This is a more general article about the "flavor" of game design we're going for, instead of a specific game design.

Each of these items is along a spectrum, indicated by the "vs". I am going to reveal my bias here and indicate that I typically favor (and would like to follow for our game) mostly the 1st of each. That said:

• The whole point of a "spectrum" is that it is not binary, sometimes 75% one way or the other is optimal, and 0% or 100% rarely is.
• Absolutely amazing games have been made at the opposite end of the spectrum for each of these, it depends on the kind of game you're making.

I've ordered these such that the ones whose choice has the most impact on a game come first.

Empowered Player VS Restricted Player

This starts with the most basic choices of a game, the controls, and the possible player actions, and how they affect difficulty, and the order of game design.

If you start by first making the player as agile as possible (fast, instant movement), making them feel in control, giving them lots of options on how to operate in the game world (many diverse actions, including building/destruction, allowing creativity), you then as a consequence have to make the actual game harder (more/tougher/smarter/faster enemies etc) to compensate. Or, you can have an idea for a game world, its enemies and its resources, and then only design the players abilities later, and find out you have to restrict the player a lot, to not make it possible for the player to easily kill or bypass enemies, bypass intended story progress, or get desirable resources too early, such as reducing their speed, slowing them down for particular actions, artificially restricting jumping and movement not to make parts of the level accessible etc.

Both these approaches can result in a game with the same difficulty level, but in the former, you feel great while executing actions, and you tend to feel more "badass" since it appears you are playing a harder game than it actually is. In the latter, these limitations often feel frustrating or artificial, and your sense of achievement is lower.

There may be reasons why maximum player empowerment is not desirable:

• Some limitations may be necessary for a survival game, where being limited along some axis is part of the point. But there, a conscious choice is made that certain limitations are fun and desirable, whereas the limitations mentioned earlier are often necessary not to break the game, but never fun in their own right.
• An empowered player generally makes for a faster, more intense game, which is not always appropriate either.
• It is also hard on content. If you've decided to empower the player, and it is discovered that some player shortcut is possible, then your only option is to rearrange things in the world for that not to be game breaking, as opposed to put a convenient player restriction in place.
• In some sense it requires the rest of the game design to be more robust/resilient, and not rely on a carefully crafted progression to be fun. It thus makes for harder quality control, and harder to ensure the game is always fun.

But it is also very appropriate for open world, creative, user driven games.

Some tradeoffs that are in some way a specific aspect of the empowered player:

Physics VS Animation

A big split in games is in whether they make the player wait for an animation to finish before being able to take a new action, or whether the player affects physics directly, and the animation tries is best to express what the physics is doing.

A player will feel more empowered if the game is physics driven, but it has the huge downside that animation often looks clumsy/unpolished on sudden movement as the animation system desperately tries to blend towards the new state of things. An animation driven game has a clean 1:1 relationship between action and animation which tends to look very neat, but can be frustrating to control.

Since at least your own animation is not visible in 1st person games, these tend to be less likely to be animation driven.

Immersion VS Art

This is the big tradeoff that drives 1st vs 3rd person games. It is interesting to consider the huge advantage 3rd person gaming gives:

• There is a much more consistent scale to the art in the world, which are seen at "medium" distance more often, allowing for art to look great at a given level of detail, compared to 1st person which has to deal with extreme scale difference of art up close (often ugly, distorted) and very faraway (camera always looking close to the horizon)
• Showing player in the world makes it easier to express what happens to the player in terms of gameplay (they can see the effects on their character).
• Easier for the player to interpret how they fit in the physical reality of the world (they can see why they are colliding with something, easier precision platform jumping, etc).
• Much easier to create flashy looking fighting effects, melee looks great. Use the whole body for animation, as opposed to the unnatural nature of 1st person hand/weapon movements.
• Easier to fight when being surrounded is a thing.
• Less scary / confusing for younger players.

A lot of these are art driven though, and it would be very interesting to see how many games in the past became 3rd person simply because the art director insisted on wanted to see prettier fights.

Of course, 1st person has a lot of disadvantages, but its one advantage, immersion, is huge. Modern engines have learned a lot of tricks to mitigate some of the above disadvantages. Melee vs projectile focus also makes a big difference, as 1st person guns can feel incredibly satisfying.

On the graphics side, it is not surprising that games in history known for pushing the graphics envelope have often been 1st person games. This is simply by necessity: if you want to make a better looking 1st person game, you need to push graphics tech harder than any other genre.

