Deadly Assault in ‘Zenless Zone Zero’ and the Difficulty of Tuning Difficulty

During the “Challenging Play” panel during the Year of Games symposium, Jon Perry made a comment about an older game of his and Derek Yu’s, Eternal Daughter. While playtesting the game, they had gotten very good at it and therefore were not the best judges of its difficulty. Perry then decided, right before its release, to double the health of one of the bosses to make it harder, a decision he expressed regret about.

This immediately reminded me of Zenless Zone Zero (ZZZ), one of the two gacha games I play. To very quickly gloss the gacha genre, they are resource management games where the player uses resources to get new characters and weapons and to upgrade them. These resources can also be bought with real money, which will matter more later. ZZZ specifically is a hack-and-slash, so the main gameplay is fighting enemies in a 3D environment.

For months now, there has been controversy surrounding ZZZ’s endgame content, the hardest modes in the game, and HP inflation, the raising of enemy health values to make the game more difficult. This is precisely what Perry did in Eternal Daughter, which has led me to further consider ZZZ’s difficulty and how it is an interesting case of difficulty design. Difficulty can be difficult to talk about in the abstract, so, given the connection to Eternal Daughter and the interesting aspects of its design, I want to use ZZZ as a specific test case to speak to some of what the panel discussed.

More specifically, I want to consider how ZZZ generates, or doesn’t generate, “pleasant friction.” The idea is to find a middle ground between being “too easy” and “too difficult” so that there is a challenge, but it is one the player enjoys. That is what tends to make difficulty fun. This will naturally vary among players, but there are general trends that can be discussed.

The main mode I want to consider is Deadly Assault. Every two weeks, there are three bosses to fight who have a massive health bar. It is, for players without significant monetary investment into the game, near impossible to kill a boss. Instead, doing damage awards points, and the player must earn 20,000 points in three minutes (for context, 60,000 points is killing the boss). There is also a 5,000 point performance score based that is unique to each boss, which will be discussed more in depth later. So, the player needs to deplete a fourth to a third of the boss’s health bar depending on their performance score in order to “clear” it.

In case it is helpful to see an example, below is a video showcasing a rotation of bosses.

To start, consider HP inflation, the original inspiration for this discussion. HP inflation provably exists in ZZZ and is by far the cheapest way to generate friction. The boss is getting harder, but not in any way that requires player strategy or skill to overcome. It is just making the characters do less damage arbitrarily. To put this quantitatively, a score that might have been 20K in a previous encounter is now 18K. This feels awful since the decrease in score is no fault of the player’s. Perry acknowledged this as a mistake with Eternal Daughter, and the same holds here. It is unpleasant friction.

For an example of pleasant friction, the recent bosses have been incorporating more engaging and boss-specific mechanics for their performance scores. Previous bosses have had performance scores based on more general mechanics, such as how many dodges or parries the player performed. These are fine, albeit generic, and are often very hard to maximize, so the player has to do more damage to compensate. Bosses would also often include stalling tactics where they were invulnerable, such as Pompey’s circle attack or Typhon Destroyer’s jump. In contrast, the stalling tactics in recent bosses have been significantly improved. The most recently released boss in Deadly Assault, the Defiler, has parry mechanics, but in particular, prescripted moments that the player can be prepared for. There is also a stalling tactic where the player must destroy the sword she plants to create a protective barrier. If executed correctly, which is not particularly difficult to do, the player can guarantee a full performance score every time. 

This is pleasant friction. While the Defiler cannot be attacked during these prescripted moments, engaging with the mechanics awards performance points, meaning that the time is not entirely wasted. It also gives the player tangible gameplay goals, instead of forcing them to idly wait for the boss’s invulernability to end. It is simply more fun.

Likewise, the Miasmic Fiend has a stalling stage where the player lures it to attack in a certain spot and then has to dodge out of the way. This makes the player engage with the game and the boss’s unique mechanics, which creates pleasant friction.

There is certainly room to improve; most bosses still feel relatively similar to each other and further differentiation between bosses would be appreciated (for an example of very unique bosses, see Cuphead). But it is still a step in the right direction.

