Valorant is a Rhythm Game That Produces Angry AI

Why does the introduction of new playable characters in live service games like Valorant incentivise re-engagement with the experience with minimal structural changes to the base game?

Hypothesis: players are primarily driven by the desire to participate in incrementally novel experiences; i.e., experiences which are new, but which are fundamentally rooted in familiar ‘safe’ circumstances.
Because this novelty is fuelled by interactions, i.e., player-to-player and player-to-environment, the only choices available are to develop novel interactions. And because the goal is to minimize cost while maximizing engagement, the introduced novelty has to be based on remixes of existing functionality, skinned with different contextual and aesthetic layers.

An experience can therefore be renewed by reproducing the existing ability set as permutations varied by the degree to which the ability interacts with the physical context of the game; for example, the range to degree of manipulation available to the player. This combines with the narrative context of the game, in the stories and aesthetics of the new playable character, to create a sense of an experience that is both novel and familiar.

Familiarity in competitive live service games is key. Pure novelty could never satisfy players in this context because of the imperative to remain dominant in the rankings of individual rounds, individual games, seasonal rankings, and informal social hierarchies. Truly new mechanics would require a seasoned player to inhabit the head-space and social-space of a newbie, open to and even expecting failure in the process of learning. The ranking system precludes this possibility, as every aspect of the game’s incentive structure is dedicated to associating an equivalence between ‘winning’ and ‘having a positive experience’. Failure in live service games is not designed to be enjoyable and entertaining; it is designed to be harsh and punishing.

This further intensifies a tendency towards (hah) convergent evolution in gameplay styles. Devs and players are in a competition to develop dominant strategies of play for each map, tool, and character. This convergence is fuelled by gaming streamers and competitive play, where dominant strategies are quickly discovered, refined, countered, and modified or discarded as developmental changes or cultural shifts make them redundant or obsolete. Rather than reducing the costs of development, the inflexibility of the base abilities and physics of the game, the competitive culture driven by a lack of failure support, and the convergence of play styles through dominant strategies make it so that any novelty a player can experience in the game must be introduced through development, rather than emerging naturally as a consequence of co-creation between player-system-player.

This convergence has morphed the average live service experience for a player in a very interesting way. Each player can now expect every character in game, both on their team and the opposing team, to act within a narrow set of actions and responses that constitute an ‘acceptable’ play style, i.e., that which inherits from the dominant ‘optimal’ play-style sold by top players in the competitive/streaming space. Every round of gameplay becomes less about outwitting other human players, and more about how well you can adhere to the optimal style, in the vein of rhythm games. If every regular player is playing that way, then what you have is closer to a multiplayer game where every player is controlled by a sophisticated AI mirroring optimal play with randomized error introduced for gameplay balance, than an experience representative of inter-human play. If we removed the randomization element where team composition, map layout, and tool selection can differ, it is more than likely that you could reproduce an exact game, kill for kill, while using completely different players.

In short, Valorant and its consumers have co-created an experience so mass-produced and reproducible that each player functions are variations on the same AI. All that happens when a new map or character is introduced is that the system now has new stimulus to run through the pipeline, with no room for emergent gameplay or variations whatsoever. The proof would be in the fact that any system like this would be fundamentally antagonistic towards newcomers in the space, or rather, antagonistic towards those who do not engage in the optimal, dominant strategies already established.

What might have initially been considered a revolutionary method to maximize revenue by minimizing development costs is instead a hamster-wheel live service companies are stuck on; they’ve created an environment where players are rewarded if they learn to be socially hierarchical, antagonistic to change, averse to failure or risk, to crave continuous, incremental novelty, and be vicious towards newcomers who might dilute or redirect cultural trends, and are now stuck trying to please the audience they’ve created. Their options are to feed the machine with new content, and add increments of low-cost novelty through character “re-balancing” to keep players on the platform until the next update launches.

