Path of Exile 2 & the “Melee” Discussion
I’ve just gotten through the second act boss, and I’m in a transition area called Sandswept Marsh. I’m trying to get to the next town to really kick off my new Magma Barrier block character’s progression. I happen upon Rootdredge, an area boss that raises zombies. I put up my barrier with Resonating Shield and have to time it in the 0.7s window where the barrier will trigger. I realize that Rootdredge doesn’t really attack you, so I’m going to have to chip away at his health. A pack of zombies from the nearby bog gets into the battle, so now I can parry them to hit the boss. It’s slow. It’s painful. I wonder why I shelved my infinitely faster staff character.
My level 32 warrior, WillItBlock, finding out that in fact, you need an attack to connect to be able to block.
We Live in a Ranged World
My thoughts above aren’t too dissimilar from those of many others in the Path of Exile 2 (Grinding Gear Games, 2024) community who’ve had the distinct “pleasure” of playing with mace skills in this early access phase (see the end of this piece for more details about my total experience). Melee has been a recurring issue for Grinding Gear Games’ Path of Exile (2013) since its release more than a decade ago, and it seems like that may still be the case for this new entry, though in altogether new ways that dredge up old discussions of gameplay feel, difficulty and time investment. Notably, the current down-the-middle melee class Warrior, along with it’s ascended Warbringer and Titan classes, is at the heart of renewed discourse about how melee feels, how it should feel, and perform in numeric terms. This in turn contrasts the inverse proportion of classes that are blasting through the game and seemingly feel none of that friction. At the heart of this debate is question of vision about game speed, but also about identity, for archetypes and for pieces of the class roster that seem joined at the hip with several mechanics. What follows is a fairly technical, but not probably not in-depth enough discussion about melee, warrior, maces, armour, and the issues facing players, especially newer ones when first prompted with the class-selection screen.
The gemcutting pane where players can select which skills to play. Rolling Slam shown with a non-reducible 1.5s added attack delay.
The easiest way to get into that morass is maybe by discussing how skills are built in PoE2 relative to the game content, along with the tools the game gives you to even build. Skills are tied to weapon types. If you’d like to play the Rolling Slam skill, for instance, you have to either use a two-handed or one-handed mace for it to be active. This entails having the requisite attributes to equip the items and skills themselves (which in this case means a fair deal of strength, which can really scale to exorbitant amounts if you’d like to use the two-handed variant that currently seems strongest). To get those attributes you either have to spend passive points on the game’s signature skill tree, or you need to find pieces of gear that give strength. You would also want passive skills that increase either your damage with maces, shields, some attack based scaling or armour that’s thematically paired with the mace/warrior aesthetic. This is a reverse engineering way of thinking, planning where you want your character to be once complete, and then to map backwards what you need to accomplish that. That process invariably leads you to consider that the skill tree, especially for new players, incentivizes picking a class that starts relatively close to that area of the skill tree, settling most often on warrior, or potentially mercenary. This is a trap.
Veteran PoE players know that skills are not attached to classes, but they are loosely localized to skill tree areas, and it so happens that the red area (where strength, melee weapons, block and armour are located) is conventionally considered lackluster along multiple axes. In a recent interview for Tavern Talk, content creators GhazzyTV and DarthMicrotransactions interviewed PoE2 game directors Jonathan Rogers and Mark Roberts. In the interview, Darth fields a question (at 54:20) about the current feel of warriors/maces being much weaker than most other archetypes, which Jonathan acknowledges is the case, while also remarking that to him the class “felt fine” during early to mid segments of the game, which is where the campaign takes place (though we can certainly have another discussion about what the “late-game” even is). Darth on the other hand called melee “clunky”, a “punishing” experience for new players and generally a playstyle for “masochists.” Notably, he does allow for the idea that warrior and maces might have been designed earliest of all archetypes, when the game was envisioned as something slower, whereas the current version of thegame becomes very fast at higher levels, with certain mechanics showering you with enemies. Mark chimes in explaining that the feeling of the archetype being subpar probably has to do with fixed attack timers, like that of Rolling Slam, which can’t be fully scaled with attack/skill speed, that most other archetypes and classes have access to, which he terms a “multiplicative axis.” Mace skills are locked into being slow, which is a creative decision we may agree or disagree with. What’s clear though is that both Jonathan and Mark feel that maces are probably closer to a design sweet-spot than it seems to the community.
