Hades II Has a Legibility Problem

This post contains mechanical spoilers and boss names for Hades II.

I wasn't just hoping to like Hades II, I was fully expecting to. I generally think there are two possible failure states to a sequel to a game like Hades where much of the fun is encounter and build design that can be expanded on almost indefinitely: either it turns out to be more of the same, which may be a bit uninspiring but can't be too bad if I liked the original (think Dark Souls III); or it takes the game in a wildly different direction that doesn't quite work, which may not be as fun as such but is almost always interesting (think Tears of the Kingdom). And at first it looked like Hades II was going to be more of the same, and I was content with that. But the more I played, the more it seemed to have an issue at the heart of its combat that just wasn't there in the original.

The game has a legibility problem, and not just in terms of the density and chaos of on-screen interactions (although that certainly doesn't help). This problem is structural: it pervades the game's design sensibilities. Whether it's a consequence of the game's wildly popular early access period or a result of the studio's anxiety about their first-ever sequel to their only massive hit, it pushes the player into a mode of play that's less fun and more frustrating. Legibility isn't just about being able to keep track of what's going on, it has far-reaching consequences for what types of play are safe in practice and thus which builds are viable. In a game whose primary mechanical driver is the joy of assembling new and interesting builds, that's critical.

Some of this springs from the new boss design philosophy in Hades II. In the original, bosses were generally built around melee attacks that did damage in some hitbox around the boss and ranged attacks that flew outwards from the boss. There were occasional situations—most notably when fighting Theseus after vanquishing Asterius—where large AOEs^[1] appeared that the player had to avoid, but these were very much the exception. The result is fights that are focused on the player establishing a rhythm where they dive in, hit the boss, position themselves to evade the boss's next attack, and repeat.

With the exception of some minibosses and Hecate (the first true boss), Hades II doesn't work this way. Bosses in the sequel are much more focused on two types of attacks that rarely appeared in the original: AOEs that are telegraphed by regions of the arena turning red usually independently of the boss's positions, and "waves" of damage that the player is expected to avoid using the invulnerability frames. When melee attacks do appear they tend to come out very quickly and often have their own AOEs, while ranged attacks come in one of two flavors: they're either very fast (often coming out of something that looks like a melee attack), or they're heavily telegraphed while also tracking the player's position relative to the boss.

I think each of these attack archetypes have serious issues, and I'll address several of them each as part of case studies of particular bosses below. First, though, I want to discuss the net effect of all of them in concert. Hades II is still about positioning in a certain sense, but in a much more global sense: the player is constrained by the areas the bosses allow them to exist in and the time the arena is clear. Quick melee attacks that hit all around the boss and tight boss tracking means the player's position relative to where the boss is facing is much less relevant in comparison to their distance from the boss. Arena-wide AOEs force the player far away, and damage waves require the player's dodge to be off cooldown.

When it's rarely safe to attack from nearby, ranged attacks end up structurally advantaged. The best place to be for nearly every boss is the middle distance: close enough to hit them with a longer-ranged attack but far enough away to avoid the radius of their quick melee hits and have time to dodge through any wave attacks they send your way. Only the strongest melee builds are viable, those that do enough damage to make the exchange rate on taking hits from the boss profitable; it's just not realistic to plan to do damage up close and avoid damage reliably. Of course, there are a sequence of inputs the player could do to survive, but the legibility isn't there to prompt the player room to make those inputs without tons of memorization and close attention to what the boss is doing^[2]. To understand this in more detail, let's look at some case studies.

Case Study: Cerberus

I like this fight on the whole, but that's because the mistakes it makes are gentle enough that I can trick myself into thinking it's a classic Hades-style fight about carefully keeping behind the enemy as he moves around and giving distance during critical sections. But that strategy only works because the things that punish it are mostly relatively low-damage and mostly by the time I finish the mourning fields my damage output is high enough that enduring those hits is viable. It does not forgive the sins of the design.

There are complaints to be made the half-dozen distinct AOE animations this fight employs or the fact that Cerberus's lunge attack shoots a projectile for no discernable reason, but by far the biggest issue is the way the attacks where he slams his paws into the ground work. There are two variants of this attack: in one, he lifts one paw high in the air and slams it down, creating a shockwave that goes all around him in a big circle. It's odd that this attack is specifically telegraphed by one side of the boss hits well past the other side, but it's a heavy-looking attack with a long windup by Hades II standards so it's not beyond the pale. In the other attack, he lifts both paws much lower and more briefly before bringing them down. This creates two smaller shockwaves as you'd expect, but those shockwaves still reach fully behind Cerberus and so will effectively hit the player if they're standing close to him in any orientation.

Because this attack comes out so quickly—faster than the recovery time of some of the slower points in weapon combos—it means that if you're regularly attacking the boss from directly behind you're just going to take damage from it sometimes. The solution? Tank the damage, avoid combos entirely, or rely exclusively on ranged attacks for your damage output.

