Studies conducted by preservation organisations have found that a substantial majority of games released before the modern digital era are not commercially available in any form. Not on a storefront, not in a compilation, not through any legitimate route.
For a medium barely fifty years old, that's a remarkable rate of loss, and it compares badly with film, music and literature — none of which have great records either.
Why games are harder than film
Several reasons that are genuinely technical rather than excuses.
Games are software. A film is a sequence of images; a game is a program that must run on hardware. Preserving it means preserving either the hardware or an accurate simulation of it, and both degrade or drift over time.
Rights are fragmented. A single game may involve licensed music, licensed likenesses, licensed brands, engine licences and middleware, each with separate terms and expiry dates. When a music licence lapses, the game becomes unsellable, and re-clearing it is often uneconomic. This is why several well-known titles vanished from storefronts.
Online dependency. A growing share of games require servers to function at all. When those servers retire, the game may become entirely unplayable — not degraded, gone. This applies to single-player games with online authentication as much as to multiplayer ones.
Studios close. Ownership of a defunct developer's catalogue frequently becomes unclear, and establishing who holds rights can be impossible. Orphan works are a problem across all media and games have an unusually high rate.
The legal barrier
Here's the part that frustrates archivists most, because the technical problems are largely solvable and the legal ones aren't.
Emulation of old hardware is mature. The technical capability to run and preserve most historical games exists and works well.
What doesn't exist in most jurisdictions is a legal framework permitting institutions to make preserved games accessible. Copyright terms are long, exceptions for preservation are narrow, and remote access is frequently prohibited entirely — meaning a researcher must physically visit an archive to view a game, which is a serious constraint for a global research community.
Efforts to expand these exemptions have met sustained opposition from industry bodies, generally on the argument that broader access would harm commercial re-release markets. Preservation advocates counter that the overwhelming majority of affected titles have no commercial market at all, which the availability studies support.
Who's actually doing the work
Largely enthusiasts, and largely in a legal grey area.
The emulation community has reverse-engineered generations of hardware. Preservation projects have catalogued and dumped enormous quantities of software. Documentation efforts have recorded interfaces, manuals, packaging and marketing material that would otherwise be lost.
Much of this activity is technically infringing, and the people doing it are the reason a great deal of gaming history still exists in any form.
There are also legitimate institutional efforts — museums and university archives with collections and expertise — but they operate under the access restrictions described above, which limits their public value.
The modern problem is worse
An uncomfortable point. Older games were physical objects. A cartridge from 1988 still exists in thousands of homes, and its contents can be extracted.
A digital-only release from last year exists on servers. When it's delisted, there is no physical artefact anywhere. The only copies are on the hard drives of people who bought it, and those are frequently encrypted and tied to accounts.
Games as a service are the extreme case. A title that exists only as a live operation, updated continuously, has no canonical version at all. When it shuts down, there's nothing to preserve — no build represents what the game was at any given moment, and its whole history is gone.
Several substantial titles have already vanished this way, and more will.
What would help
A few things that preservation advocates have proposed and that seem reasonable.
Legal deposit for games, as exists for books in many countries — a copy lodged with a national library on release.
Expanded preservation exceptions permitting institutions to provide remote access to titles that are not commercially available, with a clear test for availability.
An obligation, or at least a norm, for publishers to release server software or offline patches when discontinuing an online-dependent title. Some developers do this voluntarily and it's a genuine service.
And clearer treatment of orphan works, so that titles with untraceable owners can be preserved without indefinite legal risk.
None of these is radical and none has broad adoption. Meanwhile the medium keeps producing work faster than anybody is keeping it, and the gap widens every year.
What individuals can do
Not much at scale, but a few things genuinely help.
Keep physical copies where they exist, and keep the boxes and manuals — packaging and documentation are frequently lost even where the software survives, and they carry a great deal of context about how something was presented and understood.
Support the archives doing this work. Several preservation organisations operate on small budgets and accept donations of both money and material, and they are the reason a great deal of this history exists at all.
And when a service announces a shutdown, pay attention to whether anybody is capturing it. Community efforts to document and archive a closing game have preserved things that would otherwise simply have stopped existing, and they generally need to start before the servers go dark rather than after.