Deus Ex: Invisible War Is a Deeply Flawed Game.comments powered by Disqus
Posted on Wednesday, March 6 2013 @ 15:10:10 Eastern
Compared to its predecessor, Deus Ex, it is inferior in every way save one arguable exception, and that came at a terrible price. The following is a disordered rambling attempt to show how and why broken down by particular subfield.
Feel is a highly subjective thing. It speaks more to the reviewer than to the thing under review. Ultimately it is an expression of how all the individual elements come together. A game can have good art and sound design, mechanics, good writing and story setup and they can all be completely wrong for each other in such a way that the experience is something less than the sum of its parts. The opposite can also occur. Bioshock and Bioshock 2 have excellent art design but nothing else in those games on their own make for a real standout presentation. However, they relate to each other well enough to produce something better than their individual bits. Deus Ex: Invisible War falls flat harder here than in any of the weaknesses of its subsystems. It feels like a prototype for everything wrong with the seventh generation. Oversimplification, removal of core challenges, handholding, elevation of graphics over gameplay, poor art direction, a story that makes up for its lack of integrity or coherence by clumsily importing themes ill-attached to the plot at large.
Independent of any expectations from the era Invisible War looks surprisingly deep and well-prepared considering the general tone of commentary surrounding the game. Hyperbole is to be expected; Deus Ex is one hell of a perch from which to fall. It is unlikely anyone will play it unpatched. It is also unlikely anyone interested in it will come to it without having first played Deus Ex or Human Revolution. In hindsight Invisible War was a landmark title of the era and that is not a good thing. Those years were a major wave of what RPG Codex members call 'the decline' and someone coming in from a point of view just a few years prior would find Invisible War at best underwhelming and at worst a total disappointment in view of its more prestigious ancestor. The systems are simplified to a fault and do not work as well as their allegedly clunkier and 'tweakier' (to use Spector's term) forebears. The skill system and experience points were entirely removed. Lockpicking and electronics were joined together and there was no way to improve the efficiency of the multitool which now filled both functions. The augmentation system was heavily simplified and the consequences of poor decision-making practically nonexistent due to ubiquitous resources for changing course. There are three possible choices for each of six slots, for a total of 18 augmentations. The upgrades came in the form of canisters which were single use. Unlike the first game where there were specific canisters for each socket, and upgrade canisters were a separate matter altogether, in Invisible War canisters which unlock abilities are also used to upgrade them. The system is horribly imbalanced and there are few tough choices to make. The only other detail of note in the IW aug system is the presence of 'non-standard' upgrades requiring special canisters to unlock which make up 1/3 of the choices. They are often game-breakingly overpowered. One such mod is the Neural Interface which, when fully upgraded, allows the user to enter any locked computer or ATM in just a couple seconds. The old system of passwords and logins, key codes and the like has been jettisoned, which deprives an experienced player of the freedom to break sequences. It also compromises immersion. The interface was nowhere near as handy as the first one. The original included a Diablo-style inventory management and provided for rapid access to vast quantities of information. The system in the sequel was heavily simplified (every item took up exactly one inventory slot, all weapons drew from a common ammo pool, numerous active abilities in the first Deus Ex were now made passive to streamline the experience) and combat was also much weaker overall. AI was generally less intelligent than previously seen and enemies were rarely anywhere near as threatening. In the first game locational damage applied to the player as well as enemies. It was possible for the player to break their legs, arms, torso and head. Break an arm and one will not be able to shoot properly, break a leg and they hobble, break two and they crawl. At first it could be a little overwhelming but coming to grips with it was a rewarding experience.
While there is no question that the effective use of the Unreal 2 toolset produced visuals which were very good for the time* some of the artistic choices made along with the unquestionable crafstmanship and technical proficiency of the project are downright strange. The User Interface (which should perhaps be discussed under feel or mechanics) was, at launch, intrusive, unintuitive and highly unpleasant. Its elements were fully opaque and crammed into the center of the screen at all times. These pieces of information included a curved bar on one side for equipment, another for ammunition, another for augmentations and two more bars on the top corners of the screen for health and bioelectrical energy. All told a considerable portion of the most important screen real-estate was occupied by information only needed some of the time. After a patch (which also fixed pernicious crash-to-desktop bugs and helped with OS incompatibilities) fixed the issues with the UI the experience was overall much more approachable. However, some other choices were never fixed in an official capacity. One such choice was the male Alex D face textures being some of the lowest resolution and worst looking in the entire game. Due to the way dialogue works the player will be looking at those visibly lower resolution textures for a hefty portion of the experience. Other concerns include effects which look very nice, such as the multi-tool animation, which by their nature and style draw heavy attention to themselves and cast themselves apart from the game world.
Instead of canonizing one of the three endings from Deus Ex (Tong, Illuminati, Helios) they decided to play off all three. For a player of the original the story doesn't make much sense and it's turned the original story into a pot-luck roadkill casserole. The conceptual setup for each of the three endings was deliberately portrayed as mutually exclusive to the other two. In order to reconcile them the entire ending sequence of the first game is retconned. As a sequel, then, it doesn't really work. It doesn't build on the themes, events and characters of the first game so much as it reimagines and transforms them outright. Characters' motivations for actions are cast in new lights and sometimes their behavior is completely reworked. A single datapoint can be drawn up to illustrate just how cavalier they were with the lore they themselves had created. The Dragon's Tooth, a straight-bladed Jian-style sword made in Hong Kong by a Chinese technology firm at the behest of one of the triads, was a major Macguffin (and the best melee weapon) in the first game. Its appearance is completely different in the sequel - it's now a Katana identical to any other standard energy sword seen elsewhere in the game. Due to poor balancing and implementation it is also inferior to a common riot baton.
