay very close attention to them during the fight.
• Plinkers: typically ranged enemies that just sort of chip away at your health. They are not very dangerous, are a low priority target, but due to their range they apply a constant pressure on the player to finish the fight quickly.
• Swarmers: weak but come in large groups that can quickly surround the player. They tend to get in the way a lot and are generally just annoying until they manage to surround you. Some for of AoE is very good for dealing with them.
• Summoners: often some of the most dangerous and high priority targets since they fill the battle field with monsters which tend to act as a shield and make it difficult to kill the summoner. 
• Tanks: are interesting in that they tend to have high damage but low mobility and so are best to be ignored until the end of the fight. Their high HP makes it impractical to burst them down mid fight and so typically the player relies on kiting them while clearing out their allies after which it tends to be pretty simply to deal with the tank alone.

HARD COUNTERS:
Hard counters are something that I'm trying to move away from in RFIII though they will always have some place in the design. By hard counter I generally mean an enemy that counters some player ability purely through its statistics. Reflection, elemental resistance, protection or strict immunities all fall into this catagory. While this certainly works to counter abilities and force the player to use something else, the trouble arrises from the relative short length of the game and the linearity of the dungeon. There simply isn't enough opportunity for the player to build a character to overcome hard counters many of which he will not even be aware of until he reaches that part of the dungeon. Even when the player does successfully overcome a hard counter its possible that the same enemy will not show up again and so any resources spent are then wasted.

SOFT COUNTERS:
Soft counters are the thing I'd like to lean more heavily into in order to encourage a wider range of tactics from the player. Unlike hard-counters, soft-counters are based more in an enemies behavour and the more outward effects of their abilities rather than the raw underlying stats. Things like fast enemies that swarm you, ranged enemies that stay way out of range, tanky enemies that can soak up a ton of damage, priority targets that must be bursted down and may be hard to reach, enemies that either move themselves or move the player into challenging positions etc. 

All of these things force the player to adopt some new tactic or use an ability in some novel way but unlike hard-counters they don't simply disable some portion of the players kit. Soft-counters tend to have multiple solutions and so while the player still has to build his character to handle lots of situations, the system is much more flexible. I believe it is also more satisfying since the solutions often require some cleverness and tactics rather then just 'bring a source of non fire damage'.