Persistent World VS Respawning World

This one has so much impact because it has a bit of a snowball effect on everything else in the game: when you kill an enemy, or collect items/resources from the game world, is that change forever, or are these things gradually respawned/refreshed/regrown over time?

Through game history, the use of this has been shifting:

• In really old games, respawn was common simply because the world was divided into small levels that were independently loaded and it was technically not possible to save the state of all levels, or maybe it was too complex in terms of code/design.
• Shooters in the 90ies and many other game genres made it possible to have a more persistent experience.
• MMOs reversed the trend back to respawning, which at least they had a reason for, since it needed to work with arbitrary amount of players in arbitrary areas.
• Early Open world games made persistence again desirable and possible.
• More recently, I see a big uptick in respawn based games, in trying to make games with mmo/social/mobile flavors, more casual, more grindable etc.

There are many reasons why respawning reduces fun:

• Choosing respawning (or not) has snowballing effects on all other gameplay design, because once you have respawning, player can now grind enemies indefinitely, so you need to make the rewards from enemies low (or at least diminishing returns) to not make the game broken. It makes things more of a slow "numbers" game instead of "wow, look at this upgrade I found, I am now way more powerful" kind of instant boosts that can be very exciting.
• Respawning reduces the reward from "completionist" tendencies people have since there is no point in trying to kill all enemies or collect all the items. How much do you collect? As much as you have time for? As much until you are suitably leveled up? It starts feeling arbitrary.
• It can make difficult fights against enemies seem more annoying, knowing that they'll be back in 10 minutes. It feels pointless, and the focus is not on overcoming the challenge, but on the small rewards you'll get for putting up with the fight. In a persistent game a fight can be very difficult but just merely beating them (even if there is no "drop") can feel very satisfying since you never have to face this same fight again.
• It takes away a lot or excitement of the game. It shifts the fun of the game from the core actions the player performs to the numbers after the actions. Intrinsic vs Extrinsic reward, etc.

Of course, there are strong reasons games use respawning, the primary one being that balancing the game is endlessly easier with respawning, you don't have to worry about there being enough resources, or even enough content for the player, just make sure to cap the benefit. With respawning, you can throw in all this content and features and it can easily be ensured it won't cause overpowered on unwinnable situations.

It is not fully binary, there are some ways to have limited respawning that keep the advantages of a persistent world:

• Level designer creates explicit events / story moments when a specific invasion of monsters teleports in, but only once.
• A monster spawner object you can destroy. This can be further reduced by the spawning only happening when you're near, and/or a limited amount. Whether a monster spawner is a nuisance or a farming opportunity depends on what these monsters "drop".
• A more general respawning of any monster in the world, but super slow/sparse, i.e. if you've been playing for over an hour you may find a few new monsters in areas you had cleared, but never enough to a) feel annoyed your cleared area is not cleared and b) do not provide much of a farming/grinding opportunity.

A sub topic of persistence is of course player death:

Consequences VS Load/Save Progress

There are many possible ways to deal with player death, but the two big ones are:

• You load a prior savegame. You replay a part of the game exactly as before.
• The world state largely continues upon death, you respawn away from your death, and have to travel back to the action and/or risk losing some resources, but always in a recoverable way.

The problem with savegames is that they make you focus on the meta. Should I save now? When did I last save? Argh I died and forgot to save for a while. Argh I took a ton of damage, if I save now I may not be able to make progress later, but if I don't I will have to replay more? Should I save in a different slot (if you have one), etc etc. Automatic check points only partially fix this as there is still this "when is the next save point" anxiety that occupies your mind.

The world being persistent simplifies your thinking (no choices to make regarding saves) and focuses you on the game. Got into a tricky situation? Oh well, might as well try and make the most of it! Near death, and still made it? Feels extra exciting and accomplished, compared to doing the same thing on 3 savegame retries. You have to get more creative to get out of your current situation. Decisions feel like they carry more weight and are more exciting, more meaningful. You may have to be more careful, rather than pressing "save" and running forward, not caring about traps ahead. Having to work thru your mistakes creates more of a unique, player driven story.

Luckily, with a game intended for seamless coop gaming, the decision here is simple, as load/save with in coop would feel even more intrusive than in SP. On top of that, all these more exciting consequences of not being able to cheat death are extra fun in coop ("haha, Fred fell in the lava"), and are part of "player storytelling".