Live reacting to the stream, another content creator, Kostya Khudoshin (more commonly known as Mathil), exclaimed “did you play anything else?! Did you try playing a monk, by comparison, and not see the disparity? Maybe in a vacuum it’s fine, but what are you on about? You can’t have played two separate classes and have gone, yeah that’s ok.” Mathil has been producing builds for Path of Exile every few days for nearly a decade and provides a relatively accessible entry point for a range of players to enter the game. I take his perspective as someone who’s played in various contexts and budget points, often deciding to build creatively rather than optimally. So when he says there’s something off, there’s probably something off beyond the numbers. After playing a “thorns” warrior build for a few days, with watermarks of “YES I KNOW THIS IS BAD. PLEASE LET ME STRUGGLE,” Mathil shelved the character, explaining there was “no real way” to make it work and that the left side of the skill tree has some significant room for improvement in terms of support. Both creators’ perspectives track with community discourse on the subreddit and the forums, though we can also understanding that Jonathan’s perspective of the class being fine is not necessarily wrong either. That they differ so strongly though can probably be attributed to habit and preference for what the game should feel like. This isn’t a gotcha, clipping Mathil or taking Jonathan’s comments out of context, but rather an important divergence in viewpoints that sheds a lot of light about what the perceptions of melee are. This is a question of which version of the game we’d like to play. Mark himself lists attribute-stacking archetypes, a counterpoint, as an extreme outlier in terms of damage and speed, but in doing so produces a spectrum of game speed that we can measure characters against. Maces on the slow side, attribute-stackers on the side of speed (among other archetypes that I’m not going to dive into here).
Rethinking Who Plays How
There is a more traditional way of thinking through this problem that has its roots in thinking about habit, and how play even happens. Without getting too much into the weeds of academic discussion, this is a fairly standard analysis of habitus, a concept coined by French sociologist Pierre Bourdieu in 1977, which roughly refers to the how people respond to their social and cultural world based on their habits, skills and dispositions (or preferences/taste). To put into PoE2 terms, Jonathan and Mathil aren’t really playing the same game, despite the fact they boot the same piece of software, because they aren’t even really playing it the same way or for the same reasons. I’m not even sure that Mathil and Darth aren’t playing the same game either, because their community context is never the same, nor for the same audience necessarily. This is the double bind. There is one Path of Exile, as it is designed, but there are numerous Path of Exiles how the game is played. It isn’t even a question of skill alone, though the capacity to optimize builds that comes with a decade of daily practice for content creators certainly exacerbates the discrepancy. Mathil, Darth, Ghazzy and numerous other creators play to maximize their output, or to be creative, and by extension measure archetypes, and their take on those archetypes, in the context of every other archetype. These measurements come through relative speed, boss kills, and damage numbers on tooltip. Designers, like Jonathan and Mark, play classes to ensure that they have a coherent identity and feel. With that in mind, warrior feels like a warrior in terms of tactical feedback, visual effects and pace. It just happens that the community doesn’t enjoy that pace. That also doesn’t mean that warrior is necessarily unplayable or not viable. Creators like KalTorak and TheDoppelganker have been playing versions of the same Magma Barrier build I am and advancing to the furthest stages of the game, but my experience along with many others, posting on feedback avenues like the official forums and subreddit, is that warrior/maces/melee is in a weird spot in terms of play experience. It’s not just that warrior is slow, it’s that it’s slow in a game where everything else is fast. To that extent, there is clearly a discrepancy between how Jonathan views the archetype, how content creators build , and how the average player experiences it.
Rethinking the issue through that lens of “somewhat ok for devs” and “dire for players” can be flipped. Attribute stackers, like the above showcase, are all the rage in community videos and builds. The playstyle is largely agnostic to weapon type, and even somewhat to classes, with many expecting the playstyle to be significantly tuned down, despite their enjoyment, because of its ubiquity and performance. Mark Roberts states as much in the interview, calling stat-stackers, which are at the “top-top-end”, “absolute insanity” and “just ridiculous.” In the broader interview, there is a sense that archetypes like these trivialize the game and in some ways undo the intricate design of bosses and endgame fights. Players are effectively opting out of playing the game the way the developers would like them to. Roberts goes on to say that by comparison, it’s very easy for players to perhaps feel that “warrior will never be viable,” while explaining that there are in fact warrior/mace builds that are good in the endgame. This is, technically, correct. Other content creators like cArn_, mas0ny1 and Palsteron, have indeed been producing guides for endgame viable warrior builds, though all them rely on supplementing maces/melee with stat-stacking, automated totems or items/skills combinations that are conventionally understood by the creators themselves to not be working as intended (looking at you Polcirken+Herald of Ice and Trampletoe+Shield Wall). These builds also abuse one shot mechanics to invalidate end-game bosses, but they’re substantially slower and more fragile in other parts of the game. These are some more technical digressions, but the key takeaway is that there is a wide gulf between the game the developers envision and the players experience, and in that gulf playing a linear, intuitively built melee warrior is perhaps a ticket to a very different ride than players expect.
Clarity Matters
This is an issue, particularly for new players. If players need to dive into the kinds of builds and interactions discussed above, what does that say about the legibility of the archetype? There is a fundamental issue where players that pick maces are in many ways setting themselves up for disappointment. At the root of this disappointment is that the slower, clunkier playstyle doesn’t feel good. That lack of enjoyment is at odds with general action RPG trends of acceleration and power, where warrior is more sedate. And, to be clear, maces/warriors are slated for improvements in the upcoming larger balance patch on the horizon, but there’s also a compounding question of defenses. It turns out that armour, which is typically close to mace/melee, is also the weakest of all the defensive types in the game currently. Kripparian, a content creator often seen streaming Hearthstone who actually began as a PoE streamer, has been putting out fairly regular videos on armour math, which is its own can of worms, and explaining just how bad it is. Then, a few days later, he gets to put out another video about how armour is in fact even worse than it seems, after extensive testing.
