Live service games get criticised constantly and make billions of dollars. The two facts seem contradictory until you understand the business model — at which point they become inevitable. The criticism is structurally correct. The financial results are also structurally correct. The problem is that both are true simultaneously.
This piece isn't about whether live service is good or bad. It's about why the model produces the specific design patterns it produces, and why studios running live service games cannot easily change those patterns even when they want to.
The core economic commitment
A traditional boxed game is a single transaction. The studio earns revenue at the point of sale and then relationship with the customer is largely over. The sequel is a separate transaction with a separate evaluation.
A live service game trades that model for a continuous revenue relationship. Revenue comes in over time — through battle passes, cosmetic purchases, season content, expanded access. The business model requires the player to keep playing, which requires the game to keep asking for continued engagement.
This creates a structural dependency on retention that a boxed game doesn't have. A boxed game can end. A live service game that ends is failing. The development team working on a live service game is not making a game — they're maintaining a relationship with a player base that must never feel like the relationship is over.
What retention actually requires
Retention at the scale live service requires it is not the same thing as making a good game. Making a good game creates the desire to play it. Retention engineering creates the cost of not playing it.
Daily login rewards, time-limited content, battle pass expiry dates, friend group activity — these aren't features in the design sense. They're loss-aversion mechanisms. They make leaving the game feel like losing something rather than simply ending a session. The player who doesn't log in for three days hasn't missed out on content; they've incurred a kind of psychological debt that the game suggests they owe themselves.
Good game design makes you want to play. Retention engineering makes you feel bad about not playing. These are different relationships with the player, and they tend to feel different from inside the game. Players often can't articulate the distinction, but they feel it — something about the game's relationship to their time starts to feel coercive rather than inviting.
The content treadmill and why it resists improvement
Live service games need a constant supply of new content. Battle passes require new cosmetics. Seasons require new mechanics or maps. Events require limited-time modes and rewards. The content cadence that players have come to expect from major live service games is substantial — in some cases, the equivalent of a full game release worth of material every few months.
This content demand is fundamentally at odds with design iteration. Good design involves trying things, seeing how players respond, and changing based on what you learn. That process is slow. The content treadmill is fast. The team that should be spending three months understanding why a new mechanic isn't working is also the team that needs to ship next season's content in six weeks.
The result is that live service games often get stuck. They ship a design that works well enough to retain players, but not well enough to be genuinely good. They can't fix it because fixing it requires the kind of extended iteration cycle the content schedule doesn't permit. So the design calcifies. The game becomes the same experience, served continuously, with new surface variation on top.
Why studios enter the model anyway
The revenue potential is clear, the player numbers are verifiable, and the competitive advantage of an established live service player base is real. A studio that successfully runs a live service game has a stable revenue stream that a series of boxed games doesn't produce. The financial case is rational.
What tends to get underweighted in the initial decision is the operational overhead. Running a live service game is not making a game and then serving its community. It's a continuous production operation. The team that shipped the game doesn't get to move on to something new. They're maintaining and feeding the thing they built, indefinitely, at the cadence the player expectation requires.
The studios that have managed this well — the ones whose live service games are both financially successful and well-regarded as games — tend to have made one specific choice: they decided what the game is, committed to that, and built the content cadence around the game's identity rather than the inverse. That sounds obvious stated plainly, but it requires resisting significant pressure in the other direction.
The player's side of the trap
Players are not passive in this dynamic. The psychology of sunk cost is real: a player who has spent money on a live service game has an incentive to keep playing that is independent of whether they're enjoying it. Leaving means the prior purchase was wasted. Staying means it's still active investment.
This is compounded by social mechanics. Live service games are often most enjoyable with a consistent group. The group has its own momentum — if you leave, you lose the social infrastructure. The game becomes a meeting place that's inconvenient to stop attending even when you'd rather do something else.
Understanding this dynamic isn't a justification for the design choices that exploit it. It's an explanation for why the market is as sticky as it appears. Players who would describe their relationship with a live service game as frustrating continue to play. They're not being irrational. They're responding to a system designed to make not playing more costly than playing.