| “aw darn i was gonna write this”

                                     -Potatochan

TL;DR: This is mostly images and is not as long as it looks. If you’re still not sure, at least take a look at the introduction and the summarization of the key points before you go and check out the full description for any points that catch your eye.

Trolling for Dummies

By Defender1031

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

With feedback from

Aqua Arrow

ArgentRyins

Bedcliff

BloodFalcon4

CilantroGamer

Coolin_the_pikachu

DankDrew

Dboy1004

Dooglus

Drakyen

federico378 (aka Paulo)

Foggles17

Geek (of FreakAndGeekGames)

Jjboy01

Jjtheneko

Jurween

MarcyAugust

Martii_1205

Metalooney

Muskwalker

Nicotec

Nightbot

NikWillOrStuff

Nofeetmcgee

Not_Pagle

Pacmanboss256

Potatochan789

Pulse

RexRath

RoyceRemix

Rraeann

Sandywitch

Seanjank

ShockolateChip

Showsni

Stay_hydrated_bot

Stevo8

Suspicious_door

Symplectic

Thatfuggingguy

TheStellarStrix

Trollmakerdave (aka Troll)

Vincepor1

..and of course, the one and only CarlSagan42

Introduction

So it seems as though I've come to be viewed in the Super Mario Maker community as some kind of a troll level expert. As someone who sees himself as an ordinary ←Dummy conducting ethically questionable psychological experiments on unsuspecting SMM players, I'm still somewhat shocked at being considered an expert. But because many of you have asked for it, I'd like to share some of the observations of human nature which resulted from my unlicensed psychological research in the hopes it will encourage other level makers to perform their own unethical experiments.

I realize that this is somewhat long, but if you’re a new creator, I do encourage you to take the 30 minutes to read the entire thing, as it will likely save you weeks, if not months, of the learning experience it would take to discover these things on your own.

This document is a perpetual work in progress and will continue to be updated as the community learns even more about making good troll levels, and as new techniques are discovered. Feel free to DM me on discord (@Defender1031) with feedback, corrections, or suggestions for improving this guide, or even if you just want to chat. I’m serious, by the way, I’m not just saying that. Really. I’m here to make friends. Come say hi.

I also want to mention that as this has been a learning experience for me as well, you will certainly find examples of things that do not conform to the following guidelines in every level that I've ever made, (and frankly, in every level that I ever WILL make). I didn't start off knowing these things, and even now that I do know them, I'm still an imperfect human being, and will continue to make mistakes. So if you too are interested in being an imperfect human being who makes troll levels for other imperfect human beings, please join me on my imperfect journey. Without further ado, let's get started.

Glossary

If you’re familiar with the Super Mario Maker community and especially Carl’s stream or videos, you likely already know these terms and can skip this section. For anyone new to the community, welcome on in, and here’s a list of terms used frequently in the community:

50/50 - 90 Rule - Carl’s rule which states that a skilled troll creator can set up a 50/50 choice between two options such that 90% of players will choose wrong.

Antitroll - A setup that makes the player think for a moment that they’ve been trolled, only to find out that they’re still completely safe. Some creators even make entire antitroll levels.

Barb - Twitch streamer/YouTuber BarbarousKing

Benevolent Level Creator - Often used as a comment to refer to situations in which a troll creator visibly made something less difficult than they could have, or visibly made a troll non-lethal where it could have been.

Big Brain Play - An attempt by the player to outsmart the troll creator, often by picking a non-obvious path. Commonly thwarted by a Bigger Brain Play, in which the troll creator has made the non-obvious path a troll as well. See also “50/50 - 90 Rule”.

Carl - Twitch streamer/YouTuber CarlSagan42.

Carlbox - A spot where a player will be trolled if they stand there for too long waiting or a box into which a player is tricked or pushed and from which there is no escape. Often actually marked with a box outline.

Cheese - An unintended method of getting through a part of a level, including unintended methods of getting extra items or gaining access to areas of the level that are supposed to be inaccessible.

Controlled Jump - A jump where the player needs to let go of the jump button in order to not jump to the maximum height.

