I applaud the intent behind Celeste’s Assist Mode that allows for tweaking aspects of the game and lets players of varying skill level and physical capability enjoy overcoming an appropriate challenge. And I hate to come across as complaining about it. But the fact is that Celeste is a game that I found very frustrating and I wasn’t able to fix that with Assist Mode - because Assist Mode doesn’t let you tune punishment.

Celeste is a precision platformer. You have a set of abilities: running, jumping, wall-jumping, wall-climbing, and an air dash. Some abilities are limited and get refreshed by standing on solid ground. You must use these tools to get through a series of platforming challenges in varied environments with their own varied mechanics, such as platforms that move when you air dash or midair gems that replenish your abilities without you needing to land.

Most challenges in Celeste really have two parts: the puzzle of figuring out how to use your limited abilities and the particular environment to navigate each obstacle course, and then actually executing your solution with precise timing and positioning. To use my own terminology, this is a tactical challenge (figuring out what to do) followed by an action challenge (doing it). They are difficult in different ways and can separately be interesting/dull or hard/easy to individual players.

This is risky, because it means a player has to enjoy and be sufficiently competent at both the tactical and the action challenges in order to enjoy and progress through the game. Someone who likes charting a path through each screen but then lacks the reflexes to actually follow that path is not going to have a good time.

For me, a lot of Celeste’s difficulty felt unintentional.

First, some background about me: I have a terrible sense of direction. It’s hard for me to build mental maps of areas and to visualize where locations and landmarks are in relation to each other - and thus to figure out how to get from one place to another.

In the neighborhood where I grew up, I was once asked for directions to a building that was literally next door to where we were standing. I pointed in the wrong direction. This is not an atypical example.

I’m starting to think that most of the heated debates that happen around game design choices are due to poorly-understood differences in how our brains are actually wired.

Like, I’ve written before about how some people hate punishment in games and others don’t and how this seems to be related to how we process tension, and how it’s easy to think someone else is a wimp or a masochist for the type of gameplay they like when it actually feels different to them than it does to you. But I realized there are other factors here too - punishment is worse for players who have trouble focusing on things that aren’t novel, which, like… that’s straight-up an ADHD symptom, right? I’ve never been diagnosed with ADHD, but I’ve got a couple of symptoms including that one. Allie has more symptoms, and she’s even more bothered by punishment and repetition in games than I am.

I’ve also talked about how I don’t like games that make you work to find the quality content in exchange for a sense of discovery that rings false for me. But when I saw the following mailbag question in a Shamus Young diecast post, I realized there was something else going on:

Dear Diecast.

The modern Persona games are lauded for their fusion of turn-based combat and social sim gameplay, but I’ve always been bothered by the social sim aspect. It’s less about roleplaying and more about puzzling out the spreadsheet nightmare the designers have conceived so you don’t miss out on story content and have to replay it in new game plus to see it. As such, I always play them with my head in a guide to negate the issue so I can instead focus on enjoying the combat and story.

What’s your thoughts on games that are hard to play properly without using a guide and have you ever found them enjoyable in spite of needing to look things up constantly?

-Victor

My immediate thought was that yeah, I feel the same way about Persona and that this kind of design is stupid in general as just another way to make you work to find the quality content - but I made myself take a step back. It’s not very likely that the designers of several incredibly-popular games are all just making the same obvious mistake over and over and the fans somehow don’t understand the resulting flaws. It’s much more likely that this is another case where players have different but legitimate preferences.

Victor’s question has assumptions baked in - that if you “miss out on story content” you then “have to replay it in new game plus to see it” and that seeing all the story content is the only way to “play properly”. I didn’t notice at first that these were assumptions, because I’m a completionist so to me (and I imagine Victor) they just feel true. Like Victor, I find it hard to enjoy a game if I’m constantly worried that I’ll miss content - particularly story content - particularly if it’s a story I’m enjoying. Like Victor, I often deal with this by using a guide and then lament that the game “requires” a guide.

But like… that seems like something in the area of anxiety or OCD, maybe? I’m not sure exactly what the divide is here, but roughly speaking I suspect some people prefer certainty and control (the completionists) and others prefer exploration and surprise. For the latter group of players, the fact that it’s possible to miss some story content based on your choices is a bonus - it means that you can actually be surprised by what you see, even if you return to play the game again. To me, this is a baffling way of looking at things - but some quick internet research shows plenty of evidence that some people like surprises, some people hate them, and many people in each group do not at all understand the people in the other.