The root of the issue is, once again, legibility. In-game, armour shows a percentage reduction, but it actually scales that reduction based on the size of hit damage. That scaling is only accessible through external wikis, forum posts, community Google docs and Excel spreadsheets. This is, in terms of accessibility, a nightmare scenario because not only is conventional melee weak, the archetype’s defenses don’t fully function as advertised at first glance, exacerbating the feeling of weakness. Armour can be strong, as shown in TheDoppelganker’s armour-stacking build, but it requires such specific tuning that even experienced players need to watch a 40-minute video to understand all the layers of interactions. That isn’t ideal, and should definitely be the topic of balance in the game and display clarity. This is because, on that habitus axis, new players may in fact be opting into a much more difficult style of play than they expect to. What seems intuitive, is actually complex. What seems strong is actually weak. What seems slow…well, that’s actually slower than maybe you thought. This is supplemented by the fact that evasion or energy shield, the thematic “rogue” and “mage” defense types (though that’s a gross oversimplification) do a much better job at being strong and intuitive defensive layers for staff-based melee characters. There’s no gotcha math if you’re in that part of the skill tree. Monks are, at face value, better at being tanks than warriors are, and though at a mechanical level that’s part of PoE’s general play-it-your-way appeal, it is an issue that each archetype borders on being in a different game entirely. Warriors play a Soulslike, while other classes are playing an action RPG.
Looking to the Horizon
In the Tavern talk interview, Darth summarized the whole affair with a question:
“Would you say it’s a fair statement to say that warrior and mace skills, and that side of the tree in particular, is in a state where it has the worst of all of them. It has the worst defenses in armour. It has the worst effective hp in terms of life and basically the worst offensive capabilities as well? So you could see why I’m feedbacking so hard on this archetype in particular.”
Darth is pretty even handed about the whole affair, but it does speak to a general sense that melee in PoE2, much like in PoE before it, seems best played as least melee possible. So, hopefully the design direction will be to review both the performance axes (especially in terms of speed and defenses), and more so the clarity around the archetype’s mechanics, as this remains the biggest trap for new players. Shifting in his seat as Darth goes through the list, Jonathan began laughing, acknowledging that it's a “brutal savaging, but I suppose I can’t necessarily disagree with any of those assessments.” Aptly enough, warriors using physical skills do generally need to include the “brutality” support gem to boost their damage, so here’s hoping we see equally apt patches.
_______________________________________________________________________________________________________________
About the author
Andrei Zanescu is an Assistant Professor in Communication Studies at Concordia University. His research explores the intersection of Hollywood film and prestige television as cultural institutions and blockbuster game companies operating in the Euro-American context.
1. I want to stress that these thoughts are part of the early access feedback and subject to change as patches come down the pipeline. With that said, I’ve been playing Path of Exile since 2012, and I’ve seen melee run into similar problems in the series’ previous installment, which I’d hate to see happen here. These thoughts are based on playing in early access for roughly 150 hours (with a level 93 Gemling Legionnaire that’s cleared up to +3 bosses, an 85 Corrupting Cry Warbringer, an 81 Spiral Volley Witchhunter, a 75 Minion Titan, a 55 Crossbow Deadeye and a 71 Ice Strike Acolyte of Chayula). There are certainly folks who’ve put in more time in this early access though, so take my thoughts as one person’s perspective.
References
Bourdieu, Pierre. 1977. Outline of a Theory of Practice. Trans. Richard Nice. Cambridge University Press.
cArn_. 2024. Path of Exile 2 - Amazing Endgame Titan Earthshatter Build Update.
fuiwwjdasj. 2025. POE2 CI Deadeye Howa Herald of Ice Lightning Crossbow T16 breach & boss SSFHC.
GhazzyTV. 2025. The New PoE 2 Patch EXPLAINED! - Podcast ft. Jonathan & Mark /w @dmdiablo4.
Grinding Gear Games. 2024. Path of Exile 2. 0.1 Version. PC
Grinding Gear Games. 2013. Path of Exile.
KallTorak. 2025. 1.5 Million Damage Magma Barrier | 1 Shotting Bosses and Good Mobility! | KallTorak Path of Exile 2.
Kripparian. 2025. THE ARMOR SAGA CONTINUES! (PART 3) - Path of Exile 2.
mas0ny1. 2025. Path of Exile 2 - Str stack Sunder Totem Gemling Herald of Ice Abuser.
Mathil1. 2025. Mathil on Warrior/Monk disparity.
Mathilification. 2025. I Played Thorns Warrior So You Wouldn't Have To And The Results Are... A Mixed Bag.
TheDoppelGanker. 2025. Path of Exile 2 - Armor stacking Magma Barrier Warbringer guide.