CP - Checkpoint.

CP0 - A checkpoint that puts the player back at the beginning of the level after they’ve already collected another checkpoint, reversing their progress.

CP1 - The first checkpoint. Can also refer to a troll setup which sends a player back to the first checkpoint after having gotten the second, reversing their progress.

CP3 - Another way of saying “a CP1 troll”, owing to the fact that, barring glitches, Super Mario Maker only allows two checkpoints per level.

Damage Boost (Also sometimes referred to as Invincibility Frames or I-frames) - Using the short time while mario is invincible after getting hit and losing a powerup to make it through a saw, over some spikes, or through some other obstacle that would normally kill him.

Dev Exit - An intentional alternate method of beating the level used by the creator to clear a level for upload without having to play it the intended way.

Disrespect Block - A moving block on a track over a single-block gap designed for the player to go into and get crushed.

Disrespect Hint - A hint which taunts the player with how to avoid the troll they just fell in, when the setup itself had already made it clear without the hint.

Dick Vine - A common trick used in SMM1, with a vine placed behind a door that a player will end up climbing rather than entering if they hold up too early. Not often referenced in SMM2, as the layering order was changed and vines now appear in front of doors rather than behind.

Doodads - The decorations that occasionally pop up out of ground tiles when they are placed, term popularized by streamer Chichiri12345.

Entity Limit B (often shortened to ELB after a video by Ceave Gaming popularized the term and the tech. Sometimes called the Sprite Limit) - The limit the game places on the number of movable objects that may be spawned at any given time. Often used as a technique for controlling contraptions, especially by forcing the collection of a red coin to count against the limit.

Hard Cheese (Or The Hard Cheese Problem) - An unintended solution which is more difficult than that which is intended. Often mistakenly thought by beginners to be of no importance, the reason it is problematic is expanded upon in section 4K

Hot Garbage (Often shortened as Garbo) - Poorly-designed levels which are full of things like random hidden blocks, enemy spam, pick-a-path, doors or pipes leading to instant death, and other things typically associated with chaotic level design.

Jank - Inconsistent behavior of elements in a level based on spawn position, RNG elements, movement speed, or other chaotic or non-deterministic factors.

Juz - Twitch streamer/YouTuber JuzCook. He’s Australian, mate.

Kaizo - Difficult platforming, especially involving item abuse such as jumping off of shells while airborne.

Kaizo Block - Hidden blocks, often over a pit or other hazard such that an unsuspecting player hits the block and falls into the trap.

Keydeath - Dying after obtaining a key from collecting all of the red coins in a level, in a case where some coins were collected before the most recent checkpoint. Because of how Mario Maker works, this will destroy red coin progress and force a player to recollect all coins.

Knowledge Checkpoint - A point in the level where the player is given a VERY CLEAR indicator letting them know that the correct way to go was a tricky solution at some earlier point in the level. Often used as a way to shorten sections without the use of additional checkpoints.

Meta - The way current community trends and expectations from previous levels suggest how a given troll will play out. Constantly shifting as expectations are violated.

Nai - A failure just before the end.

New Soup (Also New Supe, the clearly superior spelling, or New Soups, as Juz calls it) - The New Super Mario Bros. U theme, stereotypically known for terrible speedrun levels and the worst kinds of hot garbage.

Poo - Twitch streamer/YouTuber GrandPOObear

Recycled Checkpoint (Also Infinite Checkpoint, Continuous Checkpoint, or any other phrase that implies Checkpoint Reuse) - The mechanic by which a level with more than two progress-saving sections can be created by placing a red coin in each section and either alternating or presenting both checkpoints together after each section to save the player’s red coin progress, with a key door that leads to the end of the level.

RNG - Game effects that are controlled by game randomness. Sometimes used also to refer to effects that vary in an unpredictable and uncontrollable manner based on the player’s movement.

Runback - The section of the level through which the player has already played, and through which they need to play again each time they die to get back to where they were before.

Safety Coin - An extra red coin placed near the key door in a red coin level (usually a recycling-checkpoint red coin level) to ensure that there is one uncollected coin until the very end to prevent keydeath.

