Difficulty levels in games

I’m reminded of this principle: learning is necessary for fun. A successful game design teaches players how to use its systems to overcome its challenges. The player who fails without learning becomes frustrated and quits. Some can’t or won’t learn or lack essential skills or need extra help (easy settings, cheats, accessibility), and some learn quickly and want more difficulty.

Because the gameplay is supposed to be fun. Players who don’t want to engage with the tools and obstacles probably picked the wrong game (or hobby), or they do need to change the difficulty, or else they’re struggling against bad game design.

Market forces. Developers want the widest possible appeal for sales with as much of the game to be reachable as possible to avoid wasting resources building unseen content, and nobody can afford to alienate interested players when there is so much competition. AAA budgets especially require compromises because universal balance for all players is generally not possible.

“Making a game for everybody is the same thing as making a game for nobody.” -Soren Johnson, reflecting on the development of Spore.

Their goal is usually not to defeat or frustrate the player. It’s more like mentoring. Players must decide how hard to push themselves based on what they want from the experience. The “losing is fun” idea (as in Dwarf Fortress or AI War) is insightful and valuable to some of us, but not all.

I’m in favor of what I call “geographic difficulty”, where most content is accessible to the player in an open-book or open-world format, rather than gating difficulty behind play time or progression or options. Think Ghost Recon Wildlands or Crusader Kings 3, where the player can tackle whatever interests them or feels appropriate. A game world can have guidance and warning signs without blocking the player’s path. The difficult content can be there to entice players even if they opt out.

A downside to this may be choice paralysis, where having too many options becomes a disincentive. It’s a bit like the map-marker spam problem, as having some absurd checklist to optimize before you play feels much worse than just exploring.

There’s also a practical reason why you don’t start the game on hard: it’s never really clear what hard means. Those who want the easiest possible setting can’t really go wrong with easy. But what does hard mean?

Does it mean, “it’s meant for players going through the game a second time who already know all the tricks and might have New Game+ items” or “we made normal piss-easy, so hard is still a joke” or “screw it, we literally didn’t do our job, we just tripled all the values, have fun banging your head against a wall”.

Can you think of a better word? I’m always up for better options for how to articulate it. But when I heard “enjoyable frustration” many years ago – I don’t even remember who is was, sadly – it felt like the perfect description of how game design works.

Don’t get hung up on the term adversarial! Think of it more like a dungeonmaster throwing down obstacles. What kind of lousy dungeonmaster puts you through a dungeon without traps and monsters? Ultimately, he’s an entertainer and a storyteller, but part of the experience he’s engineering involves monsters and traps. Obstacles. Frustration. Game design is the same way, and the adversarial component is what sets it apart from entertainment from a novelist, filmmaker, sculptor, or juggler. None of them presents obstacles. None of them has an adversarial component in the entertainer/consumer relationship.

Dude. Don’t even. :)

I have to disagree, because there’s no adversarial element to mentoring. A mentor doesn’t create obstacles to his student’s progress. A mentor has no interest in frustrating the student’s learning or challenging his ability to proceed through a lesson. I mean, teaching is its own kind of skill and I wouldn’t necessarily compare it to any form of entertainment, especially something unique like game design.


Flow theory uses the word “anxiety” or “frustration” to denote when a task isn’t going well because it’s consistently too difficult (it’s “boredom” when it’s too easy). So within the flow space, you have a sense of challenge without overwhelming frustration.

I don’t know if this chart makes sense without a full explanation, but the moments you’re talking about are the parts of the wiggly line that veer up toward frustration, before you learn how to address a challenge and the line flattens out.


Flow? I’m going to vote against using flow.

I’m pretty sure that chart really wants “flow” to be labeled “fun”, but it knows if it did, I would scold it.


By the way, I usually don’t do either of those things! If the game designer can’t be bothered to do his job, I just play on normal and let the chips fall where they may. It’s not my job to find the fun.