Philosophically, it is maybe funny that I advocating for enemies to have permadeath, yet the player respawns indefinitely. As opposed to a game with monster respawning that requires reloading on player death, which is inconsistent in the opposite way. You'd think it be better if these two choices were aligned, but we have a natural asymmetry in that we design games where the player must be able to make to the end no matter what.

Mechanics vs Realism

Some games pride themselves on realism, but if you look closely, even the most "realistic" games have long lists of very unrealistic aspects for the benefit of gameplay.

As it turns out, the choice of how to restrict a gameplay mechanic by realistic limits is always a bit arbitrary in any game, so you may as well turn things around by first thinking what mechanic is most fun, and then find the most plausible mapping to realism for it.

Realism not great because actually being a soldier under fire is so much fun (it's actually horrible, I hear). Realism is great because it reduces learning, and increases immersion: by mapping your gameplay mechanic to something players may be familiar with (from reality, movies, or simply other games), they'll know how to use it straight away, what it means, and what to expect and fun increases. Games can have a lot of mechanisms, and without the help of "realism", they'd actually be very hard to understand.

Random Input VS Random Output

Input being things in the game world that come at the player (monster, items, levels), whereas output are things that the player produces (weapon damage / spread etc)

Randomness is generally something that should be used in measured amounts in games as it helps with variety and replayability, but it is more fun if its something the player can respond to, as opposed to something that just "is" and the player has no agency over.

My Fault VS Mysteries

When the player gets hurt or dies, some games make it very obvious what happened, and the player feels its their fault and feels motivated to do better. Other games prefer to not give the player quite as much information (dislike having all these numbers/indicators on the HUD etc), and want the player to learn the game the hard way. A first death of a new thing in the game world will typically feel random/mysterious, but then the player gradually gets better at predicting the problem. This can be frustrating but also instill a sense of mastery (in e.g. Souls-likes).

Are things readable (chess) vs there being lots of hidden information (poker) ?

There's no right or wrong here, different players respond differently to each gameplay style.

Correlation

Many of the above items are correlated, where the first items point to very open, un-balanced, creative, non-linear systems driven games, and the latter to more tightly controlled, balanced, linear, story driven games.

In particular the issue of balance is one where if you want a strict balance, then all your choices have to support that, and conversely once some choice makes balancing hard, you may as well increase the excitement and variety by allowing it fully.

"open world" games make balance especially hard, unless they make things more balanced by going for a respawn based design. In addition to open world, we want to focus on "user generated content", which makes it even harder to ensure balance. You might as well make the chaos that comes from unbalance into a feature.

Specific "Game Pillars"

Here are some more topics that are to some extend more specific instances of what is discussed above:

Player Creativity

We want to encourage the player to get creative (using destruction, building, the "open" and "simulation" nature of the worlds) to make whatever playthrough seems worth it to them, and not worry if they subvert quests or do things entirely differently. Quests are there to give challenge, meaning, and variety, but should never be thought of as absolute gatekeepers.

Player Story above Designer Story.

To be completely player story driven typically requires procedural content without quests, but that doesn't mean we can't emphasize player agency at every step possible, making them feel they are directing things. We may at some point also have more procedural content.

Game mechanics over Realism.

We borrow from realism (or other games) where it serves us to short-cut having to explain how something works, but in the end, the mechanic is king, and its workings are optimized for fun gameplay interactions first and foremost.

Things have consequences.

Typically, a player action should have permanent effects on the game and world.

Feedback and Visceral action

We like the player to really feel what they're doing, especially frequent actions. It simply "feels good" to be in the world and do anything.

Non-linearity and Exploration

Where possible, we like to allow the player to explore in whatever direction, though there may be consequences / difficulty spikes. These consequences can help with the Player Story.

Solidity and Consistency of the world.

We make everything in the game out of the same "stuff", and there are hardly any exceptions on what you can do with it, everything is destructable to the last voxel, given the right tools and determination. This makes everything in the world feel very real. No hidden barriers, no arbitrary distinctions between static vs destructable scenery, etc.

Progression from weak to strong in a single world.

We want the contrast between how the player feels at the start of a world vs the end to be as big as possible, going from a sense of having nothing, having to be really careful not to run into enemies too quickly, and to be happy for quests that might give you items, all the way to feeling badass and charting your own path.

Co-op multiplayer makes all the above better.

Or rather, their design is meant to work well for multiplayer.