Shinies - Intentional distractions as the setup for a troll, such as a question mark block that might be either good or bad and makes the player hesitate for long enough to stay in the Carlbox.

SMB1 - Super Mario Bros. 1 theme.

SMB3 -- Super Mario Bros. 3 theme.

SMW - Super Mario World theme.

SM3DW (Also 3DW) - Super Mario 3D World theme.

SMM - Short for Super Mario Maker.

SMM2 - Short for the same thing as SMM is short for, but with a 2 at the end, of course.

Softlock - When a player is in a position where there is no way to either die or complete a level and their only choices are using the “Start Over” button or letting the level timer run out. An Antisoftlock on the other hand (sometimes shortened as “Anti”), is some mechanism the player is given to die in an area which would otherwise be a softlock. This is often used for trolling by giving the player a mechanism where it’s possible to die but somewhat challenging to do.

Spaghetti - Failing to execute the necessary maneuvers to complete a section, especially repeatedly.

Strat - Strategy. Often used to mean a method of performing a particular section, such as a specific series of inputs or a specific route.

Troll Density - The amount of trolls layered on top of each other in the same setup. Often used when the troll density is too high and doesn’t allow the player a chance to breathe.

Twice Twice - An area of a level which looks identical to a previous area upon first glance, but has different trolls with different solutions.

Uno Mas - A short level, usually on a single screen, that showcases and forces the player to figure out a single unexpected game mechanic or interaction.

x=9 - Glitchy tech, specifically originating from some janky behavior when certain sprites are placed in the 9th column of the level from the left.

Yatta - Success.

YOLO Jump - A blind jump off an edge trying to make it as far as possible in the hopes that there’s ground on the other side to land on.

Yump - A silly way of saying “jump”. Also sometimes used to refer to a controlled small jump, often a botched or accidental one. Originally used to refer specifically to jumping off a switch button in Super Mario World.

Z - The 26th letter of the alphabet. Tracks arranged as a Z are often used as an indicator to spin jump (owing to spin jump in Mario Maker being generally performed using one of the Z triggers on the controller), or, in a troll level, as a humorous reference to this phenomenon, often in a place where it doesn’t make sense (such as SMB1 or SMB3) or to disguise a troll which will make use of the tracks.

Key Points (Table of Contents)

Click on a key point or go to the page with the indicated number for a more detailed description of each point and why it’s important, or skip this section entirely to read each point along with its explanation.

1. Mentality Matters        9

Always remember that your goal is to make the player laugh and never to frustrate them.        10

Put yourself in the player's shoes. Look at your level as if you don't already know the answer.        11

Playtest and give feedback on a lot of other people’s levels.        12

Be aware of the "frustration meter".        13

Err on the side of too easy.        14

Be prepared to sink a LOT of time into each level.        15

Be aware of the anatomy of a troll setup.        16

2. Clear Communication is Key        17

It’s the creator’s job to communicate to the player how to complete the level.        18

Make it clear what a player’s options are.        19

Arrows, sound effects, coin trails, track boxes, and other indicators are often not enough to clearly communicate.        20

Don't make obscure or convoluted solutions, or solutions that can’t be easily found without using the editor or receiving hints from you as the creator.        21

Try to make it clear immediately what a player did wrong after a troll.        22

Don’t blame the player. Ever.        23

Avoid large amounts of invisible stuff that can’t all be checked in a single pass.        24

Every element you place in the level should have a reason for existing.        25

Avoid delicate or janky solutions that can be performed and yet the player still fails for not doing it good enough.        26

Players generally have tunnel vision and don’t look outside of the immediate area to which they have access. If something important is out there, make sure to draw their attention to it fully.        27

Hints rarely work, try to avoid them as much as possible. If a hint must be used, make sure it can't be mistaken for anything else.        28

Try as much as possible to keep setups for contraptions offscreen and inaccessible.        30

It's okay if people don't fall for every troll, but make sure that players know the purpose of all visible elements which were part of a troll they missed.        31

