For the last three decades, I've witnessed many gaming developments. From my humble roots with the 8-bit wonder of the NES to the mid-90s 2D PC scene to the rise of 3D, there were new and wondrous creations to captivate audiences at each step. As the industry grew, so did the scope of what games could be, and what they could include.

In retrospect, some of the earliest games were far more complex than we give them credit for, whether through text-based interactions or interlocking systems on a granular scale, yet over time the medium slid back toward simpler, more straightforward gameplay. Now we find ourselves in a moment where every game touts excessive length or a sheer number of mechanics, and the biggest buzzword going is the Open World.

I love games, and I truly do, but it has been a long time since I felt the sheer wonder of playing one. Part of it could be that I'm rapidly approaching middle age. Part of it could be that today's games don't carry the charm older titles did.

What I want to do, across the next few pieces, is talk about three games that shook me out of that gamified melancholia. Three games separated by years and by genre but joined by a single design philosophy. We'll start with the two that taught me the principle, and we'll start in 1999.

You may have heard of this one from your weird uncle, or in passing from various neckbeardy gamers online. You may have also seen the recent trailer, or the reactions to said trailer from the same weird uncles and gamers. The game in question is Legacy of Kain: Soul Reaver.

Developed by Crystal Dynamics, much to the chagrin of Silicon Knights, Soul Reaver is an action-adventure puzzle game for the PlayStation and Sega Dreamcast, later ported to PC. As the sequel to Silicon Knights' Blood Omen, it picks up roughly 1500 years after the canonical ending of the first game. We follow a vampire-turned-wraith, Raziel, as he embarks on a quest of revenge against his sire, Kain. That simple, direct premise is the first of many things the game gets right.

Through the opening scene, we watch Raziel betrayed and thrown into a centuries-long torment. As the audience, we understand exactly what our protagonist wants and why. This straightforward premise is joined by incredible voice acting and writing that elevate the story to a higher echelon of art.

We're quickly engaged in exploration, with assistance from a many-eyed patron who introduces the primary gameplay mechanics through realistic scenarios. Each skill is built on gradually (learn to jump, learn to glide, combine jumping and gliding, and so on), and the player is trained to remember that if they ever get stuck, they can shift into the Spirit Realm, a core mechanic in which the world warps in numinous ways, including new platforms and pathways, or simply allowing Raziel to pass through barriers. Each new mechanic is given an opportunity for the player to test and, in some cases, master before moving on.

These two factors combine into what I've come to see as fundamental for any game to work properly. The story, at least for the first entry in a series, should be direct, and the gameplay mechanics should be straightforward, increasing in complexity by combining what was previously taught. There are more aspects of Soul Reaver I want to cover, but I want to introduce the other primary example before getting too deep into the weeds.

Being a PlayStation fanboy for many years, there are titles one might expect me to have played. From Jak and Daxter to Crash Bandicoot to Uncharted, and given how old I am, there are a few that should have been on my radar.

Team Ico's Shadow of the Colossus is one of those. It has been a known quantity for a long time, yet much like Game of Thrones or The Joker, I resisted looking into it because the hyperbolic hype made me recoil. Hype is at best warranted for a good property, at worst an ouroboric recitation of platitudes echoed by the masses. But I digress. This game deserves the hype.

Fumito Ueda, the director and creator of SOTC, employs a design philosophy he called Design by Subtraction. This minimalist approach distills the essence of a game down to the most necessary parts, best typified by the sheer lack of UI elements in his work.

Shadow of the Colossus, or as it's known in Japan, Wander and the Colossus, is a fundamentally simple game. The plot is direct: a boy wants to resurrect a girl who was sacrificed because she was said to have a cursed fate, and so he seeks the means to that end. The means, in this case, happen to be the killing of giant creatures.

To accomplish this, you're given the only tools you need for the entire game: the Ancient Sword, a Bow, infinite Arrows, and Agro, your horse. Each tool has a clear use, aside from the sword, which doubles as a sort of compass and weak-point marker. To slay a Colossus, you need to find it, identify its weak point, and scale it to stab said weak point. Every encounter breaks down to those three phases, with some variation in the third (I'm looking at you, number four). Once again, a simple premise and a straightforward toolkit, but with two other aspects worth highlighting, aspects also present in Soul Reaver but earning their attention here.