A lot of us have trouble explaining what happens in our own heads, and it’s difficult to realize when something you thought was universal is only true for people with brains like yours. And it’s really hard to see where someone else is coming from if your disagreement stems from one of those things. A lot of the time we’re arguing about things like game design decisions, we’re being much more subjective than we realize, and that leads to heated and unproductive discussions that say more about ourselves than the thing we’re trying to talk about.

In short: its high strictness and punishment plus its regressive difficulty and locking mechanics behind power-ups make it frustrating to learn to play.

I’ve never really gotten into mainline Mario games, but I was intrigued by Super Mario Maker 2’s story mode, which apparently serves as a sort of extended level design tutorial. It features 120 levels each themed around particular level pieces or combinations thereof, showing you how to use them in play and hopefully providing inspiration for how to use them when creating your own levels. I find tutorial design really interesting, and Mario famously teaches through level design, so I checked it out.

Muse Dash is a rhythm game with a simple control scheme. Aside from menu navigation, the actual rhythm gameplay only requires two buttons.

The player character is on the left side of the screen, running constantly to the right. Threats come from the right side of the screen in one of two lanes - at ground level, or above it. When the threat reaches the player character, the player is supposed to hit a button to “knock back” the threat - one button for threats in the top lane (by default, any button on the left side of the controller will do) and a second button for threats in the bottom lane (right side of the controller). There are various twists and complications laid on top of this - threats that require hitting both buttons at the same time, or holding one or both buttons, or mashing the buttons repeatedly, and so on - but they are all dealt with using just two buttons.

(This is how it works on PC and I assume on Switch as well. Probably on mobile you tap either the left or right side of the screen instead?)

My gut reaction to seeing this was to assume that Muse Dash must be simple and easy, but there’s no actual reason this has to be true. In fact, Muse Dash gets quite difficult in most of the standard ways. Reducing the number of buttons used changes very little. After all, thinking back on all the rhythm games I’ve played it’s rare for the player to be required to press more than two buttons at once. Consider two scenarios:

1. In Muse Dash, a threat comes in the top lane and then one comes in the bottom lane.
2. In Hatsune Miku, there’s a prompt for D-Pad Left and then one for Circle.

The actual actions taken by the player here are very similar.

1. Read the on-screen cues.
2. Recognize you’ll want to press a button with your left hand and then one with your right hand.
3. (Miku only) Remember the controller layout and move your thumbs over the correct buttons.
4. Press with your left thumb and then with your right thumb.

The only real difference here is that Muse Dash doesn’t quiz the player on the controller layout or force them to move their thumbs. Lanes correspond directly to hands - top lane threats mean pressing with your left thumb, bottom lane threats mean pressing with your right thumb.

Personally, I’ve been taking rapid-fire quizzes on the PlayStation controller layout since 2000, so the extra step of remembering which button is where doesn’t trouble me much and I barely notice it. But to someone new to the controller and its apparently-arbitrary arrangement of buttons, this step makes the game far less approachable. They have to memorize the controller before they can be effective at the game. There’s an extra source of difficulty up front, and it isn’t the thing that makes the game interesting. The player has shown up to feel the rhythm and before they can do that they must perform rote memorization.

Now, this might be worth it. Guitar Hero on Easy or Medium is much like Muse Dash in that the player doesn’t have to move their hand and the colored notes correspond directly to fingers. Hard and Expert difficulties use more fret buttons than the player has available fingers, so they have to move their hand. Accessibility aside, I’d argue that the game is better when it forces you to move your hand because the game is about pretending to play the guitar which in real life also requires moving your hand. It’s extra effort and difficulty that isn’t strictly required - the game easily could have been designed to use only four frets - but it’s directly tied to what makes the game interesting, which is the fantasy of being, well, a guitar hero.

But what about Hatsune Miku? Your actions in this game are abstract and not a metaphor for anything specific except music itself. What matters is that they are rhythmic and flow-inducing. So what benefit is gained by adding controller memorization to the challenge?