3. Linear Levels are Less Loathsome        32

Don't have too many paths or options open to a player at once.        33

Keep the open options at any given time part of a cohesive whole.        34

Make sure the decision a player makes feels like a choice rather than chance.        35

Fence the player into the path you want them to follow.        36

Limit the amount of backtracking a player has to do.        37

4. Deceptive yet Docile Design        38

Fight the instinct to make traps that are easy for a player who already knows about the trap to fall into accidentally.        39

No softlocks. No, not even if “the player would have to be trying to softlock themselves to end up there”. No, not even before CP1.        40

No hot garbage.        41

Excessive use of sound effects can be grating.        42

Never...        43

Not every troll needs to be lethal.        44

Less is often more.        45

Try to make trolls as easy as possible to avoid once a player knows they're there.        46

Puzzley or harder to figure out sections should either be placed right after a spawn point or made nonlethal.        47

Long waits and moves that are annoying to do over and over or difficult to execute should be right before a progress save.        48

Fix ALL cheese, even if it's a harder solution than intended.        49

It's okay if people miss a troll the first time and then fall for it the second.        50

Don't use obscure tech the player has to put together or perform as the solution to a troll. DO use obscure tech that can be set up to happen automatically as the trigger for a troll.        51

Don’t use a dev exit unless absolutely necessary. Even then, ensure that the dev exit is much harder to perform, takes longer, skips only what it needs to, and is complex enough that it can’t be found by accident.        52

5. Test That Trash        53

Have your levels playtested thoroughly. Keep revising until the level flows well. There's no shame in a 52nd revision.        54

Before uploading your first draft, hit "play" in coursebot and play through it to completion while intentionally dying to every troll at least once. Go slow and pretend to be exploring. Time yourself doing so.        55

6. Mess with the Meta        56

Don't feel compelled to include a back-to-CP1 troll unless it's a really clever new setup or a new twist on the concept.        57

If you do CP1 the player, make the middle section relatively short and straightforward.        58

Come up with some clever, original, innovative trolls instead of just reusing 100% old trolls.        59

Disguise your clever trolls to look like they're simpler, older trolls.        60

Mix in simple, older trolls along with your clever, original trolls.        61

7. Cautious Creating        62

Don’t get too attached to anything in your level.        63

Start out as a new creator by making shorter levels.        64

Don't expect Carl (or whatever popular player) to play your first level.        65

8. Tips for Top-Tier Trolling        66

Mix and match creative techniques.        67

Think of ways to reuse elements from trolls you’ve already placed in your level.        68

If you have a point where you need to give or take away an item or powerup to set up the next troll, come up with a funny way to do it.        69

Harness your constraints, don’t fight them.        70

Trolls that focus on a certain theme or item or style can be interesting, but be sure you’re experienced enough at creation before adding constraints.        71

Find new tech by starting with a particular item and experimenting.        72

Find new tech by playing levels or watching other people play.        73

Figure out a funny way to kill Mario. Then disguise it with additional trolls to mislead the player into thinking the setup is something else.        74

Subtly train the player as they go to put them in the mindset you want for later trolls.        75

Coming up with a grand concept for the level as a whole can sometimes be a good starting point.        76

Above all, stay creative. There’s no one “right” formula for making a troll level.        77

Appendix A - Streamers Interested in Playtesting Troll Levels        78

1. Mentality Matters

https://www.flickr.com/photos/learningissocial/9880482233/

1. Always remember that your goal is to make the player laugh and never to frustrate them.

Self-created using popular meme, credit unknown, emotes belonging to CarlSagan42, used with permission, and icons from Super Mario Maker, Copyright Nintendo

This is the SINGLE most important thing to remember about creating a troll level. Your level should be designed to make the player (and their audience, if they're playing in front of one,) laugh. You need to trick the player while still being fair and still letting the player easily figure out how to progress after having been tricked. Even the name "troll level" is ambiguous, as trolls are generally malicious, with intent to laugh AT you, rather than WITH you. Good troll levels, on the other hand, should laugh WITH the player, rather than at them.  A better name might be "deception levels", but the name isn't nearly as catchy or easy to say.

