(wrote during some lectures but too busy 2 refine. read if u like, just some nerd feedback on boss/level design)
Although I consider PN2 to be mechanically solid for the most part, I find several parts of the game's campaign problematic. As a significant part of the Project Nexus lore most players' first introduction to the game, it's important for it to catch the player's attention and introduce them to as many of the game's mechanics as possible. Boss fights are an important and oft-utilized part of this sequence structure: They serve as setpieces that provide a unique and memorable challenge to the player, while simultaneously wrapping up built-up narrative tension, or sometimes introducing more for future gameplay. Boss levels are meant to be difficult and unique, yet achievable through what the player has learned thus far.
PN2's boss fights are usually enjoyably written and beautifully modeled, but several of them fall into the territory of becoming frustrating, unfair, and ultimately forgettable. I'd like to go over the different elements of boss design that stand out to me the most and reflect on how PN2 could achieve these goals better.
NARRATIVE - Buildup, contrivance, resolution
A sudden escalation in difficulty typically mandates some sort of excuse be made for it: An alarm has been set off, you're infiltrating a stronghold, there's a particularly difficult enemy approaching. Boss fights typically happen as a narrative climax to rising tension, and their resolution (or the act of beating the boss) acts as a way to reset tension, potentially on multiple levels of granularity. PN2 is not particularly concerned about the quality of its writing, but narratives aren't just used in artsy or moody games: Any sort of premise feeds into narrative, senseless violence non-withstanding. Players don't need context given explicitly through writing to understand what's happening, and quite often it can be more effective to show instead of tell.
The Ghoul boss fight is the prime example of what happens when a boss is thrown in without any consideration of narrative implication: It literally drops out of the sky with zero warning, you fight it once, and it drops down the elevator shaft just as unceremoniously as it came. The boss is really quite pretty and the fight is alright, but the lack of any context makes it exceptionally forgettable - It takes the minimum amount of screentime possible and no questions are asked about it nor raised. It's only ever perceived to be as much of a threat as it actually is because no steps are ever taken to build tension or suspense around it, and all things considered it was a rather easy fight. Gil offends in a similar manner, as he's treated like a god in the 30 seconds leading up to his fight and promptly forgotten after a couple of shotgun blasts. While a fight can simply come up out of nowhere as part of the game's atmosphere, most of PN2's bosses aren't presented as such (or aren't presented well enough in that fashion to make the intention apparent) and as a result these instances stick out.
While just having a boss pop out of nowhere is probably one of the worst ways to execute such a sequence, care should be taken not to end up as hamfisted and tactless as the other end of the storytelling dichotomy. Build-up through world-building lore or small touches and sequences that ramp up over time are subtle means that PN2 has utilized. We see the former done with Church & Jorge's first appearance and with Crackpot - The lore contextualizes the bosses in a tone that juxtaposes the absurdity of their character with the gravitas of their challenge (or vice versa) and contributes to the tone of madness. On the other hand, Gestalt is introduced by having us get the heck out of dodge, and his presence as an unkillable threat means there are actual gameplay ramifications to back up the premise that he's virtually invincible. The Black Knight also achieves good narrative buildup with the presence of lesser tower guards in conjunction with all of the previous hype surrounding the science tower, and all the justification he needs is him being batshit insane.
Imagine this: Maybe the level prior, you hear that rabid screech once in a while. You could see its red eyes glowing in the dark momentarily, or perhaps you could catch it flitting over a balcony or crawling up a wall. That last enemy on the other side of the map that was chucking grenades as you crossed the bridge, suddenly dragged off into the dark the moment you put your sights on him. And right as you're about to leave the mining sector, it decides it's stalked you enough to make you prey for sure.
or it could just drop from the fucking ceiling
FAMILIARITY - Uniqueness, precedence, conveyance
Bosses can distinguish themselves narratively, but there must be some sort of meat behind the gameplay to really justify them. Just increasing enemy health and damage works, but those approaches alone often make them more tedious than exciting due to the typically spongy and unforgiving result. The most compelling fights are those that force you to tackle something in a new way, but at the same time bosses should still test players on skills relevant to normal gameplay. If a fight demands previously irrelevant gameplay, it either strips the player of agency or becomes extremely frustrating: A fight with overt tutorialization makes the player feel led-on, but a fight with no tutorialization will often lead to success only by trial-and-error. PN2's boss fights understand this for the most part, but there are several instances where the game simply doesn't do a great job of teaching the player what to do. How can we quickly and effectively indicate what options the player has available, without making our fights stale, confusing, or overbearing?