Before we go further, something has to be cleared up, because the most common reading of Shadow of the Colossus gets it wrong.

The Colossi are not living creatures, nor are they automata. Ueda has clarified, in interviews over the years, that Dormin was not a benevolent being. Dormin had been split into sixteen parts and placed inside the Colossi after a separation tied to idolatry and ritualistic sacrifice. When people talk about SOTC being a game about cruelty, the cruelty isn't as simple as "don't be mean to animals." The real cruelty is what Wander does to himself in pursuit of his goals. He corrupts himself knowingly. Dormin tells him the cost will be heavy, and Wander says he doesn't care. He no longer cares about himself, and so he dooms himself in the process. Mono pleads with him to stop after the fourth Colossus falls, again per Ueda himself. That is the true cruelty of the game, the cruelty Wander inflicts on himself in service of an outcome he has decided is worth any price.

There are many unanswered questions left in SOTC, but the fundamental misreading of its premise is the greatest tragedy in its critical reception, second only to what Wander does to himself. The game is more austere than its reputation, and that austerity is key. Ueda subtracts not just UI, but mechanics and exposition in equal measure. What remains is a player who has to do the work of understanding, and a protagonist whose choices are damning precisely because nothing in the game stops him from making them.

The modern buzzword bandied about in gaming these days is the Open World. The concept is older than many imagine, going back to the Elder Scrolls series with Arena and Daggerfall, which each comprised thousands of square kilometers of explorable terrain. The problem then was procedural generation versus intentional design. Those games were veritable ghost towns with little to no activity. The Elder Scrolls is also the antecedent of many issues plaguing today's open worlds. Thanks to the precedent set by Oblivion and Skyrim, the inverse is now our situation: large open worlds teeming with content engineered to keep something interesting always within a certain radius of any given point. Games like Horizon Forbidden West are so full of interactive locations they become overwhelming. There is no reason an Assassin's Creed Odyssey or Valhalla should take over 100 hours to complete.

There is a middle ground, however, where the world is still open but the locations are crafted meticulously and intentionally. That is where SOTC and Soul Reaver both shine, and they reach it from opposite directions.

Soul Reaver builds its world like a Metroidvania, where new abilities open old paths and the map unfolds gradually as the player grows. The progression is linear in plot but exploratory in feel, and there is no wasted space. Every corner of Nosgoth exists because it serves the puzzle, the lore, or the moment, and nothing else.

Shadow of the Colossus does the opposite. The Forbidden Land is mostly empty. The ride between Colossi is long, and the silence between encounters is filled with the solemnity of your deeds. What might read in a modern review as filler is, in Ueda's hands, the emotional weight of the journey. You feel the distance you've put behind you. You feel the cost of going forward. The emptiness has been chosen, and your play, assent.

This is the principle the two games share. Whether the world is dense with purpose like Nosgoth or vast with deliberate quiet like the Forbidden Land, the through line is that nothing in either game is there by accident. Every system, every space, every silence has been weighed against the question of what it adds to the experience, and anything that didn't justify itself was cut.

That is what design by subtraction actually means; not minimalism for aesthetic reasons, but the discipline of refusing to add a thing unless it earns its place. When designers commit to that discipline, their worlds feel deeper than the literal feature count would suggest, because what remains has to carry weight, and weight is what makes a world feel real.

It's tempting to think this kind of discipline belongs to an earlier era, but Ueda himself put that notion to rest this past June, when genDESIGN revealed gen ATLAS, his first game since The Last Guardian and the first of his career to leave PlayStation behind. The premise reads like a restatement of everything above: a lone figure waking on a vast, silent, abandoned planet with no memory of how they got there, a colossal machine serving as the spine of both traversal and story, and a studio that has gone on record keeping generative AI out of all of its creative work and confining it to scheduling and other busywork. Two decades on, the man who taught me half of this is still building from subtraction.

Soul Reaver came out in 1999. Shadow of the Colossus came out in 2005. The principles they were built on haven't aged. What has aged is the willingness of the people making games to trust those principles, and that's where we'll pick up next.