Despite the fact that the idea of troll levels stemmed in a lot of ways from the Carl vs. Poo series, the community has come a long way since then, and Carl told me himself that Carl vs. Poo generally did pretty much EVERYTHING we have since learned not to do (okay, he actually called them “complete hot garbage”, but hey, I’m trying to be nice here). They were quite clearly designed to frustrate rather than entertain, and frustration can be funny for a time or two, but the appeal wears off quickly. Those levels were simply the catalyst that made Carl "realize how troll levels COULD be good if only we stopped trying to torment each other and actually tried to make each other laugh instead" (again, a direct quote). Now, we're looking for genuine humor.

2. Put yourself in the player's shoes. Look at your level as if you don't already know the answer.

Self-created using elements from Super Mario Maker, Copyright Nintendo

A common problem is that the creator of a level is inherently cursed with the knowledge of how the level is supposed to go. This leads to it being difficult for people to consider what alternatives a player may try, or what mistaken assumptions a player may make. Putting yourself in the player's mindset can be difficult to do, but it's an important skill that can be learned over time. A few techniques for how to do this will be elaborated upon below. Even using these techniques, be aware that it is usually impossible to fully predict how a player will view a level, which is why playtesting is also very important.

3. Playtest and give feedback on a lot of other people’s levels.

Self-created using Mario images, Copyright Nintendo

Doing this will let you get to know the community, give you a sense of what it’s like to see levels from the player’s perspective, and get a feel for the current meta. You’ll learn a lot about what works and what doesn’t by experiencing it and giving feedback from the other end, so definitely give it a try if you haven’t already. Engaging with the community also means that people will get to know you and be more likely to give you playtesting, feedback, and support.

4. Be aware of the "frustration meter".

Self-created

It's hard to find the right balance of difficulty in a troll level. Make the level too easy, and it will be boring. Make it too hard, and it will become overly frustrating. I often think of players as having a frustration meter. There's an optimal middle area on the frustration meter that you want to maintain.

A checkpoint will reset the frustration meter (so long as it's not a back-to-CP1 troll, but more on that later). The longer and more difficult a section is, the higher the frustration meter will climb. Excess waiting or monotony will raise the frustration meter quickly. Make sure it doesn't climb too high. The "annoying antisoftlock troll" for example, has a tough balance between "funny" and "frustrating". If, by the time a player is done laughing about their fate, they're still not close to beating it, it's probably a little TOO annoying. If they are likely to fall into it multiple times, it’s DEFINITELY too annoying.

5. Err on the side of too easy.

Self-created using popular meme, credit unknown, incorporating screenshot from Super Mario Maker, Copyright Nintendo

Most of the troll levels people create tend to have the frustration meter pushing near the upper end of the scale, and it's much easier to end up going over than going under. To keep the meter at the optimal level, the player should always feel like they're making progress through the level more often than they're being foiled. It's therefore a generally good idea to err on the side of making things too easy, as it will likely balance out in the end.

6. Be prepared to sink a LOT of time into each level.

https://pixabay.com/photos/pocket-watch-time-of-sand-time-3156771/

Troll creation is some of the most time-consuming to get right. Don't expect to be able to just spend a day or two on a troll level and be done. Most good troll levels take at least a week to create, and then go through several iterations of testing, tweaking, further testing, further tweaking, etc. Unlike other kinds of levels where a setup which can be performed consistently means a section is complete, troll levels often require adjustment for optimum levels of deception, humor, and entertainment. To really polish a troll level well, the entire thing needs to flow well from start to finish, and it takes a lot of time and experimentation to get just the right balance. It's not unusual for me that a level takes 3 or 4 weeks after beginning work on it before I'm satisfied that it's ready. If you intend to make something good, you need to be aware of what it takes to do so and be ready for the time commitment that comes along with it.

