Path of Exile 2: The Action RPG That Respects Complexity

Grinding Gear Games announced Path of Exile 2 at its own ExileCon event in 2023, framing it as a standalone game built to run alongside the original rather than a sequel that replaces it, sharing an economy and a client while offering an entirely reworked campaign, skill system and combat model. This piece isn’t a review — the game hadn’t reached players by the time it was written, and reviewing a build nobody outside the studio has touched would mean fabricating an experience I haven’t had, which isn’t a thing worth doing to anyone reading this for an honest account of a game. What follows is a systems read of what GGG has actually shown publicly: the reveal presentations, the developer walkthroughs, the design rationale the studio has been unusually open about. Even from the outside, the shape of the bet GGG is making is legible, and it’s worth taking seriously on its own terms.

The original’s reputation as both asset and liability

Path of Exile, GGG’s 2013 free-to-play action-RPG, built its identity as the loot-RPG for players who found Diablo’s model too smooth — a passive skill tree with well over a thousand nodes, dozens of playable ascendancy classes, and an item-affix system deep enough that entire community tools exist purely to help players parse what a given piece of gear is actually doing. That reputation is an asset: a decade-plus of committed players who like their systems dense and their builds genuinely their own. It’s also a liability GGG has openly acknowledged, describing new-player retention in the original game as a persistent problem, because a thousand-node passive tree is a genuinely hostile thing to show someone in their first hour. Path of Exile 2’s entire pitch is trying to resolve that tension without abandoning the density that earned the original its loyal base in the first place — complexity for players who want it, without complexity as the first thing a newcomer has to survive.

Dual specialisation as the headline system

The most consequential announced change is dual specialisation: every character carries two complete skill and gear loadouts that can be swapped in the field, rather than committing to a single build for the length of a campaign run. GGG has framed this as solving two problems simultaneously — giving players a genuine answer to encounters that punish a single build archetype (a fire-immune boss against a fire-focused character, for instance) without forcing a full respec, and giving the game room to design encounters around that flexibility rather than working around a rigid single-loadout assumption the way the original game’s boss design largely had to. It’s a structural bet that respects the same instinct that made Path of Exile’s original passive tree beloved — more meaningful choice, not less — while acknowledging that “more choice” and “more friction” aren’t actually the same axis, which is the distinction the original game sometimes blurred.

A campaign built around fewer, denser acts

GGG has been explicit that Path of Exile 2’s campaign trades the original’s sprawling act count for a smaller number of considerably denser acts, each built around set-piece boss encounters that gate progression rather than a long corridor of trash encounters between checkpoints. The stated design goal is a campaign that plays well on a first run rather than existing mainly as an obstacle veteran players want to skip as fast as possible to reach endgame content, which the original Path of Exile’s own community has spent a decade treating exactly that way, speedrunning the campaign every league restart because the destination, not the ten-plus hour journey, was always where the interesting decisions lived. Whether a denser campaign actually gets played rather than sprinted through on the twentieth character is the kind of claim only real player behaviour after launch can settle, and it’s the single biggest open question hanging over the whole redesign.

Itemisation and the affix-bloat problem

The original Path of Exile’s item-affix system grew, expansion after expansion, into something the game’s own community widely describes as bloated — so many possible modifiers on so many item bases that evaluating a drop at a glance became close to impossible without a browser tab open to a wiki or a community pricing tool running in the background. GGG has said publicly that Path of Exile 2’s itemisation is being rebuilt with that specific problem in mind: fewer, more legible affixes, a stated goal of letting a player recognise a strong item without external tools. That’s a meaningfully different promise than “more complexity,” and it’s the part of the reveal that most directly answers the genre’s dominant recent argument — the one Diablo IV and its live-service calendar represent — that a loot game’s economy needs to be simple enough to parse in a single glance to hold a mainstream audience. GGG’s counter-bet is that legibility and depth aren’t actually in tension, they were just never separated properly in the original game’s design.