I’m not ready to conclude that controller memorization adds no value to Miku-like games, but maybe trimming it out as Muse Dash does is just letting the interesting part of the game be the hard part.

Guitar Hero is played with a guitar-shaped controller. The main inputs are the five differently-colored fret buttons and the strum bar. With one hand, the player holds down the fret buttons indicated by the differently-colored on-screen notes (in a simplified approximation of holding down the right strings at the right frets on a real guitar) and with the other hand the player presses the strum bar then when the notes scroll to the right point (in a simplified approximation of strumming the strings on a real guitar).

There are five fret buttons, but if you hold the guitar controller the standard way real guitars are held, your fretting hand wraps around the guitar neck and your thumb is stuck on the back. You’re thus left with only four fingers for the five fret buttons. This is presumably intentional - it forces you to move your hand up and down a little on the neck to press the correct frets, as you’d have to do with a real guitar.

But! This is only true once you get to Hard or Expert level play. On Easy, only the first three fret buttons are ever used. Medium increases it to four, and Hard finally uses all five. This means that the entire time you are playing on Easy or Medium and learning the game, you don’t have to move your fret hand. You are trained that red notes mean pressing your index finger, green ones mean pressing your middle finger, and yellow ones mean pressing your ring finger. (And on Medium, blue notes mean pressing your pinky.)

Then you start playing on Hard. Most songs don’t begin with an orange note, so you play the first several notes the same as you used to, and then an orange note appears. You slide your hand down one fret to press the orange fret button with your pinky. But then when a green note appears, what happens? What you should do is press with your index finger, because you’ve slid your hand one fret down. But you’ve been trained that green means middle finger, so there’s a good chance you’ll press with that instead even though it’s currently over the yellow fret button and you’ll play the wrong note.

For me and for the folks I’ve talked to, by far the hardest part of Hard mode wasn’t the increased density of the note charts or even moving to hit the orange notes. It was unlearning the habits taught by Easy and Medium so that once we had moved to hit the orange notes, we didn’t use the wrong fingers to hit the other notes. It was re-training ourselves to associate note colors not with fingers but with frets so that we could figure out which finger to use based on the fret and the current position of our hand. Once this was done, going up to Expert was comparatively simple.

I remember thinking at the time that while it was obviously correct to keep the note charts sparse for Easy and Medium, all five frets probably should have been used from the beginning. (And I think this is what later games ended up doing? I don’t know - the guitar skills transfer pretty well between games so I haven’t dipped back down to Easy or Medium in a long time.) And I’ve been thinking about this again because I’m on a bit of a rhythm game kick lately and have been seeing how different games treat their difficulty levels.

Guitar Hero’s error (if you consider it an error, as I do) was in treating two kinds of difficulty the same when they were actually very different. The first kind was density of the note chart, which is commonly tied to difficulty levels in rhythm games. This makes sense - a denser note chart requires a higher skill level, so as players get better they need denser charts to maintain flow (which I argue is the point of rhythm games).

While it may seem like the variety of different notes used is another source of difficulty that should scale similarly, I think this is a mistake. Learning the inputs and how they correspond to the game’s cues is most of learning to play the game. Guitar Hero withholding some of the frets from you at the start means it’s training you on an incomplete version of the game. That’s not inherently problematic, but you have to be careful about it or you’ll teach bad habits (like “color = finger” instead of “color = fret” in this case). Much like adding the double-jump in Runner3, adding the orange notes and forcing the player to move their hand doesn’t technically change the meaning of the other color notes, but it does mean that the simplest strategy for dealing with them and thus the one the player has likely internalized is now often incorrect.

The increased challenge of the denser note charts which more closely matched the real song being played and the forced fret-hand movement which more closely matched the hand motions of a real guitarist was good - this added challenge increased flow and immersion. This is a case of what’s hard about the game also being what’s interesting about it.

But challenge that comes from the fact that the game taught the player bad habits isn’t interesting in a game like Guitar Hero. It’s frustrating. You just deal with it until you unlearn the habits and the game can be fun again.

This is something that’s foundational enough to my beliefs about game design that I want to publish an article about it that I can point back to, but it’s also simple enough that I’m not sure there’s a full article’s worth to even be said about it.

What’s hard about a game should also be what’s interesting about it.

