A game’s difficulty level can make or break the title. Games that are perceived as too difficult become boring, depressing grinds, while games that are too easy become boring and tedious, with little challenge. One of the most profound differences between World of Warcraft Classic and Retail is the difference in difficulty. Of course, every player has their own ideas about how hard a game should be, but there’s no arguing that the difficulty of a title is important.
But according to game developer Jennifer Scheurle, game developers think about game difficulty very differently than players do, which may be part of why conversations on this topic sometimes seem to break down. Her piece resonated with me, partly because it reminded me of the reasons why I became a game modder, once upon a time. According to Scheurle, difficulty is all about trust.
“At the core of the difference between how game designers and players speak about difficulty,” she writes, “is the fact that we discuss it in terms of skill progression. All difficulty design is essentially that: crafting how players will learn, apply skills, and progress through challenges.”
She then walks through examples of how this plays out in games, using the Dark Souls series as an example. DS games ask you to accept that you will die (frequently) as part of learning how encounters function. You aren’t simply being killed by mechanics you can’t master, beat, or counter, you’re learning how the game functions and how to counter incoming attacks. The game, in turn, obeys its own internal rules. Players often become angry at a game if they feel it isn’t holding up its end of the bargain in some particular, whether that refers to drop rates, spawn rates, boss difficulty, or the damage you take versus the damage you deal. She also discusses the importance of how a game teaches players to play it, and the various in-game ways that developers communicate game difficulty and associated rules. It’s a very different view of the topic than simply boiling it down into whether a game is “hard” or “easy,” and it leads to a much more nuanced view of how and why different titles may put difficulty in different places.
The article resonated with me in part because it describes part of why I became a Diablo II modder and taught me something about my own motivation. I don’t want to seem as if I’m hijacking Scheurle’s excellent discussion of game difficulty because it’s worth a read in its own right, but I’m going to switch gears a bit and talk about my own experience. To put it simply: I was pissed.
Diablo II’s Trust Fail
This was the early days of Diablo II, before the Lord of Destruction expansion had even come out. Patch 1.03 dropped not long before I started modding, to put a date on things. On Normal difficulty, Diablo II worked pretty well, but as you progressed into Nightmare and Hell difficulty modes, deficiencies became apparent.
Back then, Diablo II used a linear leveling curve in which the amount of XP you needed to gain for each additional level increased by a flat amount — the amount you needed for your previous level, plus a flat modifier. This was exacerbated by a leveling penalty, introduced in Nightmare, in which you lost XP gained towards your next level if your character died. You couldn’t drop a level due to this XP loss, but you could theoretically be 99 percent of the way to Lvl 50 and fall back to 0 percent at Lvl 49. The net result of this was that the amount of time required for each additional level increased sharply, and this became increasingly noticeable as you moved into the later game.
Now for the coup de grace: The game was poorly balanced outside of Normal difficulty. I became a game modder specifically because my Barbarian character with maximum Fire Resist was being one-shotted by mini-bosses with Fire Aura even when he used abilities that temporarily increased his HP. These mini-bosses and bosses could one-shot a character virtually as soon as you saw them. Death meant losing a portion of gold and dropping equipped items. Attempting to retrieve those items (using whatever alternate gear you had access to) was virtually guaranteed to get you killed at least once more because you’d have to drag monsters away from your corpse in order to try and retrieve what you originally had. Mini-bosses could also spawn with these modifiers in critical areas, where it was exceptionally difficult to move them away from a critical spawn point. There was no way to see the exact location of the fire aura on the ground; you knew you’d touched it when you died.
It was cheap. That’s what I called it. I didn’t consider it any kind of legitimate difficulty spike. It just felt like a way for Blizzard to make the game harder by killing players in a manner they couldn’t even fight. I became a modder because I was angry about the way that these imbalances had changed the game. I felt betrayed.
Looking back (and using Scheurle’s article for reference), I’ve realized that I was angry because Diablo II had broken trust with me. Some of these flaws existed in Normal as well, but they weren’t as apparent due to the influence of how other scaling factors impacted the title. Some of the changes between Normal and later difficulties that impacted how poorly the game scaled included the much-slower pace of leveling and the fact that there were no unique items in-game for the Nightmare and Hell difficulty modes. This made it pointless to spend gold on gambling (since gambling, at the time, only produced normal weapons). The slow speed of leveling meant that one of a player’s primary means of gaining power was substantially curtailed. There were also notable power imbalances created by the use of percentages for some metrics (like life steal). In original vanilla D2, life steal was absurdly overpowered — and absolutely essential to surviving the late game. Certain classes were locked into endgame strategies as a result of bad math and poorly balanced game mechanics. It grated on me.
The changes to Diablo II from Normal to later difficulties weren’t just the result of Blizzard trying to be jerks. It’s common for RPGs to have poorly balanced endgames because most people do not play them for long enough to actually experience the endgame. This was a topic of discussion around Skyrim when that game was new, and it explains much of what happened with Diablo II way back then.
I developed the Fusion 2 mod for Diablo II, followed by a much larger overhaul, Cold Fusion. I and a team of three other people — Justin Gash, John Stanford, and Matt Wesson — cumulatively poured in several thousand man-hours of development time into Cold Fusion. I led the effort, which was a core part of my best friend’s senior project in computer science and consumed no small chunk of my own senior year in college. I’m not sure the game files exist on the internet any longer, but you can see the original website archived by the Wayback Machine. Fair warning: I was not a web designer. Still, it gives some idea of the scope of the project, if you’re familiar with Diablo II.
While I don’t expect anyone reading this to have ever played the mod — I never released an LoD-compatible version of the project — it was a pretty major part of my life for the time I worked on it. We overhauled the entire title, tweaking drop rates, fixing bugs, and implementing a new leveling curve, a new difficulty curve, new monsters, and new unique items intended for both Nightmare and Hell difficulty levels. We developed new audio effects, visuals, and skills using pieces of code that developers had left in place in the engine and audio effects another friend created. We pulled certain unique items over from Diablo I (with Diablo I art) and reworked the skill trees to better balance the game. Our goal, in every scenario, was to build a more consistent Diablo II that didn’t just funnel characters into a single endgame build but allowed other skills to compete as well. I was quite proud of the fact that when Lord of Destruction came out, it adjusted Diablo II in some of the same ways we had, and even introduced new spells that were similar to some of the ones we built. I’m absolutely not claiming that Blizzard took inspiration from our work — it was just neat to see that we’d been thinking along the same lines as people at the company.
For example: We implemented a logarithmic curve for CF’s level scaling — one that was designed to allow a player to run the game once at each difficulty level and finish “Hell” near maximum level. Blizzard wanted a game that would require many, many, many runs through maximum difficulty to reward Lvl 99 and used a differently-shaped curve to do it — but they still moved away from the linear curve they used in the early phases of the title when they launched the expansion, Lord of Destruction.
Until now, I never really understood why I was so unhappy with the base game in the first place. Now I do. I felt as though the collective changes to Diablo II that happened after Normal weren’t just the result of making the game harder — they made the game different, in ways that felt like they’d broken the trust Blizzard had established in building the game.
It’s not often that you discover the explanation for why you spent a few thousand hours rebuilding someone else’s project in an article written 18 years after the fact. I suppose Cold Fusion has always felt a bit like a road-not-taken path for me. It had its fans, but it was one reasonably popular mod among many, not a DOTA or a Counter-Strike. Either way, I appreciate Scheurle’s discussion of difficulty and how developers think about the topic. It shed some light on an episode of my own life.