The only exceptions are when I’m basically done with a game but want to see it through for whatever reason. By the time I got to the end of Cyberpunk and was doing the last mission, I honestly couldn’t have cared less about any challenge level, so I dialed it down to easy and just cut everyone up with a sword to get to the ending cutscene.


Frustration is a problematic term because games can produce the psychological element as well as the intended hindrance.

Students do get challenged and tested. A game designer provides both training and tests.

Maybe coaching is a better analogy. A sports coach has to push you out of your comfort zone to excel and will put you in competitive situations where you can fail. Some sports are games too, but not all.

Finally, something we can agree on.

But I more generally agree with this:

If difficulty levels are harmful to the game, how do “incentives” make them right? how would an achievement or XP bonus for playing on hard make your description of playing Prey on normal any better?

By the way, the word you’re looking for in the frustration talk is just “challenge”. This carries, in the context of a game, the whole concept of obstacles that are satisfying to take on and overcome.

And incidentally, that description of Prey sure sounds like an unbalanced weapon to me, which is a different problem than difficulty levels undercutting the design. (Personally, what I’ve played of Prey was on Easy, because those Typhons are brutal as hell in at least the early game and I was getting super frustrated with the game on Normal. Easy didn’t make anything a cakewalk, it made it manageable for me with some ingenuity.)

I agree. Some more thoughts:

  • Which difficulty levels were really tested? Which ones are really balanced to bring out the experience the developers wanted, and which ones are just +1 boosts thrown around, either to player or adversaries, to cater to a larger audience? All difficulty levels are not the same, but there’s often no easy way to know.
  • Higher difficulty levels for people who mastered the game (NG+) make sense. At this point, they’ve gotten full value out of the game and are simply trying to test their skills against harder, but untested difficulties.
  • Not all games are the same w.r.t difficulty levels. Strategy games are very value-specific and cannot easily be ‘tweaked’ for difficulty levels (imagine tweaking the number of spaces a knight can move in chess). Ideally you’d pit weaker AIs against the player in lower difficulty levels, but in reality the AIs are so miserable, it’s just a matter of how big of a cheat they get.
  • The same applies to some degree for RPGs (inasmuch as they have strategic components). It’s worth noting that getting the balance right for these complex beasts is so hard as it is, that adding difficulty levels is usually just silly and turns an imbalance into completely broken systems.
  • On the other hand, twitch-based games (action etc) can be adjusted for different skill levels rather easily, by just changing enemy reaction times. These are games that require coordination, and different ages and abilities have different built-in values for them. The balance of the game is a far less intricate affair, but the key is that every difficulty level must still be properly tested and adjusted to engineer the desired experience, and this rarely happens.
  • Puzzle elements/games could also be adjusted, going from a more difficult puzzle to a simpler one, but this is also very rare as it requires extra design rather than modifying puzzles. Puzzles, much like narrative are not just quantitative: it’s not just a matter of hitting the right number to achieve success. As such, difficulty levels for puzzles are very rare.

Interesting point, Tom, but I wonder if there is a deeper psychological connection between the different forms of media. A movie or book without drama, without a central conflict, is simply not interesting. It appears that when we consume entertainment, our brains go through the ‘challenge’ of a dramatic journey, wondering what will happen next and trying to process the emotions of temporary setbacks. I wonder if that challenge, though passive, is not fundamentally so different from that of our more interactive forms of entertainment. Note how even in the story domain, the ‘challenge’ has to be ‘just right’ aka balanced: if it is too easy or superficial, the enjoyment seems lacking. I’m suggesting that we’re activating similar neurons for all forms of entertainment, and as others here have mentioned, the crossover becomes more apparent when you watch a person playing a game on twitch, thus turning an interactive experience into a narrative form.

Ah, yes! Challenge is a fundamental part of any story, isn’t it? And that’s where the interactivity of a game can be so appealing. You’re experiencing the challenge, the drama, the conflict, rather than simply spectating it.

I don’t want to push this comparison too hard, because it’s nonsensical, but imagine if you sat down to read Moby Dick and the very first page asked you to rate how obsessed Ahab is with the White Whale: a little obsessed, moderately obsessed, or From Hell’s Heart I Stab At Thee.