It’s the reason some reviewers disliked the focus on meter management in Pathologic 2, where the interesting stuff is the mystery and atmosphere.

It’s the reason why it’s frustrating to have your choices overturned by QTEs in Until Dawn, a game that sells itself as a game of decisions and not one of controller memorization.

It’s the common thread behind these house rules for card and board games that seek to eliminate memory- and inexperience-based challenge in order to emphasize interesting strategic-based challenge.

It’s the reason why my proposed changes to Akiba’s Trip combat are “specifically looking to reduce uninteresting difficulty - things that are hard for stupid reasons. This actually allows for increasing difficulty in more interesting ways.”

What’s hard about a game should also be what’s interesting about it.

Allie doesn’t play many action games. Her PS4 is mostly for Rock Band 4, though she also enjoyed Until Dawn. She’s now trying Dragon Age: Inquisition after hearing a lot of good things about its characters. Watching her play has been instructive. Here are some of the lessons I’ve taken away just from her first hour.

I mentioned how I’m frustrated when games don’t treat rewards as rewards. There’s a more subtle and complex example of this in TY the Tasmanian Tiger 2.

Ty’s central gimmick is his collection of ‘rangs (boomerangs) with varying effects. Some of them affect environmental objects, such as fire ‘rangs that can melt ice or ice ‘rangs that can put out fires.

In TY 2, the Bush Rescue HQ is a safe zone where you can experiment with your abilities, similar to Peach’s castle in Super Mario 64. It’s dotted with minor platforming challenges that reward small amounts of in-game currency. But some of them require ‘rangs you don’t start with.

Tucked away on a particular rooftop is a broken-down generator. It’s a little hard to find and a little hard to get to. A mobile platform carrying some currency hovers in midair nearby - but too far away to jump to. It’s obviously a tutorial for the electric Zappyrang that can start up machinery - you’re supposed to wonder how to turn on the generator, remember it when you get the Zappyrang, come back and use it to get the currency, and then be all set for generators you encounter later in the game. The currency itself barely matters - it’s a small amount, just there to make sure you have a reason to remember the generator.

So when I got the Zappyrang, I excitedly went back to HQ to turn on that rooftop generator. And sure enough, the nearby platform came closer… and then moved away again. It went back and forth and even at its closest, I wasn’t sure it was quite close enough to jump to. I tried over and over and kept missing. Each attempt was separated by taking several seconds to climb back up to the rooftop (longer if I didn’t turn back in time and fell off the hill the building was on and had to get back up that first) and then waiting for the platform to come closer.

I didn’t even care about the damn currency - I could easily get more in less time. It was the principle. It was the fact that I’d gotten the Zappyrang and remembered to come back and find the generator. It was the fact that the setup implied that was supposed to be the hard part. I’d earned the reward, and collecting it was supposed to be easy.

But I never pulled it off. I eventually gave up and moved on. I still don’t know whether I was supposed to be able to make that jump or if there was something else I was supposed to do.

So to me, that moment failed in three ways, each of which increased my frustration.

First, by teasing a reward that was then withheld. As a Zappyrang tutorial, collecting the reward should have been dead simple: activating the generator should have moved the platform very close by and left it there so that it was easy to jump to.

Second, by being overly punishing. The platform could easily not have been positioned such that jumping for it meant going off a cliff if you missed and didn’t turn around in time. There was no reason for each attempt to take so long to get back to - this wasn’t part of some kind of endurance or mastery challenge.

Third, by providing unclear feedback. A ledge that you can almost jump to is actually the exact example I used to illustrate why challenge profiles should be clear - I had no way to tell whether I was supposed to be able to make the jump or not. It seemed unreasonably hard, but I saw no indication that there was anything else I was supposed to do to get onto that moving platform. If there was in fact another step, that should have been made clear; otherwise the jump should not have been borderline-impossible.

By itself, this was not enough to make me stop playing. But it absolutely decreased my confidence that I could trust the game’s designers to provide a quality experience. And when I ran into other issues later on, I considered whether they were worth dealing with and found that I was unwilling to give the game the benefit of the doubt. I put it down and haven’t looked back.

Much of the confusion in the difficulty debate is because we often talk about difficulty when what we really care about is focus.

