by Tom Ingram
The Tower of Latria is the next level in my clockwise run through the game. The first section is the “Prison of Hope”, a clammy dungeon full of wailing undead prisoners. In the outdoor areas, you can see shadowy outlines against the night sky, hinting at something unspeakable above. Inside, there is not much floor space. Each cell block has narrow walkways against three walls and a bottomless pit in the middle. Anyone attempting to escape will have trouble eluding the Cthulhuvian sorcerer-guards.
These creatures are nasty, especially early in the game. Their magical attacks, which they can deploy quickly, will drain most of your health in one hit (except for the one that paralyzes you, opening you up to the other attacks), and at this stage in the game they take quite a few whacks with a sword before succumbing and descending into some slippery nautiloid hell. Getting off the first attack is crucial, and not letting the monster get a word in edgewise even more so. Not being seen is therefore paramount. The Tower of Latria demands stealth. This fact is a little problematic.
It has often been said that the stealth sections in otherwise non-stealthy games never come off well. This is because stealth as a mode of gameplay requires several mechanics in order to work properly. You have to be aware of how alert the enemies are. You have to know, not just graphically, whether the game considers you to be concealed or not.* The level design and other details need to follow suit: you have to provide a challenge without creating a level that is impossible to be stealthy in. Gadgets like Sam Fisher’s sticky cameras are also helpful.
More generally, your dominant mode of gameplay (melee fighting, shooting, stealth, avoidance, etc) is the star around which everything else orbits. Your levels, your mechanics, your character models, everything must serve this central concept. It might actually be impossible to make a good job of switching between modes. Certainly so much would have to change that levels involving different modes would not be recognizably part of the same game. Great games are usually highly unified in this respect: Half-Life is a shooter through and through; the Legend of Zelda games are strictly about hacking and slashing and suffer whenever they try to incorporate anything more than the crudest projectiles.
Even more generally, the class system favoured by many RPGs might be flawed at a basic conceptual level. Demon’s Souls has a bevy of poorly differentiated classes that can all be grouped under familiar headings: warrior, thief, wizard, etc. But here’s the problem: each one implies a different style of gameplay. Even beneath these headings, there is a not inconsiderable amount of difference between playing as a knight and playing as a wanderer (as I did).
I learned this when I saw my brother playing as a knight. The reason that the knight enemies in the Boletarian Palace are so difficult is that they are designed to be fought in a certain combat style: big swords and shields, lots of parrying, long one-on-one fights. If you play as a knight (as the creators evidently assumed you would), then there is no problem. But if, like me, you positively relish the idea of being Fifth Business and choose a less-played class, you’re fighting an enemy that was not designed to be fought by you, and it just feels frustrating and senseless.
This effect is mitigated somewhat in party-based games like the original Final Fantasy. But its classes give only an illusion of variety. Whether or not you provide for them, you will need the ability to fight, heal, and magick. There is some latitude in which specific combination of classes you choose, but by necessity your party is going to function pretty much the same way no matter what choice you make.
Unless the gameplay changes on a pretty basic level depending on the class you’ve chosen (as in the 3D Sonic games, which are bad, but for entirely different reasons), or there’s not much difference between the classes (as in, again, Baldur’s Gate: Dark Alliance), you’re going to have at least one or two classes that are utterly useless. It’s not that they’ll be weak, it’s that their strengths are not tailored toward the game’s difficulties, and their unsuitability will manifest itself at the most frustrating times. Neither condition prevails in Demon’s Souls, and so the farther you get from the knight class, the worse off you’ll be.
This level, so carefully designed to be a stealth level, will ultimately fail because the rest of the game is not designed to accommodate stealth. You will bumble through gracelessly until your character is strong enough to take out the mindflayers (as I later learned they are called) in one or two hits, at which point stealth is irrelevant.
All that said, from a conceptual and visual point of view there is some very good level design in this game. It’s hard to explain the intellectual pleasure of an environment so coherently designed and presented that you can always mentally place yourself within it. The Boletarian Palace level evokes this sense of continuous logical architecture. You feel you could draw a map of it from memory. The Stonefang tunnel is rather more random and chaotic, but then the Tower of Latria is another home run. The first half of the level carefully pulls you toward a certain style of gameplay. The second is a virtuosic feat of capital-W Weird visual design.
The level’s flavour text says something about a queen imprisoned by an insane monk. This is the stuff that could make for a good yarn, but unfortunately it’s nothing more than flavour. Nothing is ever made of it. There are several plots that are never played out except in flavour text. The insertion of actual characters and conflict into the levels** might have been an interesting diversion.
But since this game is supposed to take place after almost everyone is either dead or zombified, we couldn’t have the plots play out in the usual way. There are some interesting ways to handle this. For example, the game might have achieved poignance by including signs that these conflicts played out centuries ago between now long-dead characters (and certain moments near the end of the Stonefang Tunnel approach this), but this was mostly avoided. It feels like there is an overarching plot, but it’s one in which you are neither a participant nor a stakeholder. This makes the outcome difficult to care about, and indeed I never actually beat the final boss because the effort required was disproportionate to the enjoyment I would get from it.
Since most of the level’s first half involves pissing around in the somewhat samey Prison of Hope, it’s relieving once you get to the boss fight in the big stone church-castle. The boss is the “Fool’s Idol”, a sorceress with the ability to duplicate herself and set invisible paralyzing traps on the ground. This is in many ways the most conventional boss fight in the game. The duplicating boss is a common fixture of certain kinds of video games, and while it’s an easy fight to get stuck on, this is more because of its length and tedium than its difficulty. When the Fool’s Idol is defeated, a pair of gargoyles fly down and carry you up to the next section of the level.
* The Splinter Cell games, which have very well-designed stealth mechanics, have the occasional graphical oddity where a dark spot isn’t registered as shadow and a light spot is. This is why the light meter, which tells you infallibly what the game considers hidden and unhidden, is indispensable. [return]
** Not the loading-screen NPCs, who are narratively problematic for reasons I’ll get into later. [return]