At launch the PC version was a poorly documented mess. It required a video card which supported a certain version of the Shader toolset. Support for a given version of Shader at the time was, like Hardware Transformation and Lighting, not something well-documented or advertised with video cards so buying the game (or buying hardware upgrades to play it) was something of a gamble if one did not know how to dig up documentation prior to purchase. One could not simultaneously have Bloom and anti-aliasing enabled or the game would crash. Certain effects in Invisible War would not work properly with Bloom enabled until the first or second patch came out. The UI was seemingly poorly designed but with a couple menu tweaks and a single line edit in the user.ini file it actually worked just fine. A patch made those changes the default values. While most users with the target hardware would have been running Windows XP by the time the game came out it was not properly compatible with XP and the only way to run the game at all was in compatibility mode for Windows 2000. After the last official patch the game was perfectly stable (hardware and driver conflicts notwithstanding, but that stuff is largely out of a developer's hands) and all of its features worked as intended. It could finally be evaluated as it deserved to be. It was markedly inferior to the game which preceded it at that point. Human Revolution would exceed it in every single way.
The Lone (Arguable) Exception
The one area where Invisible War consistently outshines its predecessor is in visuals. While questions of art design and UI implementation are a separate matter the developers made good use of their software. Due to hardware limitations or the quirks of Unreal Engine 2** level design was heavily constrained. Deus Ex's large, open, seamless levels are replaced with smaller levels joined together by convenient bottlenecks. Everything in those levels sports a (for the time) generous amount of detail with high-resolution textures and high polygon counts aided by a new lighting system. While modern computers could easily brute-force several levels joined together (the Detroit outer area in Human Revolution is larger and more complex than all the load zones in Invisible War's Seattle areas combined) at higher detail than seen in the game at the time, for whatever reason, such an offering was not in the cards. With the costs in view for the visuals, the weaknesses of the other parts of the game come full force. The most obvious weakness after completing a playthrough (which the expense of creating content at the desired level of fidelity certainly exacerbated) is that the game is much shorter than the first one. To some extent the developers bear little blame for this. In 1999 System Shock 2 reminded people there were more ways to do a first person shooter than Doom. Reviews whined about the low-detail visuals stemming from the limitations of the Dark Engine. In 2007 Bioshock heavily simplified or outright eviscerated everything that made System Shock special and introduced a much-diluted version of the experience. Six years running and people still give it praise it does not deserve. Even the seemingly intellectual set are taken with visual bells and whistles.
With each system generally weaker or just not used as well as in the first game and the plot a mess it is far and away a lesser achievement compared with its predecessor. Now it's time to talk about why. First there's an obvious sense of not knowing their audience. The people who played Deus Ex and responded well to it were very often the same people singing the praises of System Shock 1 & 2, Cybermage, Ultima Underworld and the S.T.A.L.K.E.R. trilogy. They would find Human Revolution a better follow up to Deus Ex than Invisible War. It is fair to say that with all the simplifications and strippings down of systems they were trying to target a new audience, one that was not very good at video games in general or first person shooters in particular, and this mythical creature also wanted lengthy meditations on political philosophy front and center. In other words someone who would rather watch and talk about games than play them. To the extent that such an audience exists*** they already had Metal Gear Solid. Part of the problem is that Warren Spector was not actually directly in charge of development. At one point Harvey Smith, who would later go on to create Dishonored, considered eliding an inventory system outright. Thankfully Spector vetoed that idea. Other problems seem to be a function of limited budgets grappling with higher production costs. Finally, many of the more irritating choices which also caused clash with the earlier game were the result of focus testing aspects of the game prior to release. Alex's appearance and style of dress, the time of the setting, and many other details seem to have been the result of that kind of feedback. Focus testing is also responsible for the first US X-Box controller, the 'kill everyone' ending of I Am Legend, and the bad SEGA Genesis controller.
*It made particularly good use of UE2's DirectX 8 features. It contained detailed dynamic lighting and shadow casting before Doom 3 hit the market and the Havok middleware worked seamlessly with it. Overall, the visuals helped work with the sense of tactility they were trying to create. Since this is a largely negative article I append the following complaint: that tactility was mostly a gimmick and accomplished almost nothing at what must have been severe financial expense. These same features would prove far more valuable to Thief: Deadly Shadows.
**Pseudointellectuals on either side of the debate have poisoned this particular well. If it really was Unreal 2 that forced their hand in the direction it did then how can one account for Unreal Tournament 2004? Its Onslaught maps are large enough to house a considerable percentage of the entire Deus Ex: Invisible War world all at once. If it was hardware, again, UT2004 shouldn't have come out around the same time with good performance numbers on mid-range PCs. A more likely explanation, to my understanding, is that if they wanted to make changes to one version it would have been easier to change both, and as these games were among the last made in the era when quality assurance and post-launch support were handled by developers or at least at their discretion it would not have been feasible to maintain two separate SKUs after launch. A similar problem may have affected Thief: Deadly Shadows and in the end ISA closed not long after T:DS shipped.
***There is some evidence it does. Spec Ops: The Line, despite playing Abel Farerra to Invisible War's Martin Scorsese (New York New York Scorsese, not Taxi Driver Scorsese, this is a neg post, remember?) didn't completely crash and burn and apparently the publisher only lost a small amount of money on the game if any.