Case Study: Polyphemus

I think this may be outright the worst fight in the game. It's not tremendously difficult, but the difficulty that is there is all in mechanics that encourage the player to memorize the fight and play conservatively (which is to say, not very fun). In that way, it's kind of a microcosm of the game's combat design as a whole!

This fight has, by my count, five distinct animations indicating waves of damage flowing outwards from the boss. The bread and butter one comes when his jump attack lands, where a bright red circle expands out from the landing spot—but the circle is only visible after a second or so, because it's intentionally obscured by the cloud of dust as he lands, laying a trap for players who aren't used to the fight yet or anyone focusing on the adds he summons and trusting their peripheral vision to alert them of threats. There are also smaller versions of this circle that appear overlapping one another as he walks. These are distinct from the small circle he creates when he punches the ground, though!

There's also a very different animation that happens when he throws a rock at the ground. This creates multiple thicker red projectiles which fly off in all directions in several layers, one behind another, so that the player can't easily dodge through them like they can the slower circle. Why a rock thrown at the ground would diegetically create such a dramatically different effect than a cyclops landing, let alone produce two distinct waves, are clearly not questions the designers are interested in, but these questions are important for a player to be able to easily pick up on what's happening.

The fifth and final animation is a series of circles that expand outward from him in a semicircle after stomps the ground and explode upward after a few hundred milliseconds. This attack may be his most vexing. It's diegetically incoherent again (if punching the ground creates a small symmetrical effect why does stomping it produce a massive asymmetrical effect?^[3]), it covers a full 180 degrees around the boss so moving to the side doesn't work even if the player has internalized that it's a directional attack (so why make it directional in the first place?), and the circular pattern with a width that's a significant proportion of the player's dodge distance makes dashing over it (the only real possibility for evasion) finicky. It's only made worse by the fact that the coloration before the explosion is dim, making it simply hard to distinguish visually on top of everything else, particularly in the Rivals version of the fight.

Any study of Polyphemus would be incomplete without mentioning that his jump prior to landing and producing the first wave I described above also, completely inexplicably, does damage if you touch him at all while he's in the air. The game otherwise never has contact damage for enemies—only their attacks deal damage. The result is that any attempt to stand anywhere near him during this attack is going to be punished, particularly because he can turn nearly full-circle in less than a second and jump directly behind him.

Again, this doesn't mean the fight is all that hard. Once you know how to dodge the attacks, most of them (other than the stomp) are pretty easy to avoid. It just requires playing it extremely safe and largely avoiding any attack that doesn't have good range. And if your build after Ephyra is focused on melee attacks, especially repeated flurries? You're in for a long, dreary boss.

Case Study: Melinoë

Before I wrap up, I want to talk a bit about the way Melinoë's kit itself contributes to the game's legibility issues. Legibility is about the player being able to form an accurate mental model of what will happen in the game, and part of what happens is always going to be their own actions. It's even, by this definition, about the build management aspect—whether players accurately predict how different boons or items will interact. And there are numerous places in the game where there are just arbitrary limitations-by-fiat to keep players' builds from becoming too strong. I think this is almost always a mistake in build-oriented games like this. When much of the appeal is specifically trying to find combinations of abilities that are crushingly powerful, running into a place where that power is curtailed in a way that feels artificial feels like getting your shiny new toy taken away. If a player manages to make a build that can push massive Scorch buildup if they can get in successive hits (already not a trivial feat!), seeing it cap out at 999 is a real letdown.

Some of her weapons suffer from the same kind of "no fun allowed" moments. In particular, the Black Coat and the Aspect of Artemis sword and sickle both have the ability to block incoming damage while charging their omega attack. But some sources of damage—again, just by fiat—can't be blocked. Not because they're ambient like fire or coming from the wrong direction, just because these particular attacks are flagged as unblockable. You even see a little "UNBLOCKABLE" notification pop up in the game! There is of course no way to tell which attacks these will be until you try and fail to block them, so it's just another thing a player has to memorize if they want to use those tools effectively.

Maybe the most frustrating legibility issue is actually part of Melinoë's universal loadout. The cast^[4]'s primary purpose, the only thing it does by default when not omega-charged, is binding enemies in place. But whether it actually binds them and to what extent is completely arbitrary, varying from enemy to enemy and in some cases even from move to move (the miniboss Phantom will move much slower in a binding circle during its slash attacks, but will ignore it entirely otherwise). If a player wants to plan its deployment effectively the only solution is once again to memorize which enemies get stopped, which get slowed, and which ignore it entirely and recall this in the heat of battle. And even with that done, there's no getting around the sheer misery of the first time you lay down a binding circle and an enemy walks right through it and whacks you in the face.

What's Legibility For?

Despite finding these issues tremendously frustrating when first making my way through the game, still running into them with some regularity as I learn the Rivals versions of the boss fights, and having numerous other resounding complaints about the game^[5], I'm still playing Hades II and by and large enjoying it. Legibility is, in a sense, a beginner's resource: it's there to help players understand what's going on before they have a lot of iterations, not so much to help them succeed as to allow them to feel like their failures can be understood and fixed, and so to encourage them to brush themselves off and try again. But I tried again regardless, and now that I have a lot of iterations under my belt, most of the issues I was running into early are consigned to memory.

