Working out whether your computer can run it

Almost every publisher of a desktop game publishes two lists of hardware. They are usually labelled minimum and recommended, and the difference between them is not a matter of degree — the two lists answer different questions.

Minimum is the publisher's statement about where the game will start and run at all. A machine that just clears the minimum will typically run the game at a low resolution, with visual detail turned down, and with a frame rate that is playable rather than pleasant. It is a floor, not a target.

Recommended is the publisher's statement about the configuration the game was built around. On a machine at or above this list, you can usually leave the default settings alone and stop thinking about the subject.

Two things are worth knowing about both lists. They are written by the publisher, and a publisher has a commercial interest in the minimum list being generous. And they describe the game as it is today: a title that has been updated for a decade asks more of a computer than it did at launch, which is why the version of the list you want is the one on the publisher's current page, not a copy of it on a third-party site.

Every requirements list in this category contains the same five items, whatever order they appear in. Here is what each one is, in the plainest terms, and how much it tends to matter.

Two items are frequently left off these lists and still matter: the type of storage, since a solid-state drive shortens loading noticeably compared with a mechanical hard drive, and the screen, because a very large or very high-resolution screen asks more of the graphics chip than a modest one does.

You do not need to install anything, and you should be sceptical of any site offering to scan your computer and tell you what it has. Every desktop operating system reports its own hardware, and the information is a few clicks away.

On Windows

Open Settings, then System, then the About page. It lists the processor, the installed memory, the system type and the edition and version of Windows. For the graphics chip and free disk space, open Settings, then System, then Display and Storage respectively; the graphics information also appears in the Device Manager under Display adapters.

On macOS

Open the Apple menu and choose About This Mac. The panel names the model, the chip or processor, the memory and the macOS version. The More Info or System Report view lists the graphics and the storage in detail.

On Linux

Most desktop environments have a Settings application with an About or Details page carrying the same four facts. From a terminal, the commands lscpu, free -h, lspci | grep -i vga and df -h report the processor, the memory, the graphics device and the free disk space respectively.

Write the five answers down somewhere you will find them again — the Step 2 checklist has a place for them. You will use the same five numbers every time you consider a game, for years.

With both lists in front of you, the comparison is mechanical for three of the five items. The operating system either matches or it does not. The memory figure is a number you can compare directly. The free disk space is a number you can compare directly.

Processors and graphics chips are harder, because the lists name specific models rather than giving a score, and the model named is often a decade old and unfamiliar. The reliable approach is to look up the exact model name in your own machine on the manufacturer's own product pages — Intel, AMD, Apple and NVIDIA each publish specifications for their parts — and compare the release year and the market segment with the part named in the requirements. A chip two or three years newer in the same segment is generally at least as capable. This is a judgement rather than a calculation, and it is fine for it to end in "probably, at low settings".

If the comparison is genuinely ambiguous, the sensible order is: install, run the game once at the lowest settings, and judge with your own eyes. That costs you a download and an hour. It does not cost money, and it beats any amount of further reading.

The storage line on a requirements list is the one most often underestimated, for a simple reason: it describes the installation, and the installation grows. A client in this category can be tens of gigabytes on the day you install it, and updates through the year add to it rather than replacing it neatly.

Two practical rules follow. Leave roughly half again as much free space as the requirements list asks for, so an update does not leave you deleting things in a hurry. And keep an eye on the total: a desktop computer with two or three installed games of this size has a storage problem long before it has a performance problem.

The download itself is the other half of the question. The initial download is the same size as the installation, and in Australia that is a number worth checking against your plan before you start, particularly on a connection with a monthly data allowance or on a mobile or satellite service. Updates during the year add more. If your household shares a limited allowance, the considerate approach is to schedule the first download for a time when nobody else needs the connection, and to find out whether your service has an off-peak period.

For a game played against other people in real time, raw download speed matters much less than people assume. A connection fast enough to stream video is usually fast enough to play. What matters is latency: the time it takes for a message to travel from your computer to the game server and back, usually quoted in milliseconds.

Latency is governed largely by distance and by the number of hops in between. Australia's geography means that if the server you are matched to sits overseas, the round trip is longer than it would be for a player in the same country as the server, and no amount of paying for a faster plan changes the distance. Publishers in this category typically operate regional servers partly for that reason, and the practical question to ask of any title is which region you will be connected to.

Three things are within your control and worth doing before you blame anything else. Use a cable between your computer and the router rather than wireless, if the room allows it — this is the single largest improvement most households can make. Find out whether anyone else in the house is doing something heavy on the connection at the same time. And test at the hour you actually intend to play, because evening congestion is real and a midday test will flatter your connection.

A judgement, not a rule: if a short video call from the same room is steady, the connection is very likely adequate for this category of game, and anything that still feels wrong is worth investigating at the server end rather than at yours.

Our hypothetical household has two candidate machines. The five-year-old laptop reports a mid-range processor from its year, 8 GB of memory, graphics built into the processor, and 40 GB free. The second-hand desktop computer reports a slightly older processor, 16 GB of memory, a separate graphics card of unknown vintage, and 300 GB free.

Reading the table in section 2.2, the laptop's problems are the built-in graphics and the free space, and only one of those is cheaply fixable. The desktop computer's unknown is the graphics card, which is settled by looking up the model name printed in the system information against the manufacturer's own page. The household's decision is to install on the desktop computer, and to leave the laptop out of it — not because the laptop certainly cannot run the game, but because testing one machine properly is better than half-testing two.

They also check their plan's data allowance before starting the download, and schedule it for a weekday morning. Again: a hypothetical, written to show the shape of the decision, not a report about anybody.