Gear Design In Early Video Games Vol. 1 (D&D, Akalabeth, and Rogue)

The screenshot below is from a game called pedit5, also known as The Dungeon. As you can see, it’s a simple game. It was never sold in stores. And it was made by a single person, Rusty Rutherford, as part of a research grant rewarded by the University of Illinois. No one’s entirely sure if the original version from 1975 still exists. This is from a recreation.

It was, in every practical sense, the world’s very first digital RPG.

It will probably come as a surprise to absolutely no one that this game, like so many back then, was an early attempt to adapt Dungeons and Dragons.

Dungeons & Dragons (1974)

Before we can talk about D&D, we should talk about an earlier game called Chainmail. It was a 1971 miniatures wargame created by Gary Gygax and Jeff Perren. Gygax is a name you may recognize because he, along with Dave Arneson, would publish Dungeons and Dragons a couple of years later. And D&D’s original rule-set was so intertwined with Chainmail’s that the instructions actually assumed you owned both games.

There were three pieces of armor in Chainmail— leather, chainmail, and plate — each with their own armor class. The lower the armor class the harder it was to hit the character. There was a handful of weapons, but they all served similar functions. Regardless if it was a sword or spear, you rolled a d6 for damage every time. These limitations were inherited by D&D’s initial release.

The arrival of the Greyhawk Supplement in 1975 aimed to fix a number of issues with D&D’s initial publication. First and foremost, it removed the game’s dependency on the Chainmail and made it easier for new players to learn. I say removed, but it was really just re-branded. The serial numbers were filed off, but it was still Chainmail. The most important improvement, at least to me, a 30s something writing this in the year 2020 with the benefit of 20/20 hindsight, is that weapons were differentiated from one another through different dice rolls:

Daggers dealt 1d4 of damage while swords dealt 1d8. And to make the weapons even more differentiated from one another, they actually dealt different amounts of damage depending on the size of the target. Some weapons did more damage to huge monsters like dragons while others were better for fighting your standard human-shaped bad guy, like an Orc.

Greyhawk was one of the first instances of “good” equipment design in gaming. At least in theory, if not in practice. A morning star in AD&D did 2–8(2d4) damage to a small target. A trident, on the other hand did 1d3 +1d4. The morning star is clearly better, all other stats being equal — I’m not accounting for speed factor or space required here, because, quite frankly, no one ever did.

The morning star vs. trident example is reversed if you’re fighting a larger foe. The trident is clearly dominant, doing a whopping 3d4 worth of damage. This example also brings us to a concept that I think is important to discuss before we get much farther.

That 3–12 points of damage that the trident is doing to a giant monster is really, really good. Like… It’s OP as hell. It’s even better than many people will realize because the damage it does is not actually entirely random.

Roll 2d6s, and the probability of a total of 6 is fairly good, but the probability of getting a 12 or 2 is pretty damn low. That’s because the sum of random numbers is not random; the more rolls of dice you pile on, the less random the outcome.

This is great for game designers. Without chance, games become boring, but if something like a weapon’s damage is chosen totally at random, it’ll start to feel unfair. So we can use this to find a balance between randomness and predictability. Always remember, players want to feel like anything is possible, but they still need to be able to predict an attack’s outcome.

Akalabeth: World of Doom (1979)

Akalabeth: World of Doom, another homage to D&D, entered the scene in the late 70s. It was a passion project created by then high school student Richard Garriot, the game designer who would go on to create the Ultima series a few years later.

Akalabeth was, in many ways, a step forward from pedit5. There are two classes to pick from, more than one dungeon to explore and a grand total of three weapons to wield — an axe, a rapier, and a bow. Four weapons if you include the shield, which you can also use to attack with. You have access to these weapons as soon as you have the gold to buy them; in theory you could even pick them up at the start of the game if you’d rather have the equipment than the food, but my guess is you’d starve soon after — I’m not really sure how much of a choice that is. Akalabeth’s opening hours are tough as it is.