Well, sure. I love watching videos of boardgame playthroughs. It’s a drama as sure as any soap opera or Russian novel, and because it 's a boardgame, it has to focus on the player rather than any fancy graphics engine!


I find any uphill battle juicy and will revel in it, but I have a friend who is the opposite, he thrives on victory, and will pile onto that and mercilessly drive it in…however, if he gets into a spot of trouble, he is VERY quick to give up completely.

A multiplayer game with him is a difficult business for me at times.

I think I prefer resistance to frustration. A game without resistance isn’t a game, overcoming the resistance is what makes it a game.

Yeah, I know the feeling. We played Unrailed last night and 3 out of 4 of us were enjoying it on medium. It was difficult, but a fun challenge to (try and) cooperate around. My friend just seemed to be getting frustrated with it (and us!), and whenever we got back to the lobby he kept sneaking over and switching the difficult down to easy, and we then ran over and bumped it back up again. Anyway, eventually he caught us out and victory: we got to play on easy. It was sleepy: open fields, no real terrain to plan, cut and build through, no bottlenecks. We didn’t really need to communicate a great deal which, as far as I’m concerned, is a problem if a game is supposed to be co-op. He mellowed a lot at that point and even said he had his feet up and was idly playing. The game’s unique pressure cooker co-op is far too delicious for me to sweep under the easy rug!

That’s pretty much it. The only incentive I want or need from a game’s difficulty is sufficient pushback to force me to engage with the game’s (hopefully interesting and robust) systems and mechanics. Not enough pushback: I’m facerolling through and don’t need to think. Too much pushback: it’s frustrating. I’d much rather err on the side of too much pushback because that’s a more interesting space to learn and grow in (depending on the game).

Managing Expectations on Offworld Trading Company comes to mind. Anyone who has played it remembers how difficult that tutorial/trial was at first but beyond it lies untold real-time strategy riches.

I like that!

Earlier in 2020 I played The Dark Mod, a free Thief-like game, that allows you to adjust the guards’ sensitivity to sound, and maybe increase their visual acuity? One of the things that struck me was that, after years of hearing folk grumble about blind, deaf and dumb stealth game guards, if you crank their senses up it totally breaks the game, and I suspect it would in most other stealth games. At the very least you end up spending longer waiting for them to settle down and blame the rats or wind.

While I’m not generally a fan of having that kind of under-the-hood access (leave that to the designers), it was humbling to see just how game-breaking certain (popular) suggestions can be. That said, I’m all for accessibility options if it means more people get to play more games. Ideally, accessibility options shouldn’t diminish the intent of a design but if concessions have to be made then so be it.

I think back to SOMA and Frictional adding a safe mode where you can’t die. At first I was like ‘but that’s undermining the tension!’ (and I still feel that way to some extent) but then if more people get to experience that brilliant story, atmosphere and maybe a few ‘softer’ scares too, then I think that’s a great thing.

Also, Thief had the best difficulty options. Everything about the game stayed the same, you just got more objectives, had to find more gold, and weren’t allowed to kill anyone.

Yeah, of course stealth games the enemies are half blind for a good reason. In real life I can see someone 200 mts away, so ‘let’s make it realistic!’ would be a bad choice. :P

So basically, if a game designers includes two experiences in one game they failed? If they’d just done a call of duty-tuned shooter it’d be fine, if they’d done a spectacle game with minimal pushback, it’d be fine.

Would your response be different if they’d done both as separate releases?

It also changes how much health Garrett has, such that getting attacked makes you significantly less likely to survive the encounter. (But, unlike most games, you’re trying not to get in fights in the first place, it just gives you less margin of error.)

Yeah Thief is a great example. Even within this one game and given my own mood, I might want to take an easy tour through the game, or I might want to test my skills to the utmost. The same person, depending on the time of day, may wish to experience a game differently. And this, I think, is because games really do contain more than one system working simultaneously. Making the stealth system easier doesn’t affect the narrative system very much.