Take Pathologic 2. It’s a game about trying to survive and solve mysteries in a plague-ridden town. It’s got a strong narrative and prominent survival mechanics. The reviews I’ve seen lament how much the latter gets in the way of the former. From Brendan Caldwell’s review:

There are a bunch of meters: stamina, hunger, thirst, health, exhaustion, immunity. . . . On paper, you should be exploring the town, figuring things out, piecing this unearthly story together. But most of your time is spent trading bits of old brain and rusty scissors for a tin of food just to keep a meter down. It takes over everything, an unwelcome distraction from the intrigue of murder cases and bizarre architecture of the town’s stranger buildings.

After receiving a lot of feedback to this effect, developer Ice-Pick Lodge is adding a “difficulty slider”. Their announcement post is worth reading in its entirety, but the takeaway is “[W]e’d rather give people a tweaked experience than none at all.”

That’s great, and I applaud Ice-Pick for responding in this way. But I also find myself wondering - is this really about difficulty? Here’s another quote from Brendan Caldwell’s review:

If the intent here is to follow a Soulsian “hard is good” philosophy and apply it to the survival genre, this is misplaced. Souls games are about reaction, movement, and practice. You can’t practice finding a piece of bread. If the intent is only to keep the player feeling oppressed, strapped for time, exhausted, hungry and weary, well, that doesn’t mean I won’t also resent having to spend so much time doing the most boring species of meter-management in what could have been a captivating mystery.

Caldwell doesn’t mind that the game is hard or even that it’s dark - he minds that he has to devote so much time and attention to the least interesting parts of it. The bits that aren’t this game’s unique selling point, the bits that have been done many times in many games.

Ice-Pick haven’t said exactly what their new slider will do, but in practice for players like Caldwell it won’t really be a difficulty slider so much as a mechanical focus slider. Turning down a game’s ‘difficulty’ rarely means that characters now speak in Basic English and complex moral decisions get replaced with clear-cut black and white choices. What it does mean is that the survival mechanics (or in most games, combat systems) are less demanding and you can pay them less attention.

The difficulty debate (and certainly any response of the “you cheated not only the game, but yourself” variety) often misses that there are multiple different reasons to play most games. In addition to skill levels, physical capabilities, and amounts of free time, people differ in their interests. When someone is drawn to a game’s world, art, characters, story, etc., and finds that mechanical systems are getting in the way of what they came for, it’s natural to want to adjust the game’s focus to reduce those obstacles. Because of the misleading way these settings are often labeled, that means they are “playing on Easy.” But they aren’t here for the combat, or for the survival mechanics, or whatever. Mastering it is of no interest to them and is not worth their time. Like Peter Gibbons, it’s not that they’re lazy; it’s that they just don’t care about this particular source of difficulty. They’re trying to focus on what actually appeals to them and what actually creates value for them.

A bit ago I speculated on why different people feel differently about punishment in games, and I have a new theory thanks to Arnold’s levels in Bubsy: Paws on Fire!

The short version is that different people have different thresholds for maintaining interest in repetitive content. The long version follows.

So, one of the positions in the whole Sekiro/Soulsborne/difficulty/easy-mode debate is what I would sum up as “easy mode should be there for the players who need it.” A couple of examples:

• Jarrod Johnston explains why he’s no longer anti-Easy, mentioning that previously he “never thought that someone might not have the ability to improve, and it took a while for [him] to understand that getting through a tough game can be more than just a matter of time and commitment.”

• Mark Brown explains his complicated view on questions like “Should Dark Souls have an Easy Mode?”, making a few statements like, “I’m in favor of giving people more ways to play if they need them…” but emphasizing the importance of making sure players know what the intended experience is. He also describes Celeste’s Assist Mode by saying, “This isn’t an easy mode that I might switch down to if the game’s kicking my butt. It’s an assist mode that’s just there for those who really feel they need it.”

On the surface, I’m in agreement with these folks - we all want easy modes or similar accommodations in games. We’re on the same side of the overall debate. But I’d sum up my position slightly differently, as “easy mode should be there for the players who want it.” It’s only one word of difference, but I think it’s an important one.

