Photorealism Leads to Inferior Video Games

A few weeks ago, I came across a YT video making an interesting point, that I had also made about 5-8 years ago. It compared the quality, realism and re-playability of video games since the first 3D FPS games from early 1990s to today, and pointed out that quality of gameplay had actually decreased in the past decade (2013-2023). While this might sound shocking to all those loyal “consumers” who are constantly filling the large empty holes in their pathetic cucked lives by buying the “latest current” thing, this observation is absolutely true and is part of the reason why current video games (but especially from AAA studios) suck so badly. Other reasons include buggy and broken gameplay, endless patches, very underwhelming paid DLCs and an overall tendency to extract as much money as possible from these “loyal consumers”.

Anyway.. back to the topic at hand. So why do I think that increased photorealism in video games leads to inferior video games? Surely, Wolfenstein- The New Order (2014) looks far better than Wolfenstein 3D (1992) and Doom (2016) looks better than Doom (1993).. right? Sure, they do- but that is not the point! The real question is how fun and entertaining they are to play- especially play again. Furthermore, Wolfenstein (2014) and Doom (2016) are deliberately not photorealistic- even though both have great graphics. In fact, these reboots are among the few modern video games that have carefully avoided the trap of photorealistic graphics in favor of a distinct, consistent and visually engrossing artistic style. And this brings me to how another classic video game (Half Life 2), or more precisely some its later photorealistic mods, showed me that focusing on photorealism was a dumb idea.

Before we go there, let us quickly talk about the decade with the biggest advancement in video games graphics aka 1993-2004. The start of this decade began with Doom and ended with Half Life 2, and the difference in level of graphics between them was day and night. We simply have not seen a similar jump in video game graphics since then, and for good reasons. If you are too young to remember that era- just go on YouTube and have a look at Doom released in 1993, itself a trail-blazing game, and compare it to Half Life 2 (HL2)from 2004. But here is the funny part.. both games have very high replay value. And this brings me to the topic of photorealistic mods of HL2. While there are many mods and some are quite good, almost every single photorealistic mod of the full original game plays worse than the original. This issue persists even if the mod is implemented as a demo in a much newer game engine such as Unreal 4 or 5. But why would this happen? Shouldn’t the results be better?

Let us start with what I consider to be best demos of photorealistic HL2 made using Unreal 5 game engine. It looks amazing and gorgeous as a demo, but any gamer can quickly see a glaring problem with photorealism. If you watch the entire two minute video, it becomes obvious that the photorealism comes at the cost of visual clutter aka the inability to quickly and clearly see important stuff in the game. But why should this happen? Wouldn’t more photorealism translate into better object perception in video games? Well.. it does not, and there is an interesting reason why it does not. To understand what I am getting at, we have to talk about the difference between viewing a collection of objects using binocular vision and doing the same for flat pictures. In a real 3D environment, our eyes and brain can isolate an object from the background using a combination of stereoscopic information, innate pattern recognition and a number of other processing tricks which our brains have mastered through the process of evolution. But this works only in a full and real 3D environment.

This system does not work well when we view flat (2D) pictures, because our eyes and brains did not evolve for that purpose. This is why careful posing, lighting, selection of lenses etc is so important for movie making and professional photography. This is also why well-made movies look “real” even though every shot is carefully staged, lighted and photographed. In other words, obtaining photo-realism when the image is on a flat surface requires a careful consideration of how images look on such a surface in addition to a deliberately chosen and consistent artistic style. This is also why games with a well chosen artistic style can appear far more real and fun to play than much more photorealistic ones which lack such a style. To summarize this part, any game whose graphics are shown on a flat surface will always be far more dependent on clever game design, in game lighting and consistent artistic styles than photorealism.

For an example of what good art style looks like in a more photorealistic reboot of a famous game- take a look at the numerous trailers and gameplay videos of ‘Black Mesa’- the 2020 reimagining of the Half Life 1, which was originally released in 1997.

But what about 3D games? Shouldn’t 3D glasses eliminate this problem? Well.. I have bad news for you. 3D glasses, if anything make these issues worse. To understand what I am getting at, here is a simple experiment you can perform. Go sit in a busy mall or other public place and focus on somebody walking in your general direction, a hot girl perhaps. You will immediately notice that your eyes can isolate one or more people from the background such that there is object separation with minimal background blurring. Now ask yourself- how do you eyes and brain pull that off? The answer is they can use slight differences in focus and 3D position (including motion) of nearer objects to process them differently. However 3D glasses, as they exist, cannot replicate the slight differences in focus and changes in relative 3D position which occurs when you look at a real 3D scene- and your brain can tell the difference.

This is why playing the vast majority of video games using 3D glasses feels so weird and unpleasant- however good the graphics and resolution might be. To make matters worse, most in-game human faces, movements still look really artificial and weird. To be blunt, a 3D Mario is far more believable than a 3D NPC, even one that is very well designed and animated. But why does any of this matter? Well.. as it turns out, for the past ten years, many of the large gaming studios have been putting a lot of effort into making their game “photorealistic” while ignoring other far more important things such as world building, game design, consistent art style, re-playability etc aka the stuff that matters. This dumb movement towards more “photorealism” is largely led by bean counters and MBAs, who have not finished a video game in the past two decades. However, in large corporations, these parasites make all the decisions about funding etc. In other words, large gaming studios are going to keep churning out increasingly expensive, buggy and forgettable crap in the foreseeable future.

What do you think? Comments?