The following are my final comments on depth.
Interplay Isn't Everything
The main reason why I wanted to clarify depth in this article series, specifically how it's something more than just interplay, is because pure interplay isn't everything. I can design a game right now with near infinite back-and-forth interplay counters. Let's call it 1UpperCut. This is a two player game where each player tries to 1up each other by saying a higher number than the other player. To win, just say the higher number followed by "I have now won the game." Before you say "game" your opponent can cut you off and say a higher number. As long as your opponent is saying a higher number you cannot win. This game has lots of back-and-forth interplay, but it's also fairly meaningless.
The point is interplay when in service of gameplay should move the gamestate in a direction. Stalling, static space, and infinite loops can work against depth and the meaning of gameplay by keeping the consequences and player progress from advancing.
You can have games that are very complex with very little interplay (Dominion, Acquire, Guitar Hero, time trials in racing games). You can have non-complex games with small loops of interplay (Rock Paper Scissors). You can have these same style games with more complexities (The Dinosaur King arcade game). There are "games" with some interplay, some complexity, and no player choice (War: the card game). And there are games with many options and interplay that suffer from dominant strategies or broken characters (Street Fighter 2 Akuma, Street Fighter 3 Third Strike). None of these games are very deep (or as deep as they can be without bans in the Street Fighter cases). Now it's clear that gameplay depth applies to a very specific quality.
Deep Gameplay Starts with Risk Reward
There is a drawback to playing deep, complex games. Because the depth makes more complexities relevant for success, deep and complex games create greater learning barriers for beginning players. Another way of looking at this problem is that players have to learn much more before the gameplay "clicks" in their minds; before they have enough basic know-how to understand the gameplay system at its core.
So much of what is great about gameplay -- how it is a rich interactive learning environment, how it reflects players and their mental states, how it squeezes, and how players have agency in the system -- only occurs when players make informed decisions. As obvious as it seems, in open gameplay systems, before learning the basics of gameplay, players are still in a learning stage. In this stage players are unable to make informed decisions about what to do because they cannot evaluate the gameplay scenarios accurately. Keep in mind that some rules are more core to gameplay than others and there are surely many nuances that highly successful players never need to learn. Accurately evaluating gameplay scenarios first requires understanding the risk-reward of options. Deep gameplay of interesting choices and informed decisions only begins when players understand which options are risky and what the return is for success. Knowing the risks involves grasping how the consequences may ripple forward according to the dynamics of the system. With this information, players can start to express their agency in a deep, open gameplay system.
There is no risk for doing an action if players are not committed to the consequences. Moment to moment if the player can simply take a move back, get a do over, or abuse save systems there is almost no gameplay. With no real commitment to actions and their consequences, players merely explore the possibilities of the system. Doing so is not a necessarily gameplay. Risk comes from potential consequences and deep gameplay starts when these consequences cannot be taken back.
Players don't risk anything until they commit. Commit to an animation. Commit to a move. Commit to a turn. Commit to a position. Commit to something. This is why I tend to dislike cancels in game design. Cancels tend to delay the gameplay of a game by allowing players to cancel some of the consequences of their actions. Cancels can stress the core interplay design of a game and raise the learning barrier by increasing the importance of complex, dexterous, nuanced strategies. Basically, if everyone is canceling many of their consequences, then everyone must cancel to reach the new level of "normal" play. Cancels also tend to raise the action frequency of gameplay. For all these reasons, I do not prefer L-canceling in my Smash Brothers or snaking in my Mario Kart.
The Potential Depth for the Video Game Medium
With depth of any kind of work, we're looking for how complexities build more complex ideas and experiences, and we're also looking for resonance. You may recall that resonance is a part of my CED story analysis system. Resonance happens when certain ideas or complexities have elements that are similar in some way to other elements. In stories we typically look for ways that the themes resonate with the story content. Resonance when well balanced gives a work a very cohesive feel. When all the major parts resonate with each other, it communicates a very intentional, clear, hand crafted execution. Because clear complex ideas are very hard to convey (or very easy to muddle), resonance practically screams a sense of intentionality, skill, and craft.