7. Be aware of the anatomy of a troll setup.

https://pixabay.com/illustrations/human-skeleton-human-body-anatomy-163715/

The best troll setups are those which violate the player’s expectations. In order to do that, a setup needs to make the player think they know what they need to do, have them fail in a humorous way when they do it, and then let them know what it was that they were supposed to do instead. If any of these components is missing, a troll setup will fall flat. If a player doesn’t think they know what to do, they will not have an expectation to violate. If they don’t fail or there’s no humor, then they’ll wonder what the troll was. And if they don’t have an idea of what to do right the next time, they’ll just end up confused rather than amused. Make sure that each setup has all three components.

2. Clear Communication is Key

Modified from https://pixabay.com/photos/vintage-telephone-phone-old-call-1750817/

1. It’s the creator’s job to communicate to the player how to complete the level.

Modified from https://pixabay.com/illustrations/megaphone-speaker-speakers-speak-3272935/ using Mario image, Copyright Nintendo

It may seem counterintuitive and contradictory in a troll level, but it’s actually even more important to communicate to the player how to get through the level than it is in a non-troll level. Confusion over where to go amounts to frustration. Yes, it IS possible to simultaneously deceive the player about what to do and make it clear to the player how to progress through the level. The key is in letting the player know where to go, but not what will happen to them when they go there, at least until after it’s happened to them once.

2. Make it clear what a player’s options are.

https://www.pexels.com/photo/grey-metal-hammer-34520/

Since trolls are all about violating a player’s expectations, you need to first make it clear to the player what options they have available (or at least, what options you want them to THINK they have available) before you can amuse them by making them choose the wrong option. The player should never have a point at which they feel as though they have tried all the available options and have no further ideas because it seemed as though all the options they had were wrong.

3. Arrows, sound effects, coin trails, track boxes, and other indicators are often not enough to clearly communicate.

Self-created using elements from Super Mario Maker, Copyright Nintendo

In most levels, these things would be communication enough, but in a troll level, players always have reason to doubt whether an indicator is telling the truth or not. Indicators can generally be used to point the way at a POSSIBLE path that the player may not otherwise see, but should not be relied upon to DEFINITIVELY communicate the correct solution unless the solution also definitively shows itself to be correct as well (e.g. an arrow pointing into what looks like a trap reveals the actual way out if you go the wrong way).

4. Don't make obscure or convoluted solutions, or solutions that can’t be easily found without using the editor or receiving hints from you as the creator.

https://pixabay.com/photos/map-tourism-lost-direction-guide-4152197/

Often, things will look quite different when the view is locked to Mario's position rather than freely scrollable in the editor. There are also things such as overlaid elements that can't be easily detected in-game but can be found more easily in the editor. You want the level to speak for itself, to be playable and enjoyable with no extra information. This is especially pronounced in SMM2, where, at least at the time of this present writing, the ability to open downloaded levels in the editor has been removed from the game.

5. Try to make it clear immediately what a player did wrong after a troll.

Self-created using elements from Super Mario Maker, Copyright Nintendo

In general, a troll should make it clear what the player did wrong immediately after they fall to it. Depending on the troll, exactly how this works can vary. If the troll is a sudden enemy in your face, then the obvious solution is to jump to avoid it. If the troll is "this looked like a carlbox, but was really actually the right way", then make sure that the fake right way actually shows the real right way out of the fake wrong carlbox (boy, that was a mouthful). For example, I once tested a level that had a spring contraption that knocked you into a carlbox that was just a small square room with a saw at the top of it. Avoiding the spring led one down a path to a dead end. The saw in the “carlbox” actually had vines behind it and the real strat was to damage boost through it and climb the vines. However, the dead-end path never showed it, and left me wondering what else I could try. If, instead, the dead end path had led back around to where the top of the saw were visible with the vines coming out of it, it would have made it immediately clear what the solution was, and would have been a properly communicated troll. As it was, it made it nearly impossible for a player to figure out on their own.

6. Don’t blame the player. Ever.

https://pixabay.com/illustrations/mario-mushroom-video-game-turn-on-1081471/

It’s easy to get frustrated with a player for not seeing the “obvious” solution, or for not enjoying a section, or for thinking a section is jank. Any such things are on you as the creator to fix. If the player is having trouble, it means you did something wrong, not the other way around. As has been mentioned several times, things look different from the creator’s perspective, and what may be obvious, enjoyable, or consistent to you may be very different from the player’s perspective. You should ALWAYS trust the player’s perspective over your own when creating a level. Instead of getting defensive or frustrated at a player when their perspective is different, take the opportunity to see where it is they’re struggling and think about what you can do to communicate more clearly or make the section more entertaining or more consistent.