There are the beginnings of interesting questions here; do you buy the better equipment at the cost of more supplies? When is it worth the risk? This dynamic, between purchasing better weapons and playing it safe, is something even present day titles exploit, and while Akalabeth isn’t quite there, it’s certainly interesting to see the same design way back at the beginning of the medium.

Once you save the gold to get the gear, you’ll find out that the rapier does about 0–10 damage. The axe does 0–5, but it can be thrown. The bow does 0–4 on each hit, but it’s the game’s only ranged weapon. The initial damage from each weapon is selected at random. The game then adds .2 points worth of damage per point in strength to the final hit. Combat definitely feels weird and unpredictable, at least in the beginning.

But there is something that Akalabeth does very well, especially for how archaic the game can be. Each of those 3 (or 4) weapons does something that the others absolutely can’t. You can throw the axe. The bow can fire at range. The shield reduces the enemy’s hit chance. They’re differentiated by more than just the amount of damage they can do.

Gear design’s boring when it’s all just lesser and greater versions of the same weapon. The choice between a sword that does 4–6 damage and a sword that does 6–8 damage is no choice at all. It’s 100% empty game design.

Rogue (1980)

Rogue came out a year later. In it players explore several floors as they seek the Amulet of Yendor on the dungeon’s bottom level. As you probably already know, the floors, enemies and loot are all procedurally generated, unique to each playthrough. If you die, that’s it. You have to start over.

From a gameplay perspective, Rogue’s a leap forward from Akalabeth and its contemporaries in many ways, especially with how it uses gear drops to maintain player interest . Akalabeth and so many early RPGs struggled with this — the player gets a handful of upgrades early on to maintain their attention, but there isn’t much else. Rogue is one of the first examples of a game that supplies its player with loot all the way through to the bottom floor. And those small victories along the way make a huge difference.

Jesse Schell called this an “interest curve”. It’s a simple graphical representation of a player’s interest over time. You want to introduce new mechanics, levels, and challenges so the interest curve ends up made of these hills and valleys. Without them you risk losing the player’s attention.

Interest curves are an incredibly important concept to consider when constructing reward systems. And Rogue is one of the earliest examples of getting it right. Making sure there’s a steady stream of rewards along the way is sometimes more important to the average player than finishing the game or vanquishing the bad guy. And that’s true today just as much as it was back in 1980.

Rogue released ahead of its time in other ways as well. Scrolls and potions make up the bulk of the game’s one-use items. Each has randomly determined effects when used. Most of the time, you won’t know if what you’re about to drink will help you or hurt you. Scrolls of identification can be used to, well, identify the item, but you only get so many of those. Same thing with rings. And weapons. One of the core skill competencies of the game is deciding when to use scrolls of identification and determining which items you can safely use without identifying. This is probably the first time a game rewarded a player’s in-depth knowledge of a game’s item pool.

Defeating Rogue, which, full transparency, I’ve never come close to doing, usually comes down to a matter of luck. That’s not to say that’s bad per se, many people love it, but… I think it’s an acquired taste.

There’s a balance to find here, between deterministic rewards, where the player knows exactly what they’re going to get, what they’re working towards, and totally random rewards, where the player has no idea what they’re going to receive or when they’ll receive it. It’s probably worth mentioning that I’m not talking about technically random, but how it is perceived by the player. A game can have somewhat predetermined rewards and still feel random. And that’s definitely something we’ll revisit later.

Before we get to part II, you might be tempted to ask why I want to focus on games and series that are largely irrelevant to today’s gamers. And my primary reason is actually pretty simple… A lot of these games, especially in the early 80s, are… not very sophisticated. And that makes it easier to introduce design principles before layers and layers of complexity are piled on by the better-equipped, better-funded developers of the late 90s,

But, even more than that, I think these games are interesting to look at because the people who made them… didn’t really know what they were doing. And I don’t mean that to be disparaging… This was the wild west. Games were made by one or two people in their basement with very little to go on. And nowhere was this more apparent than in the realm of gear design.