Game Design - Creating a game experience

[Gilberreke]

Okay, so I'm writing notes on game-design. It's still a bit rambly, dense and lacking in places, but I'd like to get some feedback. Especially games where you see some of these principles at work, things you don't understand and I should clarify or stuff that just seems random and convoluted. I'll create a second post, showing some examples from Minecraft and BTW to better illustrate these ideas.

Creating a Game Experience

Creating an experience is all about sculpting it with good design sense. You want to offer the player tightly designed gameplay, but also allow for lots of meaningful choices, we call this agency. Most of the examples here are geared towards open world games, with a survival element, but most of it is applicable to other games too.

Expansive Gameplay

Expansive gameplay is wide gameplay. It's all about creating varied gameplay through varied circumstances. There should be a wide range of mechanics to interest the player. We shouldn't force the player to do all of them, rather, it's about giving them the choice. We can accomplish this by dividing the gameplay into three layers.

Downtime

The downtime phase exists because the excitement can't always be high, or it is meaningless. This layer is used as a generator of texture, that provides chaos to the other layers, to make them more varied. We still have to make it interesting and we do this by providing the player with activities that will allow them the agency to shape the chaos.

The player gathers resources, tends to issues, explores the map and customizes their gameplay. This is also the perfect moment to tease what's coming and hopefully, provide a choice. The player will be more attentive to details at this time, so we can show off the bigger set pieces, such as a faraway mountain or architecture.

It's important that we make sure this layer gets out of the way during moments of excitement. We don't want to yank the player out of combat to quench their thirst. We can model adrenaline to get a gauge of the excitement.

Periodical

This phase is all about creating expectations and predictability, to give rise to strategy, an important form of agency. By creating a cycle of events, the player can predict them and plan accordingly.

The classic example is a sleep-cycle. By forcing the player to rest from time to time, we give him time to plan. Gameplay should be shaped so that certain actions are only possible a few times per day. Several mechanics interlock, so the player has to choose only one, this creates extra agency.

A good game design rule here is: "reward the player for good behavior, don't punish them for bad behavior". We do this by creating a status quo through expectations. Through predictability, the player expects certain things to be available and we never take those away. Instead, we reward him with bonuses. A player that's well-rested receives increases to stats, for example, we don't decrease expected stats for not having slept enough.

A period might have several peaks of excitement, we don't want to require the player to do maintenance after each round of combat, this creates tedium. This also limits the resources further, creating even more agency.

Visceral

Excitement is at its highest in this phase. We want the gameplay to be action-oriented and feedback to be immediate. Visuals play an important role here, in communicating this feedback. The most important form of agency, tactics, comes into play here.

The most common form of excitement is combat of course, but we are not limited to just that. There are other forms of excitement, such as stealth or time pressure. Be careful not to take away agency though. A classic example is quick-time events. While they are a good way to create excitement, they also take away all meaningful choice from the player.

We are also allowed to punish the player for doing something wrong here. If the player gets hit, he is damaged, but if we want to incur longer-term penalties, you should instead transform them to the other phases. This way, all the layers inform each other and the penalties get out of the way of fun.

Across Layers

Let's recap with an example: we take a large hit during combat, it damages our health and causes bleeding damage over time, but only for a few seconds. It also causes a wound which slows health regeneration, but this doesn't matter much during combat.

If we don't treat the wound by the end of the day, it will turn into a long-term injury, which permanently decreases our maximum health.

We can now choose to take some downtime to search for herbs and apply bandages, which will slowly heal the injury. For the next few days though, we have to find a strategy that relies less on health being available.

Scaling Gameplay

Scaling gameplay is deep gameplay. We create a sense of progression and provide goals. The player indicates which part of the game they like and we make these parts more interesting. We do this by making those parts more interesting and downplaying the rest of the game. Progression is best done by making the easy parts easier and the hard parts harder, as well as removing tedium.

The best examples of progression unlock new forms of gameplay, they don't just make the game harder or easier (usually by "making the numbers go up"), they make it different, transform it.

Examples of progression are a crafting tech tree, abilities, and equipment. Most of these are gated, so it's hard to skip past parts or break sequence. We provided plenty of variety with the expansive gameplay, we can funnel the player a bit more here. Especially, since the progression unlocks new, novel gameplay, amply making up for the hand-holding.

Specialization

It's important that we make the parts the player likes more challenging. We have to increase the challenge of activities the player likes, as well as providing them more opportunities to experience these activities. Meanwhile, we also remove the necessity of other experiences.

For example, if the player likes to sneak, we offer him more advanced and challenging stealth gameplay. We can't make him better at sneaking as that makes the interesting part easier, while still allowing situations to happen where we can't use stealth. It's the opposite of what the player wants.

Case Study

I'll provide a few examples here in the future of games/mods and how they treat the different aspects of gameplay.

[Taleric]

Who is the target audience? The concepts are good but as you said very dense.

A reader or listener would need some in depth gaming experiance to unpack things.

For examples you should probably pick the most sold titles in various genres and tease out the features you are trying to describe from there.

[Gilberreke]

Who is the target audience? The concepts are good but as you said very dense.

A reader or listener would need some in depth gaming experiance to unpack things.

For examples you should probably pick the most sold titles in various genres and tease out the features you are trying to describe from there.

I tried unpacking the notes for a wider audience, but that turns it into a large series of articles that I haven't written yet. Current audience intended is other game designers.

I think I'll start by giving some examples of real-life games using these techniques first. See if that helps. Any particular points of interest in there?

[Gilberreke]

Some notes on expansive versus scaling:

So the expansive notes go straight into the layered model but don't go over the basics. The idea is that wide gameplay offers multiple ways to accomplish the same thing. Usually, this massively boosts replayability too. For example, branching storylines are expansive, but not a very good way to do it. Another way is random or procedural generation, but the real holy grail is varying the different paths to take. Most roguelikes suffer from this, they have branching storylines and procedural generation, but playing the game twice doesn't feel that different, regardless of choice. Meanwhile, lots of RPGs do this well. The first time, you play a mage, second time a thief and the game becomes a completely different beast. These games lock you in early (character creation), even better is not locking in at all, of course, allowing the player to choose exactly how he wants to play the game.

An important note at this point: in many cases, the game that tries to do everything is less well-designed as a result. Offering everything I mention here is the holy grail, but really really hard to pull off. A more sensible design is usually to focus on parts. BTW doesn't offer a lot of expansive game-play because it increases complexity in a procedural game exponentially. Too much complexity and you lose all control over the design.

The difference with scaling gameplay is that it is deep instead of wide. Instead of options, it focuses on linear progression. Making sure there's always that next thing to learn. You can combine both, but again, design tends to suffer if you try to do too many things. The smartest move is usually to do a few simple things that lead to emergent game-play. Chess is a great example of this, being both very expansive (lots of choices) and very scaling (early decisions carry through until end-game), only with a few simple rules and emergent game-play.