Maybe the most effective means of conveyance are the most basic: If it's interactive, just paint it in a bright color or make it emissive. This approach works especially well given the mostly monochrome palette. Other more subtle methods of guidance include camera shifts as players near objectives (situational, as it can throw off aim) or clever environmental design, like arrow-reminiscent patterns or geometry that siphons the player towards their goal. And if all else fails, the designer should do their best to ensure everyone will eventually figure out what to do with the least amount of frustration: Character barks are a more benign way to achieve this than arrows or glowing items.
However, if we teach the player what to do prior to raising the stakes they've got plenty of time to figure things out. Simply having the same mechanics appear earlier works well, and if we can demonstrate it to the player then that's even better. For instance, there's already lots of crates and moving parts in the Sheriff level, so why not have the player drop a crate at any point prior? You could even have an enemy try and drop one on the player, that'd REALLY catch their attention. Other fights that could use some level of precedence include the Church & Jorge mining sector fight that introduces Church's homing fists, the Crackpot fight that has projectiles that home and stun, and the fanatic boss wave, as there is zero tutorialization for the fact that Sanford can catch weapons out of the air.
Being difficult isn't necessarily a bad thing, but there are multiple ways in which something can be difficult. Making a fight confusing or unintuitive will cause players to take a longer amount of time on it, but having them retrace the same steps for the wrong reasons is a great way to dis-incentivize them from ever beating it.
APPROACH - Setpieces, opportunities, experimentation
We've hyped up our boss, we've made sure the player knows what to do. Now we just need to make sure the player will actually do what we want them to - easier said than done.
Given a single highly effective approach, most players will exploit it as much as possible to win, often instead of making things harder for themselves or following prescribed steps. Have to fight a large block of corpus? Grab a shotgun, keep it on you until something meaty comes up and blow your troubles away in a shower of buckshot. Tackling this doesn't require that we retroactively reduce the player's options in the rest of the level: It can be as simple as coming up with an excuse for robbing the player's weapons before-hand, or that we require an entirely different means of defeating the boss. With that said, for an encounter as noticable as a boss fight players will be aware of when their options are arbitrarily restricted. With this in mind, how can we design fights in a way that allows players to approach them as we intended, without making it overly evident that they're following a carefully choreographed sequence?
For certain bosses there's really no way at all to hide the drastic change in gameplay, but we can still take steps to reduce the perceived drop in agency players may feel when told to beat an encounter via setpiece. Allowing players to experiment and learn to beat the boss without being directly told how to is generally the best means of achieving this, provided there are safety nets for the lowest common denominator and that relevant mechanics are made intuitive. The Black Knight does a good job of showing the player what to do and being decently compelling, but determining the order in which player takes the steps to beat it (Sanford, then Deimos, then Hank) is ultimately arbitrary: A good way to make the fight seem more free-form while having the player take the same steps would be to let the player perform the relative actions in any order, just context-sensitive to the current character. Gil is another fight that falls short in this manner: The immediate threat discourages us from ever trying to figure out what the meat patties are for, and the convenient 8 shells of buckshot at hand provide a much more appealing option.
If we let the player roam free, we ought to make sure they can eventually figure out what to do, or that we show them what to do if that never comes to happen. Giving the player room to learn can be rather difficult in tandem with the elevated difficulty of a boss fight, which is why I stressed previewing or demonstrating relevant mechanics earlier in the level. If that's not a possibility or the trend becomes too noticeable/formulaic, creating different stages in the boss is a good way to give the player some breathing room to learn while retaining the intended climax in difficulty.
However the player is meant to defeat the boss, they should be given the means to figure out how to without being hand-held (or without feeling that they're having their hand held) whilst ensuring the boss remains beatable for the vast majority of players. For a climactic, narratively vital segment, the worst thing we can do is take the player out of the fight or render them unable to resolve it.
As always, I will note that these are my opinions and not necessarily fact: My judgements on the bosses are drawn from my experiences with them and my knowledge of game design, I know other people see certain bosses are less/more problematic (ik gangrene hates the sewer fanatic spam idk) so other people's thoughts on them would be much appreciated. With that said, I do happen to know more about games than ur avg joe. I think PN2 has really cool bosses, but the design aspects around several of them just need a lot of work for them to shine.
i dont know how 2 end this lol. i can provide more examples/suggestions for these tenets if u think this any amount of made sense