A substantial share of the industry has moved toward games designed to operate continuously, updated over years, funded by ongoing spending rather than a single purchase.
Many of them close. When they do, the result is frequently that nothing remains, and that is a genuinely new situation.
What ending actually means
For a conventional game, support ending means no more patches. The game still runs.
For a service game, the servers hold the logic, the content and the player's progress. Shutting them down means the software on the player's machine does nothing.
Which means the game does not become unsupported; it ceases to exist.
Purchases made within it — cosmetics, expansions, currency — cease to exist with it, and the terms of service generally make clear that no refund is owed, since what was sold was a licence to use something within a service.
Why they close
The economics are demanding.
Continuous operation requires continuous staffing — servers, content, moderation, support — and that cost does not fall as the player base does.
Revenue in these games is concentrated among a small proportion of players who spend heavily, which makes income volatile and dependent on retaining specific people.
The market is also saturated. Player attention is finite and a small number of established titles absorb most of it, which means new entrants struggle regardless of quality.
The result is a high failure rate, including for games with substantial budgets and reasonable reception.
The preservation problem
The consequence that matters beyond individual disappointment.
A closed service game generally cannot be archived. There is no complete copy, since the essential half ran on servers that no longer exist.
Archives and libraries collecting games can preserve the client and cannot preserve the experience.
Which means a category of work — some of it culturally significant, some of it representing years of collaborative activity by large communities — is being lost as it goes, with no mechanism to prevent it.
Film and music have archives, deposit requirements and physical copies. This has none of that.
The private server question
Where communities have responded and where the law is contested.
Enthusiasts have reconstructed server software for closed games, allowing them to be played again. This has happened for a number of significant titles and represents substantial technical work.
The legal position varies and is generally unfavourable, since the client software remains under copyright and reverse engineering the server may breach terms or circumvention provisions.
Rights holders have responded inconsistently — some tolerate it, some have sent legal demands, a small number have released server tools deliberately.
That last option is the obvious solution and it is entirely within a publisher's gift.
The consumer protection angle
An area where regulation has started to move.
Campaigns have argued that selling access to something which can be unilaterally destroyed, without refund, is a consumer protection issue rather than merely a commercial risk.
Proposals have included requiring publishers to leave games in a playable state, to provide server tools on shutdown, or to disclose expected support periods at the point of sale.
Regulatory interest has grown in several jurisdictions, and the industry position is that mandated end-of-life provision would raise costs and discourage the model.
Both of those are probably true, which is what makes it a genuine policy question rather than an obvious one.
What a buyer can reasonably do
Assume the game will end, and treat purchases within it as consumption rather than acquisition.
Prefer publishers with a record of leaving games playable, which a few have established deliberately.
Note whether an offline mode exists, since a game that functions without servers survives the shutdown.
And recognise that the money spent is buying the time spent, which is a reasonable trade if made knowingly and is not what the presentation implies.
The wider pattern
This is the same problem as software subscriptions, cloud-dependent devices and streaming catalogues, in a form where the loss is total rather than partial.
The common thread is that access has replaced ownership across every category of media, and the consequences become visible only when the provider stops.
Games are where it is starkest, because the thing that disappears is not a file but a place that people spent years in.
What a good shutdown looks like
Some publishers have handled this well and it is worth describing.
Long notice, generally months rather than weeks, so that communities can arrange final events.
Suspension of further purchases once the decision is made, which prevents people spending on something about to end.
Refunds for recent spending, which is not legally required in most cases and has been offered.
And in a small number of cases, release of server software or an offline mode, which preserves the work entirely.
None of that is expensive relative to the cost of having built the game, and the difference in how communities remember a title is substantial.