The key problem here is player novelty and development costs. Any solution that requires an upheaval of the engine would be more expensive that could ever be worth the cost. And any solution that limits the continuous novelty provided by the current flow would be met with tears and anger from the player base. The solution would have to demand enough developmental resources continuously to keep developers employed, but without requiring the degree of crunch that is emblematic of work in the gaming industry (of course, the problem with crunch isn’t related to ‘necessity’ so much as ‘greed’, but let’s pretend). An ideal solution would transfer the demand for novelty from the system onto individual players/rounds instead. But it shouldn’t change gameplay at all; or rather, you can only change gameplay in ways that are culturally approved of, e.g., rebalancing, or adding new characters and maps.

Our Goal:

• Have players feel invested in continuing to play Valorant
• Encourage regular players to encourage new players to join Valorant
• Reduce the necessity for development to provide all novelty in the gameplay experience

Core Considerations:

• External vs Internal Motivation: People who are invested in Valorant, play Aim Labs religiously, where the only entertainment value is the internal experience of progress against their own shadow, and the hypothetical potential of progress in Valorant.
• Cost of Failure: The cost of failure prevents players from developing a positive relationship with the experience of playing, which impacts their desire to engage with randomizing elements, e.g., new players and experimentation with play styles, which would organically introduce novelty to the game.
• Reproducible Experiences: Players know what they want from Valorant, and they know that each gameplay loop will reproduce a specific experience they cannot find anywhere else; changing the functionality and aesthetics of the core gameplay loop risks alienating core players.

The Solution:

Keep the gameplay the exact same and retain the same incentives for winning, but incentivise risk taking by removing the cost of failure: i.e., wins increase rank, and losses do nothing. Culturally, players already communicate their abilities by the ‘highest’ rank they reach, not their current rank, so allow players to increase their rank through quality plays, without punishing failure. Gameplay-wise, players are actively more likely to play another round when their rank is ‘shielded’ by a rank increase than when not. Punishing losses has always been an artificial decision made at the start of development, the costs of which now outweigh the benefits in this context.

Redirect competition away from external sources, especially away from team-members, and towards intrinsic growth: rather than highlighting team statistics during and at the end of gameplay, introduce and highlight evaluations of personal skill and execution of tactics. With the move towards laddering casual players towards competitive gameplay, integrated skill tracking that presents data already being catalogued to calibrate rank match-ups can be leveraged to redirect motivation inwards. Simple stats that are already displayed at the end of a round/death, can also be repurposed.

The Caveats:

While matching is already controlled by an internal ranking system that differs from the visualization provided to players, the cultural role of the visible ranking system, especially in structuring team hierarchies in-game means that a ‘line go up’ method may match individuals who are locally close but have a wide gap in seasonal ranking. The problem here would be a general dissatisfaction with such a presentation, however, the presentation is very much the root of the problem. The issue is cultural, not functional. Mediating this would require an incremental reworking of the system in tandem with the other solutions presented.

While the purpose of integrated skill tracking would be to transmute external competition into internal motivation, players might use it as a way to showcase their accomplishments in a way that intensifies competition, especially given the existing cultural traditions of the user base. A practical (and entertaining) solution might be to pair the positive skill growth with indications of where the player absolutely shat the bed, so that any screenshot would expose them to their peers. Those who continue to share anyway would be those with thick skin, willing to laugh at themselves with others, which would only support collaborative competition between teams.

Amusement and meme-ing are the central foundations of digital cultures. Novelty and enjoyment are bread and butter to the generations that were raised with internet access. The only solution is a playful one that accounts for the systems at play, not just functional development. If a systematic view is taken, the maybe the first reaction when someone says “I quit playing Valorant” won’t be “Congratulations!”.

Amusement and meme-ing are the central foundations of digital cultures. Novelty and enjoyment are bread and butter to the generations that were raised with internet access. The only solution is a playful one that accounts for the systems at play, not just functional development. If a systematic view is taken, the maybe the first reaction when someone says “I quit playing Valorant” won’t be “Congratulations!”.