殺戮尖塔 2

Put every fact on the table and the sequel is a tower still under construction whose frame already stands tall. The overall 7.8 holds both truths at once — the brilliance of the combat (9.2) and the turbulence of early access (a 63% reputation). It's an honest compromise.

A 574,000-concurrent phenomenon, and the two review-bomb waves it earned

Start with how big it is. 5.3 million copies in the first month, roughly $108 million in revenue for Mega Crit, and a launch peak of 574,638 concurrent players — almost matching Silksong's record, briefly outpacing Marathon, Arc Raiders, and Resident Evil Requiem, the largest Steam launch of 2026 and the highest-concurrent roguelike in history. For a minimalist indie with no major publisher and no marketing spend, those numbers were snowballed almost entirely by the goodwill the first game banked — and by a price of admission low enough that almost any PC could run it. In March 2026 it reportedly topped Steam's monthly best-seller chart outright, ahead of much larger releases like Crimson Desert.

But attention cuts both ways. Two review-bomb waves — the first over the infinite-combo cull, the second over Doormaker, an Act 3 boss that shut down too many archetypes at once — dragged the rating from Overwhelmingly Positive down to Mixed. Nearly 80,000 negative reviews have piled up, roughly fourteen times what the original drew across nine years. At its core this isn't a collapse in quality; it's a public tug-of-war between "the fun players want" and "the long-term health the developer wants."

Mega Crit's stance is refreshingly candid: "this progress will not be linear, and no change is necessarily permanent" — and it later rolled several contested nerfs back on beta. A Steam analyst put it well: judged by the raw numbers, "in any other case you would say this game is dead," yet the bombing barely dented sales at all.

Who should climb this tower now, and who should wait

The practical call: if you love the original's draw-and-trade tension and can live with a living build that re-tunes its balance monthly and lurches between versions, jumping in now pays off — wider builds than the original, two genuinely new classes, friends to climb with, and a front-row seat as the tower takes shape. But if you want a settled, finalized balance and would resent your favorite build getting axed by the next patch, the rational move is to wait for 1.0. And anyone simply testing whether they vibe with the genre at all can start with the original — finished, cheap, and rock-solid. In the end, much of that 63% is people rating "am I having fun in this exact patch," while the 9s from critics are rating "how good is this underlying design." They aren't grading the same thing — and your job is deciding which one you care about more.

There's a more grounded reference point, too. That Steam analyst noted that judged purely on the scale of the bombing, "in any other case you would say this game is dead," yet sales barely flinched — 5.3 million in a month, a steady 100,000 to 300,000 concurrents long after launch. That's a rare and telling signal: the overwhelming majority of people actually playing voted with their feet, and the negative reviews stem more from the emotional whiplash of each balance change than from any verdict on the game itself. So if you're on the fence, drop that glaring 63% back into this context and read it again. It isn't a quality alarm. It's the sound of a tower popular enough that every controversy gets amplified, taking shape in full public view. All you have to decide is whether you want to be on the scaffolding now, while the design is still being argued into shape, or show up once the dust settles and climb a tower someone else helped finish.