I think my ironically-favorite part of the discourse around difficulty and easy modes in Sekiro and Soulsborne games is that anti-easy folks make both of these claims:

1. Adding an easy mode to a Soulsborne game would ruin the intended experience of overcoming challenge through persistence and learning.
2. Soulsborne games already have an easy mode since you can summon a friend.

I don’t know if I’ve heard any individual person say both these things together - the kettle logic might be a little too obvious if you’re actually saying “Easy mode would be bad, and anyway they already have it and that’s good!” But I also don’t think I’ve seen anyone really address the contradiction here.

So, a failure mode with buffs is when they present obligatory inscrutable math problems. Buffs aren’t the only thing that does that - this is an example of a more general failure mode in games. Jay Wilson, game director for Diablo III, calls it the “spreadsheet moment”:

There is a point I call the ‘spreadsheet moment’ where the variables have become so complicated that the average player can’t make a decision about their character or an item in their head. They require a spreadsheet, not for perfect optimization, but just to get an idea of what they’re doing. I’ve been known to map out abilities for a game I love in a spreadsheet for optimization. As a designer I want that level of depth in a game I make, but there is a point where you dump so much math on a player that they’re no longer capable of just making a decision and running with it because it’s what they like. When that point occurs we all portal back to town and alt-tab to the internet so someone who is better at math can make the decision for us.

Gamers are smart. They can feel when the math has gotten so complex they can’t process it in their heads, and the reaction is to assume that there ‘must’ be a clear correct singular decision. We desire that correct decision because humans like to apply order to things that seem chaotic. When we’re that sure of a singular right way to do something, no one wants to be the dummy who did it the “wrong” way. Unfortunately when players enter this mindset they’re no longer making the choice they want to make, for the sake of fun, but rather the choice they feel they have to make for the sake of being right.

I would argue there’s more going on here than just the complexity of the math - as I mentioned, Valkyria Chronicles grades and rewards you based on your efficiency, making it increasingly painful to make “wrong” decisions in combat - you’ll have fewer resources for the next mission making it even more important to avoid “wrong” decisions in a self-reinforcing cycle. But in a game like Borderlands an enemy drops the same loot whether you kill it in five seconds or six, so it’s less important to pick the absolute best gun. So even though the math on which guns are better is arguably more complex than the math on whether you should buff your soldiers, I’m much more comfortable eyeballing the guns and proceeding based mainly on feel.

And the clarity of the math matters a lot too. Explicit numbers with obvious meanings can still interact to create the “spreadsheet moment,” but if the numbers or their meaning is obscured, things get worse. As Shamus Young put it:

A good system is clear and understandable right away, but will reveal interesting trade-offs once you get to know it. (Diablo 2)

A boring system is straightforward and there is always a single, optimal answer. (Most BioWare games, at least until they abandoned gear-hoarding altogether.)

A bad system is one where you have no idea what decision you’re making and you have to do a bunch of homework and run a spreadsheet before you can know which hat you should wear.

A horrible system is one where you have to read the wiki just to know what options you should be researching. A system where you have over twenty attributes and most of them are synonyms. A system where you can read the wiki, a forum thread, and a leveling guide, and still not really have any idea what you’re doing because nobody agrees on how the stats work, which ones might be bugged, or if all of their calculations were made obsolete by the latest patch.

It’s easy to miss this amongst the other complaints, but there’s another key factor here as well - “interesting trade-offs” are part of a “good system” in Shamus’s nomenclature. He said this as part of a discussion of stats in World of Warcraft, where he argued that complex stats were particularly obnoxious because they generally boiled down to a simple trade-off of “efficiency versus survivability”. In his view (which I agree with), “you should never have more complexity than depth” and should present “decisions [which] are easy to understand but difficult to make” instead of ones where you have to do a bunch of research or math to find the unambiguous but uninteresting correct answer.

Extra Credits refers to the difference in these kinds of decisions as “calculations” (which are fundamentally math problems to optimize specific variables like DPS with options that can be cleanly quantified and ranked) and “incomparables” (which are more qualitative trade-offs like between a stealth or or combat build with no objective best option). With incomparables, players must make an actual choice based on personal preference and play-style - but we run into problems when games “mask calculations as incomparables”. Trying to make a calculation more interesting by making it more complex doesn’t stop there from being an objectively correct option - it just makes it harder to tell which one it is.