For pure gameplay systems the meaning and the depth comes from how the different complexities are balanced against each other from a functional, goal achieving perspective. This is one kind of meaning that video games can convey, but there are many other kinds. In my series Metaphor Meaning Matriculation, I explored how video games convey ideas and meaning through their symbols, forms, and the interaction found in the gameplay. Because video games can feature so many different kinds of art forms or sub-media from text, music, photography, film, sculpture, and dance, video games have many different ways to convey meaning and therefore countless of ways to create resonance and depth that have never been possible before.
Multiplayer Stress Test for Depth
If a game survives the pressure of multiplayer competition by maintaining a healthy balance of viable options and strategies, then it is almost assuredly a deep game. The competitive gaming environment is one where most players have a playing-to-win attitude. For the most part, these players will do what it takes to acquire new skills and gain advantages over their opponents. This is why dominant or unbalanced strategies can have a poisonous effect on the community. Though it's possible that players can play a very shallow game competitively, continue to play a game after all its depth has been uncovered, or fail to uncover significant portions of a game's depth, these possibilities are less and less likely to be the case the longer a game is played competitively.
Tightness and Depth
Depth isn't the highest level of design for a game or any other art form. As I described in part 1, a work can have a lot of depth from including certain complexities, but it can also have a lot of complexities that don't support this depth, other meanings in the work, or anything significant. Just to clarify, how efficiently a work achieves its depth is what I call tightness. This is the same idea as the "water tight" or "air tight" concepts I've used before. Imagine a work where every bit, every word, every note, every scene, or every gameplay rule was a carefully tuned addition that supports the deepest meaning efficiently without detracting from any other part of the experience.
For the record, I think Inception and Wall-E are two incredibly tight films. If I had more time, I'd draw a line between every scene, every line, every concept, and every theme in the film and show you exactly how it's all woven together. Video games are a bit different. Surely linear, straightforward games can have a comparable kind of tightness to other linear art forms. For example, Rhythm Heaven is an incredibly tight game in terms of its variety and cleanness of gameplay from so few element. Typically, because video games are large products with lots of different types of content, we consider individual modes for their tightness. So for the same game, one mode may feature many more complexities than another. For example, The Legend of Zelda: Spirit Tracks has a large single player adventure and a much smaller multiplayer mode. Both share some gameplay assets, but the multiplayer mode does a lot for the few assets it uses.
Depth, Challenge, & Level Design
While gameplay systems can be very deep, level challenges tend to restrict the openness of gameplay possibilities. As I concluded in my series Linearity. Emergence. Convergence openness in level design naturally works against gameplay. If a level presents a specific challenge to test players, then the gameplay isn't about all the different ways to find other solutions. The challenge is finding and executing on the few right paths. So while gameplay systems can be designed with lots of interactive depth, level design seeks to squeeze these possibilities down to a few. All of this should sound familiar to you by now. Put simply, multiplayer thrives on open, deep gameplay systems; level design and campaigns do not.
Dominant Strategies & Solvability
It's pretty clear that dominant strategies are the enemy of interesting choices and therefore a common type of gameplay depth. However, I should clarify that the term "dominant strategy" is a bit nebulous. A strategy is just a specific plan of action. It can be as complex or as flexible as you can imagine. And even in deep games like chess, we can construct elaborate strategies where all possibilities are taken into account. After all, isn't this how chess AI works? But we wouldn't say that chess AIs make chess a shallow game.
What we're most concern about when we talk about dominant strategies is how straightforward, obvious, easy to figure out, or non-complex a dominant strategy is. If a dominant strategy is impractical to pull off because of its difficulty or requires many adjustments to be effective, the gameplay system can still be deep. This also applies to gameplay challenges that can be solved or challenges that players can create complete strategies for. Checkers has been solved using super computers. With the complete strategy of checkers, a player cannot lose. This discovery shouldn't ruin checkers for most players. Unless you play anything short of a grandmaster or a checkers AI using the complete strategy data, you probably have enough agency to win or lose on your own.
Keep in mind going forward that depth isn't a check box where a game either has it or not at all. Depth is a scale. Game depth comes in degrees. And there are many ways of creating depth. Stay tuned for my next series on the elegance of meaning in simplicity.