PC Gamer has two Autumn Sale roundups up, and they quietly contradict each other.

The first is nine games released in 2026 that are already discounted, including three that launched at $70 and now sit under $50. The second is eleven underappreciated games picked by individual staffers, each one chosen by somebody who plays too much of a single genre and wants company.

Buy from the second list. The first one is evidence, not advice.

A calendar year is the new shelf life

The $70 price tier was sold to us as a response to rising development costs, a permanent correction after fifteen years of $60 holding steady. What it actually functions as is an opening ask. Three games from this year’s release slate have already given up roughly a third of that number, and we are not even through the first full sale cycle for some of them.

So the launch-day buyer paid a premium to be a QA tester for a game that is now cheaper, patched, and running better on their GPU. This is not a trap anyone fell into by accident. Publishers have been running this schedule long enough that it is a schedule, and players have noticed, which is why preorder culture now depends so heavily on cosmetic bribes and three-day early access. The bribe is the tell. If the price held, you wouldn’t need to be paid to pay early.

I don’t think the $70 tier survives in any meaningful sense. It exists for the first eight weeks of a release, for the people who were always going to buy on day one, and then it dissolves. Everyone else is shopping in a $40 to $50 market.

Eleven sickos beat the algorithm

The staff-picks list is the more useful artifact, and it’s useful for an unglamorous reason: Steam’s own sale discovery is terrible. The front page is a wall of things you already own or already declined, shuffled by an algorithm optimizing for conversion on known quantities. Nothing in that system is designed to surface a strange tactics game with 900 reviews.

A named person with a documented obsession, writing one paragraph about why they refuse to shut up about a game, is a better recommendation engine than anything Valve has shipped. It is accountable in a way a carousel isn’t. You learn whose taste overlaps with yours and you follow that person forever.

Worth saying plainly: both lists are also commerce. Sale roundups are a revenue event for every outlet that runs them, affiliate links included, and the volume of them in late November is not journalism responding to reader demand. The staff version at least makes its bias legible. Someone’s name is on it.

Three $70 games are under $50 before their first year is out. The thing I’d watch is how fast the next one gets there.