However, these bosses have also introduced some issues. The Defiler and the Miasmic Fiend are tailored towards Attack and Anomaly characters, respectively. What these mean specifically isn’t important; essentially, the characters deal damage in different ways. Defiler has a 25% reduction in Anomaly build-up, and its sword takes no Anomaly damage. This makes it extremely difficult to beat with Anomaly characters. Meanwhile, Miasmic Fiend takes extra damage from Anomaly.

These essentially just exist to force players to have a team of each type. This has always existed to a degree: all bosses have elemental weakness and resistances, requiring different teams, and some older bosses, like Sacrifice Bringer, have similar team restrictions. Given the nature of the game, there is a reasonable expectation to have different teams to take on different challenges. But in cases like Bringer or the Defiler, their mechanics are overtuned such that it is near impossible for other team types to beat them, rather than just more difficult. It can be fun to play unconventional or unintended teams, but the aggressive nature of these mechanics prevent that. It limits player expression, creativity, or just ability to clear, producing unpleasant friction.

Notably, this doesn’t have to be the case. Unknown Corruption Complex is a boss originally intended for Attack characters since its legs cannot be broken by Anomaly damage. Breaking the legs applies a debuff to it and also grants performance points, making it mechanically vital. Despite this, Anomaly teams can beat him just fine; if anything, it’s one of the easiest bosses. Having mechanics to work around if using unintended teams can still generate pleasant friction, but the overtuning of the mechanics is what makes it unpleasant.

These problems also apply to the buffs the player selects before entering the fight, which recently have shilled hard for the new characters. The buff this week for Rupture characters requires Lucia, the newest character in the game, to activate. No other character is capable of doing so as of the time of this writing. And even if you have Lucia, there are only Rupture two characters can take advantage of the buff, Yixuan and Manato. This “encourages” players to get the new characters, and by extension, spend money. However, it just comes across as obnoxious. Generalist buffs can support a variety of teams instead of trying to force getting and upgrading the newest unit just so that the buff will have an effect. This is a clear example of a choice not made for gameplay or to generate pleasant friction but rather to funnel players to spend money.

This provides a solid overview of the trends in Deadly Assault: the good, the bad, and the ugly. Being a gacha, ZZZ walks a tightrope between pushing the new characters and potentially alienating those who don’t want them or can’t get them. There have been improvements in Deadly Assault: the new boss designs are trending towards more interesting mechanics. But, what once were more general buffs are now overly catered towards the new characters, locking a significant amount of potential damage behind them having them. These are conflicting as a player, as I am happy to see improvements in gameplay but frustrated by the increase in shilling of new characters.

All of this could also change significantly overnight. ZZZ is a live service game, so it’s constantly getting updated. It’s happened before: Decibels, the resource characters generate to use their Ultimates, used to be pooled, so that the whole team shared. In practice, this meant that only the main damage dealer would ever use their Ultimate. That changed in 1.4, and now characters have separate Decibel counts. But then that genuinely necessitated a spike in difficulty to balance around the change, which started to lead into the problems discussed here. ZZZ is always in active conversation with its player base, even as the artistic angel and gacha devil on its shoulders whisper in its ears. 

But, for as much as I will criticize the poor design decisions in their difficulty balancing, I do at least partially understand it. Part of me that longs for the days when Deadly Assault was a real struggle, when it was an accomplishment to beat. When I’d grind Anomaly Butcher for two hours with my Grace/Vivian/Astra team, getting 18K runs before miraculously hitting 22K. This week, Deadly Assault took me four tries over three bosses. Crushing it was cool the first time it happened; I felt powerful and skilled. But it’s starting to feel boring. As Billy Basso mentioned, we “beat” games. We want to win, but winning only feels good when it’s not easy, when there’s pleasant friction. Even acknowledging that it is a conceptually poor way to create friction, there is part of me that wants the bosses to have more HP, just so it can be challenging again. So it can be fun again.

Perry was right when he said he shouldn’t have doubled the HP. But at the same time, I understand why he did it.