Ascendancy classes as genuinely separate identities

The original Path of Exile’s ascendancy system lets each of its base classes branch into three or four specialised subclasses late in the campaign, each granting a small dedicated passive tree tacked onto the edge of the shared thousand-node web. GGG’s reveal material for Path of Exile 2 describes a more ambitious version of the same idea: a larger roster of ascendancy options per base class, each intended to feel like a meaningfully distinct build identity rather than a modest bonus grafted onto a shared trunk. Given that the shared passive tree was one of the original game’s most celebrated systems and simultaneously one of its most intimidating to a new player, the stated goal of making ascendancy choice feel like the moment a build truly becomes its own thing, rather than a footnote applied after the real decisions were made on the main tree, is a direct attempt to give players an earlier, clearer sense of identity before they’ve had to parse a thousand nodes to find it.

Weapon swap and monster AI as combat-facing changes

Two further announced systems matter less as headline features and more as evidence of how thoroughly the redesign reaches into moment-to-moment combat rather than stopping at menus and trees. A weapon-swap system lets a character carry two full weapon sets and switch between them mid-fight, extending the same “give the player an answer rather than force a respec” philosophy behind dual specialisation down to a single encounter’s tactical decisions. GGG has also discussed a substantial monster AI overhaul, enemies that group, flank and react to a player’s positioning rather than converging in a single undifferentiated mass the way the original game’s weaker monster archetypes often did. Combined with boss encounters explicitly built as checkpoints gating waypoint access rather than optional detours, the intent is combat that asks more of positioning and timing throughout a run, not only during the campaign’s marquee boss fights.

What the genre’s history says about this bet

Diablo II: Resurrected is worth holding up against this reveal precisely because it’s the genre’s proof that unforgiving complexity, obscure runeword recipes, punishing act transitions, builds discovered by the community rather than explained by the game, can sustain a dedicated audience for over two decades without ever becoming more accessible. Nioh 2 made a comparable bet more recently, grafting a genuinely deep itemisation engine onto punishing combat and trusting that the audience for both halves would overlap enough to sustain the hybrid. Path of Exile 2’s dual specialisation and denser campaign read as GGG trying to keep Diablo II’s kind of density while removing friction that was never actually load-bearing to that density — accessibility aimed at the campaign’s front door, not at the thousand-node tree waiting behind it.

Why GGG’s track record is the reason to take the reveal seriously

Grinding Gear Games launched the original Path of Exile in 2013 as a free-to-play alternative built specifically against what its founders saw as Diablo III’s oversimplified endgame at the time, funded and developed by a small New Zealand studio before Tencent acquired majority ownership in 2018, a change that hasn’t visibly altered the studio’s release cadence or its willingness to run genuinely disruptive multi-month league experiments that reshape entire systems rather than simply adding new items. A decade of quarterly league updates, some of which fundamentally reworked core mechanics like the passive tree’s scaling or the crafting bench, gives GGG a track record most studios announcing a sequel can’t claim: they have publicly tried, failed, and revised complex systems in front of their own playerbase for years, which is a different kind of credibility than a polished announcement trailer alone provides. That history doesn’t guarantee Path of Exile 2 lands its stated goals, but it’s a legitimate reason to weigh GGG’s design claims more heavily than an equivalent pitch from a studio with no comparable public record of iterating honestly on its own mistakes.

The honest caveat

None of this is a review, and it shouldn’t be read as one. Everything here comes from GGG’s own public presentations and developer commentary, not from hands-on time with a build, and a studio’s stated design goals are not the same thing as a shipped game meeting them — GGG’s own history includes systems announced with confidence that needed real reworking once players got hold of them. What can be said with confidence is that the announced direction is a coherent, well-reasoned response to a real design problem the original game had, argued by a studio with an unusually strong track record of following through on its stated priorities over a decade of league updates. Whether the campaign stays dense on a twentieth replay, whether the new itemisation actually reads clearly at a glance mid-fight, and whether dual specialisation solves build rigidity without introducing its own new kind of busywork are all questions only real playtime can answer, and this piece will owe you a proper review once that playtime is honestly mine to report.