Let’s Talk About Difficulty

Super Mario Maker is a pretty amazing thing. Nintendo gave everyone the ability to make their own Mario levels and the results have been fascinating. You have some really neat and innovative levels, some really well-made traditional levels, and even some phenomena that don’t make any sense, like autoplay levels. It’s great to see so many people given their first chance to explore game design.

Of course, as is the case any time you try something new, many budding designers have made some noticeable mistakes. It’s part of the learning process, and I did a lot of the same things when I was making my first levels years ago. So, I’d like to extend a helping hand and discuss difficulty a bit today. If you play Mario Maker, particularly the expert version of the 100 Mario challenge, you’re inevitably going to play a level designed to be devilishly hard. I think this in itself is great; there’s a subset of the population that loves a good challenge, and Mario Maker is a great avenue for advanced players to finally play levels that challenge them.


For everyone else: welcome to hell

The problem is that most of the levels aren’t fun. There are levels that kill you within seconds of the level loading, levels that overload the screen with enemies, and perhaps most frustrating, levels that rely on liberal use of hidden blocks to either kill the player or to act as little puzzles with arcane solutions.

Ultimately, the problem is that the player does not feel in control of the actions. These deaths are the fault of the designer, not the player. After all, how could the player predict that they would have to run within seconds of the level loading? Or that there was an invisible block placed right in the middle of that jump sendingmecareeningINTOTHECLIFFANDARRGGH!

Ahem. Sorry.

It may sound strange, but players want to feel that every failure is their own fault. For players who enjoy tough games, their own failure is seen as a challenge, and they persevere only because they know that they can personally improve and overcome the obstacles.



So without further ado, here are some tips for making a tough game or level that players will find fair and actually enjoy:


Give your players proper time to prepare and react

Starting with the obvious: you need to give your players proper time to react. In other words, if the player is going to need to change course to avoid danger, they need to see that danger far enough in advance that they can properly process it and make an informed decision. This of course also means not immediately placing the player in danger the moment a level begins; the screen has likely just transitioned, so asking the player to process an entire screen’s worth of information to decipher your hazard within a split-second is usually too much for most players. Consider that even Super Hexagon, a difficult game that can make surviving 15 seconds feel like an accomplishment– begins each round without any dangers on the screen. If Super Hexagon can show that much restraint, so can you.

For games that don’t want to kill you in seconds, this means that the potential ramifications of your long-term decisions must be made clear at the time you are making the decision. You shouldn’t get far into a game only to realize that you should have levelled up combat instead of speech because there’s no talking, or that you should have invested in harvesting resource A, which is crucial late in the game, rather than resource B. This doesn’t mean you cut all uncertainty or customization out of your game, but that any inequality and potential risks in your choices needs to be communicated up front.

Regardless of speed, information must be easy-to-read for the player. The faster the player needs to react, the more important elements must contrast with everything else to stand out. Look at the Mario Maker image below. No rhyme or reason to what’s occurring on-screen, making it very hard to tell what’s going on.



Contrast this with an image of Geometry Wars below. I selected Geometry Wars because the screen is very dense and fast reaction times are a must. Yet still, if you look at the image, the colors and shapes make it very clear what’s safe and what’s not. Additionally, you can see clear groupings of enemies and a couple clear choices for the player’s next move. Sure the player could try and punch a hole in the enemy forces and go straight through them, but it seems smarter to go either right or downward, taking a few enemies as they go.



Even Super Hexagon, which appears to break this readability rule by constantly swerving the screen to and fro makes some important concessions. Starting with the obvious: the simple color palette makes it abundantly obvious what’s a threat or not. And like the color palette being limited for readability, the motion is similarly limited. The screen rotates at the same speed in one of two direction and always waits a few seconds between direction changes to give the player a chance to reorient themselves. And since the hexagons move at a consistent speed, even if the player gets disoriented, they can mentally time how long they have to react in order to dodge the hexagon.


Ease players into your mechanics…

I’ve already been over tutorialization, but just because a game is supposed to be difficult doesn’t exempt it from the usual rules governing a good tutorial. At all.

In particular, I want to draw emphasis to the idea that you need to teach your player how to play in a safe environment. The player should be able to demonstrate a skill without risk of failure. The first level of Super Meat Boy has no danger whatsoever, despite being very willing to kill you shortly thereafter.


Look how innocent that first level is. There’s even a bunny!