So am I making much ado about nothing? If legibility is a problem that solves itself as the player continues playing, and Hades II's build management is compelling enough to get me—someone the friends who have listened to me rant would probably call "a Hades II hater"—to push past these issues, is this a design issue at all?

It is. What's going on here isn't just a problem that's limited to the game's first thirty hours; that's just where the symptoms are the most visible. The game's legibility issues don't exist in isolation. They weren't created because of a sequence of poor design decisions that coincidentally happened to point in the same direction. On the contrary, they serve an underlying design goal, albeit one that evinces a profound anxiety in the designers.

That goal is to slow you, the player, down. We can see this that this is a design concern very directly by looking at the Eris mechanic. Not the boss fight^[6], but a mechanic most players likely never saw: if you start making too much progress on your path to the House of Hades too early, Eris will show up and curse you, making enemies do more and more damage every room and making a "clean save" run if not impossible then nearly that. No fun allowed!

I suspect the ham-fisted nature of this approach is the reason it only appears in runs that are truly dramatically early (I don't know the number, but I'm pretty sure it stops appearing within the first ten). But the goal is still there: the designers want to make sure the player engages in all their precious dialog and meta-progression. They want Hades II to feel like a bigger game than the original, and they apparently don't have confidence in the depth of their build options or challenge runs to bring that about organically. This despite the fact that Heat was such a brilliant challenge design it was adopted by all sorts of run-based games in the five years since Hades!

In a sense, dialing down legibility solves this issue perfectly. It creates barriers to progression that will throw players back to the crossroads a few times, but because these barriers are largely solvable by sheer memorization, players can progress once they've faced a boss a few times. I'm sure the graph of average player progression per run looks beautiful on someone's office wall.

Unfortunately, these designs come with real costs. Fights become feast or famine: when so much of them is raw memorization, they fall over to players who have memorized them. And when many of the illegible attacks specifically (and surprisingly) punish attacking up close, they severely limit entire the game's build diversity, which was the core strength of Hades and remains so in Hades II. The player is forced into two choices: play carefully from range or produce massive damage. There is no other option. That in turn feeds back into the encounter design and produces bosses that are contorted more and more towards making chipping in from range as interesting as possible for the builds where that's viable, and leaves builds where it's not in the lurch.

I can't help but wonder if the game's year-and-a-half-long early access period contributed both to the manifest insecurity of the design and to the specific shape the solutions took. When the vast majority of playtesters are the most devoted players of the original game, it makes sense that the designers would worry about the game being too easy for Hades players to blast through. When you have the same cohort of playtesters for so long, it makes sense that legibility issues would fade into the background. It would take a strong design vision to push past that kind of feedback, and a strong design vision is not among this game's virtues. It's sad for me to see, but... I guess I'll just do one more run.

---

1. "Area of effect" attacks that damage the player if they're in a broad area when the attack activates and does nothing if they're outside that area. ↩︎

2. The Dark Souls archetype of boss design does ask for close attention to what the boss is doing and, at least in the extreme cases, memorization of its windups and timings. I like those games a lot, so what's the big deal here? A run-oriented game puts bosses in a different context than a bonfire-oriented game does. When you're spending 6+ minutes to reach the first boss and 24+ to reach the last one, the pain of losing a run because you just didn't know what was going on (as opposed to not reacting correctly or having poor strategy) is much higher than when it takes less than a minute to retry. Note that the Elden Ring: Nightreign nightlords generally have dramatically more telegraphed attacks than a typical Elden Ring boss to compensate for being in a similar structure. ↩︎

3. There's definitely some use in this game, as in the original Hades, of the Street Fighter conceit of martial attacks producing energy projectiles. By and large, this works fine: a punch or kick in the air, especially one that's visibly charged with energy, can be easily read as a precursor to a ranged attack. The issue arises here where the animations don't do enough to sell the specifics of the attack: a stomp doesn't feel as directional as a kick to the ground would, and expanding circles don't feel as clearly like an outcome of a shockwave as a literal wave would. ↩︎

4. I detest this name, by the way. Hades already had something called "cast" which is far more similar to the starting weapon's special, to the extent that it took me probably ten hours to stop accidentally taking cast boons that I thought would apply to my special. What's more, there's a resource named "magick" in the game and it has nothing to do with the cast (or at least no more than any other move). For the life of me I cannot understand why they didn't call this move "bind", since it's described as creating a binding circle. ↩︎

5. I'll save my bitter disappointment in the writing for my review proper. ↩︎

6. I don't have enough to say on this to do a whole case study. I think the fight is largely pretty solid, with the notable exception of the telegraphed aim lines. The fact that these rotate to face you much faster than the actual attacks that they're telegraphing is deeply misleading and makes the fight feel very unpleasant until you learn to evade right before she starts shooting. I also think it's the only fight I've seen so far that's substantially improved in the Rivals version. ↩︎