An article was recently published on Equity Arcade concerning so-called "Classical Games", and their alleged comeback on modern systems. Equity Arcade is a "bad-ass media channel" focused heavily on the financial side of game development and the gaming industry; there are certainly some interesting pieces written on the site, but this one I particularly disagreed with (the article can be found here though I'll be pulling quotes from it in this post.
The article opens with "Who can forget Half-Life 2’s release; impossible expectations matched with impossible awesomeness led to one of the best games of all time?" So, right off the bat we know what kind of gamer this guy is, which is to say, a meme loving fuck. The deification of Valve and the memetic status of everything from Steam sales, to Gabe Newell, to the Half Life series itself does not lend much credence to its fans. The fact that this game was awarded "Game of The Decade" in 2012 at the Spike Video Game Awards show, which basically just proves my point. Still, choosing Half Life 2 as the exemplary "classical game" is somewhat helpful, as it was the first game to be developed in the Source Game Engine, a point I'll get to later on.
You get a few paragraphs about how Valve won't make a sequel because it might not be as good and how this apparently what all developers think, disregarding the fact that just about any remotely successful series gets at least one or two sequels in today's market. Now, there's definitely some merit to re-glorifying older games: "Evolving operating systems, hardware, and online services have made them more difficult to be experienced by their loyal followers or reaching a new generation." This is completely true: games don't have a reliable and effective archival system in place the way other mediums do. But these aren't the words of the writer, rather a Blizzard job posting, concerning their desire to "re-glorify" older games for new players.
He talks a bit about emulators, but ultimately the message of this article is split between lines; I'll try to condense it here:
[T]his might signal a new trend in the industry. If Blizzard is updating their games, who else might be doing it? Perhaps companies see a new value in this market, and will go on to invest in it.
...
Resurrecting forgotten (or plain outdated) games from the dead is nothing new.
...
[W]e are talking about something much bigger. This is not simply porting a game from an outdated console to Windows 10. What we’re seeing is potentially additions to the existing platform, leaving everything as is, except for the new goodies. Sure, graphics might need to be overhauled, and some mechanics are outdated, but this won’t be an emulator by any means. A full-fledged game that awakens an old-time favorite, nothing could go wrong here.
...
While surely moves like these would hype the world, the risk is low, and the pay could be high.
Fuuuuuck me. If remakes were so low-risk, then why has it taken this many damned years for a Final Fantasy VII to even be announced? The fact of the matter is, games are not just more expensive to make today, they are just straight up more difficult to make. FFVII for example was a lot of super low-poly models on pre-rendered static backgrounds with no real 3D environments aside from like, the submarine section maybe. Basically none of that is usable in a remake, you can't just "tighten up the graphics on level three" and call it a day; basically the entire game aside from it's dialogue is having to be made from scratch, on a modern 3D engine. And that basically brings me to the bigger point, concerning the Source Game Engine, and modern 3D game engines in general.
On one hand, we have amazingly robust tools available for anyone to use, in the form of software like Unity 3D, and yet none of the software we use today is as effective as the older game engines at prototyping and design iterating. We've all seen this one before:
Loading Image
3D level design has changed, sure, but so have the tools: no one is making story maps in a Call of Duty map maker the way they used to in Star Craft, Half Life, or Quake. Older 3D games were often made using geometric modularity, meaning you could stretch objects this way and that, rotate, resize, retexture etc and incur very few problems. If a hallways seemed to narrow, move one wall to the side and stretch the floor and ceiling to match, iterating was just simple. But what if you're making a modern game and encounter the same issue? You could move the way, sure, but when you try to make everything else fit, then what? Textures don't scale properly and now you have to redo that in the appropriate dimensions, open up another program or communicate with someone in another department. Before art even comes into play, consider the rendering methods utilized by the Half Life Games, that is, Binary Space Partitioning or BSP. The level of autonomy BSP allows level designers is incredible when compared to the modern methods, namely Modular Design. modular design is ineffective in the workflow, but more importantly it requires designers to become dependent on artists or become artists themselves, neither of which are good or constructive and both of which have many downsides. The lower fidelity of modern level design platforms is due in no small part to the demands of the modern market, including the market for "remakes of classics". Especially when older games were as long in hours as Final Fantasy VII, the mere notion of remaking a game using modern level design software is nothing short of nightmarish.
This may work for Blizzard, sure, in that the engines they built Star Craft, War Craft, and Diablo II on are all still around and modernized at that: there are already sequels to all of those games, still being updated and developed, with the exception of War Craft which just lives on its spiritual successor anyway (pull in some 3D assets from WoW, pop them into the SC2 engine, and you're off to a good start for a proper War Craft sequel). The dissonance between level designers and artists which becomes more problematic as time goes in development and more art is finalized thereby restricting the number of minor changes to levels a designer may wish to make is no worse than that between level designers and programmers, and programmers are the ones making the new level design tools, so, it's no secret why this is such a problem. Minor issues can disrupt the fragile flow like if a wall is 90 units high but all the ladder meshes are multiples of eight.
To suggest that the high payout of remakes justifies the development time and makes them low risk projects sounds completely insane because it's an assumption seemingly made by someone who doesn't quite understand just how much time and especially money that actually comes out to, and I hardly blame the writer of the article because level and generally game design are two disciplines not often understood by designers themselves in a pedantic sense.
Making levels entirely out of meshes as opposed to geometry works in some cases, many cases even, but it also limits what level designers are allowed to do in that a level designer without 3D modeling skills is basically considered useless, and indeed I started reading ebooks on how to use blender long before I even picked up my first book on C# or the Unity engine, it's just a fact of these modern platforms. Not wanting to fall too far from the topic of remakes or "re-glorifications" and into a discussion of level design in the current decade, I would like to at least conclude by saying modern game design and iteration is incredibly difficult and remaking old games with "complex" level designs is way harder and way more intensive a task than making new games with simple designs.