Discrete Depth and Complexity:
- Create depth and complexity from the simple interaction of many simple mechanics rather than lots of numbers and complex formulas.
- Each individual element should be very simple, complexity should arise from the way in which these elements interact.
- When possible, properties and interactions should be describable in pure English without resort to numbers or formulas.
- When numbers are used they should be extreme i.e. half or double.
- Everything is overpowered when used in the correct situation.
- Ex. Shocking an enemy in water deals very high damage.
- Ex. An axe used against swarms of weak enemies will quickly clear them.
- Everything is hard countered when used in the wrong situation.
- Ex. Elemental resistance must be effective enough to really nullify attacks of that element.
- Ex. A reflective enemy is nearly impossible to kill with standard ranged attacks.
- A character is powerful when he can exploit many situations and has solutions to many counters.
- A character should not be strong enough in a single area to overcome counters and allow one-trick play.
- This can be achieved by keeping the focus on mechanical interactions rather than stats that can just keep growing.
Tight Focus: Tactical Dungeon Crawler:
Limit scope to tactical combat, character development, and strategy. Don't get bogged down building an everything simulator.
Tactics:
- I believe the game is already doing a good job of creating tactical combat situations that rely on player knowledge and skill. Continue to expand upon this.
- Success in combat should be determined by the players skill and knowledge more so than raw statistical advantages.
- Continue to add challenges that hard counter single tools and tools that hard counter single challenges.
- The skill comes in knowing and using the right tool for the right job.
- Continue to emphasis movement and re-positioning. Have many more enemies, traps, auras, level features etc. that encourage the player to move.
- There currently many negative reasons to move in combat i.e. to avoid some effect. Add some positive reasons such as beneficial terrain or auras.
Character Development:
- Provide more choices for the player to make regarding character development. Ex. The 3 choice treasure rooms.
- Do not however allow the player complete control. Reacting and adjusting to randomness should always be a key factor.
- Rather than restricting item and talent drops to only be useful to the class being played. Work on making all items and talents useful to all characters.
- Limit and discourage the creation of one-trick characters that solve every challenge with one or two overpowered abilities.
Strategy:
This is an area in which I think the game needs a significant amount of expansion and improvement.
- Building your character in a way that provides unique solutions to the wide variety of challenges posed by the dungeon.
- Adjusting your build and play to the random resources provided by the dungeon.
- Making intelligent use of the limited times in which you can actually choose.
- Navigating the dungeon based on specific zone challenges and your unique character build.
- The order in which zones are tackled
- Knowing when to back off from a zone.
- Knowing when to take a risk and when to avoid,
- A knowledgeable player should be able to have at least some expectation of exact risk and reward even in the face of randomness.
Variety is critical both between and within runs. The length of the game should increase slower than the addition of new, varried content.
Player Character:
- Classes, races, gods should be distinct enough that each character plays differently.
- The player should have to branch out from their starting class/race.
- Two characters with the same starting race/class should not end the game looking identical.
- Focus on the interaction between various elements in a character build to create multiplicative complexity.
- 10 distinct classes and 10 distinct races produce more variety than 20 distinct classes assuming races are orthogonal to classes.
Dungeon Inter-Run Variety:
- The dungeon should be different between separate runs. This is achieved by having more content than game length and maintaining this ratio throughout development.
- More content than game length means that the game can have many OR cases ex. a single run will have The Crypt OR The Iron Fortress. This can then be applied recursively.
- There are 5 unique enemies in a particular zone, only 2 spawn per run.
- There are many special levels and unique generators per zone, only a few will spawn each run.
- The Crypt can have The Lair of the Necromancer OR The Undead Temple as its special level.
- Each of these special levels has a set of unique loot that can spawn at the end but only a subset of that loot spawns in each run.
Dungeon Intra-Run Variety:
- Each zone of the dungeon should be distinct from the others.
- Unique graphics, enemies, items, level generators, special levels etc.
- Zones should have consistent themes both aesthetically and mechanically.
- Again this can then be applied recursively:
- Within a single zone there should be sub-zones i.e. levels that are distinct from the other and follow some sub-theme.
- At the point where two zones connect a special sub-zone may be generated ex. Iron Forge connected the Iron Fortress and Core
Aesthetics and Interface:
- Aesthetics and interface are just as important as the underlying mechanics.
- Continue to iterate on the art style, add new sounds and music to the game, improve the visuals of the interface.
- Focus on consistency of presentation rather than having a mix of different styles and quality.
- The game should be hard because the game is hard, not because the interface is overly complicated.
- Provide tons of options for players to customize their interface and controls.
- Anything that is tedious should be automated. If the then automated tedium is overpowered then it should be nerfed. Don't force the player to choose between optimal play and not driving themselves crazy with tedium.
Experts and Becoming One:
- Design and balance for the expert player while providing support for new players to progress to experts.
- The interface should provide all the 'power options' that experts would expect in order to play quickly and efficiently even if these take learning to get used to.
- The simple mouse interface however should be readily understandable by beginners and provide constant tool tips that introduce the shortcuts.
- Since rogue-likes are designed for repeated plays, the game should be balanced to challenge experienced players.
- The beginner will die (a lot), don't make the game easier simply to suit the beginner who will then quickly find the game to easy as they improve.
- Rather try to make it as clear as possible why a character died and what other actions could have been taken.
- As much information should be presented in game as possible. Assume players will consult a wiki and so design the game to be difficult even with perfect knowledge then provide this perfect knowledge in game.
- Don't make things that are difficult if you don't know the one thing that makes them trivial.