7. Avoid large amounts of invisible stuff that can’t all be checked in a single pass.

Self-created using elements from Super Mario Maker, Copyright Nintendo

This is really more of an application of the previous couple of sections, but this specific case is easy to miss without specifically calling attention to it. Things like long rows of hidden blocks, spikes behind the goalpost, or blocks behind saws can make for some very amusing trolls, however, collections of hidden elements are something that looks very different from the creator’s perspective than the player’s. To the creator, a collection of hidden items is just that, a collection. To the player, however, if testing one area of the collection is immediately lethal, all they know is that that one spot was not safe. For example, if a player jumps into a spike wall at the goal at around the 2 block height, that doesn’t let them know whether the spikes exist at the base, or at the height of a max jump, and they’ll need several more attempts to confirm whether there’s a way through or not. If a long row of hidden blocks contains boo rings, a player will only realistically be able to hit two or three before dying, leaving them unaware if the entire area is covered or if there’s a gap. In general, try to shrink the invisible collection (e.g. by narrowing the path through the goal or the area blocked off by hidden blocks to a gap of two or three blocks) or make it non-lethal to check all of it (e.g. by putting solid ground rather than deadly spikes behind the goal or by putting boos rather than boo rings in hidden blocks, etc.).

8. Every element you place in the level should have a reason for existing.

Self-created

Everything you put in the level should be in some way related to setting up, positioning the player properly for, or disguising, a troll. Extraneous elements will make the player wonder if the element is relevant in some way, and confuse them into focusing on that and not seeing the actual solution as a result.

Additionally, items, powerups, and keys should generally make it clear soon after the player gets them whether they were necessary or not, either by forcing them to be used immediately, forcing them to be discarded immediately, or triggering a troll that prevents progression immediately when you try to get them. Leaving a player with access to an item for too long will make them wonder where and when they should have used it. Don’t put unrelated trolls in a section between the point where the player gets an item and the point where they need to use it.

9. Avoid delicate or janky solutions that can be performed and yet the player still fails for not doing it good enough.

Modified from https://unsplash.com/photos/geNNFqfvw48

There’s nothing worse than a player trying what is essentially the right solution and then still dying because they were a pixel too far to the left, or because the contraption janks out if you catch it on a particular frame. When the player tries an idea and it doesn’t work, they’re going to assume it was the wrong idea and not try it again. They will then be completely lost and out of ideas of anything else to try. Don’t let this happen.

10. Players generally have tunnel vision and don’t look outside of the immediate area to which they have access. If something important is out there, make sure to draw their attention to it fully.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

Don’t assume that just because something was visible onscreen that a player will see it. If the level is being designed for live streams, you can rely somewhat more on chat’s help, but even then, it’s a good idea to make the important area of the screen draw attention. I personally like to use the "heavenly light" sound effect (usually triggered automatically by a bomb at the same point as the area becomes visible/important), but you might have a different approach, so long as it fully captures the attention of the player, that’s what’s important.

11. Hints rarely work, try to avoid them as much as possible. If a hint must be used, make sure it can't be mistaken for anything else.

Created by Drakyen using popular meme and elements from Super Mario Maker, copyright Nintendo

I learned this the hard way, by creating levels with hints that ended up being hard-to-interpret and left players confused. It's very hard to communicate with visuals alone. Simple hints like an arrow pointing from a question mark block to a mushroom to indicate that the shiny the player didn't trust was actually what they wanted will generally work. Complicated hints that involve multiple steps and pieces that are supposed to look like an earlier section of the level generally do not work unless that section of the level was EXTREMELY distinctive. New creators often feel the need to make hints more and more complicated in an attempt to “make sure the player understands”. The problem with this is that the more complex the hint, the more elements there are that can be misinterpreted or confusing. The player being run through the hint will often not even realize that it IS a hint, and think it’s simply a thing they need to do for the level. The need for such complex hints may also be an indication that the section in question may simply be too hard to figure out in general. If so, the section may do better in a puzzle or uno mas level than it does in a troll level.

