Abstract
Problem: How should RPGs handle difficulty? Should they offer customizable settings, and should games get easier or harder as the player progresses?
Approach: Tim Cain draws on decades of RPG development experience (Fallout, The Outer Worlds) to discuss difficulty philosophy, concrete implementation techniques, and the tension between challenge and friction.
Findings: Difficulty settings are an accessibility feature, not a design failure. The best approach combines numerical tweaks, systemic changes, and reduced handholding — all as player options. Games should get both easier and harder as the player progresses.
Key insight: New areas should scale to the player's current level while old areas remain at their original level, creating a natural curve where players feel powerful revisiting old content but face real challenges in new territory.
In Favor of Customizable Difficulty
Tim opens by stating he supports customizable difficulty settings. He's encountered a common argument from designers and programmers: offering options means you couldn't make a decision. He acknowledges this is sometimes true — and that options make balancing harder — but frames difficulty settings primarily as an accessibility feature.
His example: a level designer friend who always plays on easy because he wants to see the game, not get stuck on combat encounters. Tim considers this a perfectly valid way to play. Difficulty sliders let players adjust the game to their personal comfort level.
Numerical Changes
The simplest approach to difficulty scaling involves pure numbers:
- Enemy progression adjustment — If your enemy scaling algorithm is based on player level, pass in a modified value. On hard difficulty, pass
player level + 5. On easy,player level - 3. For extreme difficulty,player level × 2, which starts slightly harder but diverges dramatically at higher levels. - Hit points — Give players less HP or monsters more.
- Health regeneration — Slower regen, no regen in combat, no regen outside combat, or requiring rest at specific locations (inns, camps) that cost resources.
- Skill checks — Rather than hardcoding check values, mark them as easy/medium/hard/very hard and use a lookup table. A "hard" check might be 50 on easy mode, 75 on normal, 85 on hard, and 95 on the highest difficulty. This requires zero code changes — just a different table per difficulty level.
Non-Numerical Changes
Beyond tweaking numbers, difficulty can come from systemic requirements:
- Survival mechanics — Requiring food, water, and sleep on harder difficulties (as The Outer Worlds did with Supernova). These cost resources and money, adding economic pressure.
- Fast travel restrictions — In The Outer Worlds' Supernova mode, fast travel only returned you to the ship. Leaving the ship meant walking everywhere, encountering enemies repeatedly, and managing time.
Tim acknowledges the counterargument: these changes add friction rather than true difficulty. He concedes this is valid for virtually every difficulty increase — the line between friction and challenge is subjective.
Non-Combat Difficulty
Most RPGs only apply difficulty settings to combat, but stealth and dialogue deserve attention too:
- Stealth — Increase enemy detection radius or speed up the detection meter on harder settings.
- Dialogue — Raise skill check thresholds (as with combat checks). More interestingly, make evidence harder to obtain. Tim references Fallout: you couldn't just tell the Master his mutants were sterile — you needed proof from a Brotherhood of Steel scientist. On harder difficulties, evidence could be moved to harder-to-reach locations, locked in tougher containers, or gated behind additional quests and conversations.
These require more design work upfront but trigger automatically once the difficulty framework is in place.
Reduced Handholding
A third category Tim calls "less handholding" — reverting to how older games worked:
- No quest markers — NPCs give verbal directions ("the bandit's cave is somewhere up north"), but no map pin appears. Players must explore and pay attention.
- No maps — Many older games had no local or dungeon maps. Players drew their own or navigated from memory.
- No quest log — The most extreme option. If someone gives you a quest, you have to remember it yourself: what to do, what to get, who asked, where they are. Tim notes this is trivially easy to implement (just disable the log UI) and some players find it deeply immersive.
He recommends making all of these optional because for many players, the line between difficulty and annoyance is personal. What feels challenging to one player feels like pointless friction to another.
Should Games Get Easier or Harder Over Time?
Tim's answer: both. This is why he dislikes global level scaling.
- Old areas should feel easier — Returning to the bandits by the bridge who used to crush you and stomping them at high level is satisfying. It makes the player feel powerful.
- New areas should present fresh challenges — When entering a new area, its level should be set to the player's current level, ensuring meaningful encounters.
This avoids the common late-game complaint of "nothing is challenging, I have tons of money and nothing to spend it on" while still rewarding player progression. The key distinction is between general enemy scaling (which Tim dislikes) and area-level locking on first entry (which he prefers).