In Demon’s Souls, you can’t encounter your first enemy until you’ve figured out basic movement. It then displays the attack controls on the ground before you have a chance to engage the enemy. And even though they do teach you basic attacks on a real enemy, it’s one who moves incredibly slow, is hestitant to attack,  barely takes any damage, and goes down quickly. Of course, anyone who has played Demon’s Souls knows that the end of the tutorial involves facing a very tough enemy who almost inevitably kills you, so it’s not as though a game needs to be kind much beyond the initial learning period. Which leads me to my next point…


Be consistent

Set your rules and stick to them. If you’re going to make a hard game, make it hard the moment the player has a basic grasp on the mechanics. And keep your playstyle consistent. If the player must perform with fast reflexes and high precision, then they should have to begin demonstrating basic capabilities in those skills early in your game. Alternately, if the player needs to be careful and observe their environments at the hardest part in the game, then you should cultivate this pace early on. It’s frustrating to have a laid back game suddenly begin demanding precision, or to have a fast-paced game suddenly come to a halt because the designer wasn’t crafting consistent challenges.

If you want an example of consistency done right, look no further than the Souls games. These games get away with so many things that other games wish they could get away with. From the first bonfire in Dark Souls, there are precarious paths along cliffs into which you can drop, tough enemies if you head the wrong way, and ambushes waiting around corners even if you head the right way. Most games couldn’t get away with any of those things, let alone all three, but because Dark Souls is consistent in using these tactics to set a deliberate pace, it works.


Welcome to Dark Souls. You’re going the wrong way. You died.


Keep your penalties low, especially when it comes to the player’s time

When you make a tough game, you naturally expect your players to fail a lot; which is why you need to make sure that the actual act of failure doesn’t drive players away from your game. If a player is going to lose a lot before they win, you shouldn’t hurt them for failure. In particular, don’t take resources away from them. Mechanics like lives and money loss tend to discourage players more than anything, because they make failures hurt worse and can even turn victory into inevitable failure down the line. You finished the level? Congratulations, but you only have one life left, so you better get the next one right on your first go. Good luck!

Even if the player doesn’t lose anything for failing, you have to be careful to manage their time. If you put your checkpoints too far apart, or force the player to endure a lengthy loading screen after deaths, that tends to produce similar effects of frustration. If the player spends more time working through parts of the game they already completed than trying new challenges, they will be frustrated.

It may be cliche at this point, but Super Meat Boy is a great go-to example for how to do this properly. Levels rarely take more than a minute, and those are the longest ones, and when you die it takes about a second before you’re going again without any life penalty (at least in the main game). This frees players up to fail again and again, and to experiment on different approaches without penalty, resulting in an experience that is challenging rather than frustrating.


Bonus points for replaying all of your failures in hilarious fashion when you do complete a level.


A different approach comes from roguelikes, which don’t let the player retry at all when they die. This may seem counterintuitive, since surely not retrying at all is worse than a limited amount of retries. However, in roguelikes the challenge is randomly generated each time, so players face a new obstacle in the early phases of the game each time, and they rarely take too long to complete a successful run. Thus, even though the penalty may seem high on paper, these games are usually pretty good at respecting the player’s time and the player isn’t forced to repeat large sections of the game.

Even the Souls games obey this principle to some degree, although it certainly doesn’t seem like it. Levels in the Souls games are long, often taking an hour to beat your first time, and when you die you lose all of your souls (ie: currency and exp). In Demon’s Souls and Dark Souls II, your maximum health is even limited after you die. These are significant loses in both resources and time, and understandably repel some players from the series.

But it’s not actually as bad as it seems. Those hour long levels are rich with shortcuts that allow you to skip large chunks of content on repeat plays, knowing where enemies are saves you time, and some of the stronger enemies don’t even reappear if you eliminate them once. All of which allows the player to regain their progress in minutes if they are careful and observant. And the soul penalty? It seems bad, but the game allows you to recover your souls if you play carefully and don’t die before you retrieve them, most bosses give you large amounts of souls right before a place where you can spend them at no risk, and player skill is emphasized more than level, allowing some crazy people to beat Dark Souls at level 1.

As for health penalty? Yeah, that’s a pretty poor design choice. Still 2/3 isn’t so bad.


Curses are also a jerk move.


It all comes down to respect. Players may like a good challenge, but disrespect from the designer is an immediate turn-off. All of the tips above extend from the basic question: does this design decision respect the player and their time? Does the player have ample time and information to properly react? Did the player have enough time to learn that mechanic before they had to use it? Have I acquainted the player with that style of play early enough in the game? Are they able to retry my challenge without wasting a bunch of time or resources? If you can answer those questions yes, then you’ll know you’ve done challenges right.

So for the love of god, can we please stop posting Mario Maker levels that look like this?


Pretty please?

Add Comment