Another big problem with hints is that players may not see them the first time. If the hint is only visible in a path which is obviously the wrong way to go, a player who doesn’t see the hint the first time will likely never see it, as they will likely never try that same path again, knowing it was wrong and not knowing there was a hint there to be seen. If you’re going to use a hint, make sure that it is impossible for a player to miss the fact that there is a hint there, and remove anything extraneous that the player might interpret as part of the hint which isn’t.

Note also that the difference between a hint and showing the player the right way as in section 3E is that generally a hint makes reference to another section of the level or to some kind of tech a player would need to perform, whereas showing them the right way keeps them in the current area of the level and simply shows them something about the current area that they couldn’t previously see.

12. Try as much as possible to keep setups for contraptions offscreen and inaccessible.

Modified from https://en.wikipedia.org/wiki/File:Wii_U_controller_illustration.svg with elements from Mario Maker, Copyright Nintendo

Contraptions that are onscreen may give away a troll, or worse, confuse a player about their purpose. Likewise, setups for contraptions where you can get inside the inner workings may confuse a player into thinking that’s the way to go. I played a level that had a contraption with a moving launcher that came up out of a pit from below the screen, but that pit was accessible and there were all sorts of items and things that looked like possible ways to go down there. Use elements like saws, bumpers, semisolids, and one-ways to ensure that you keep your mechanisms and the playable parts of the level separate.

13. It's okay if people don't fall for every troll, but make sure that players know the purpose of all visible elements which were part of a troll they missed.

https://www.flickr.com/photos/unloveable/2401166750

Some people will avoid some trolls some of the time. This is okay. You don't need to overoptimize troll setups to ensure everyone will always fall into them. However, visible troll setups that can be easily dodged without foreknowledge and aren’t obviously troll setups can sometimes confuse a player, making them think that since they didn’t see a troll there, that setup HAD to be for something. They may then proceed to fixate on that setup, continuing to dodge the actual troll but then get stuck on the following troll unaware that the setup they’re looking at has nothing to do with the solution. In general, if a troll setup has visible elements, then it should either be set up in such a way that it’s nearly impossible for a player to dodge without foreknowledge, or that the elements visibly perform the trolling contraption even in the case where the player dodged the actual troll.

3. Linear Levels are Less Loathsome

Self-created

1. Don't have too many paths or options open to a player at once.

https://pixabay.com/photos/road-signs-war-confuse-the-enemy-2292144/

Don't create branching paths. Too many paths and options will overwhelm and confuse the player, and make them wonder which choice is correct, or whether any is correct or there's something else entirely that they're missing. It will lead the player to wonder whether the solution might be 5 rooms back and whether everything they did was a wild goose chase. In general, try to keep it clear what the player's choices are and try to keep those choices limited to two or three at any given time. I'm not saying that branching paths can't ever work, but in order to work well, there can be absolutely no ambiguity about whether a path is wrong or not anywhere in the level, and that's EXTREMELY hard to accomplish in such a way that it works for all players at all times for all sections.

Bear in mind that a level can potentially have the player go through the same area multiple times and still be linear, so long as the changes made to the area are such that the player can never be confused as to where to go next, nor even think “oh, maybe I can get through there”. This means closing off all previous and future entrances and exits to an area and leaving only the current ones open. While this is possible, be aware that it is difficult to dejank and decheese, and may not be something you want to try until you are more experienced at level creation.

2. Keep the open options at any given time part of a cohesive whole.

Modified from https://www.pexels.com/photo/black-and-white-blank-challenge-connect-262488/ with Carlbox image belonging to CarlSagan42

Your level will feel like a mess if you give the player options that aren’t cohesive. Give the player a set of two or three options for how to progress past one challenge, let them figure it out, then send them on to the next. Try not to leave multiple, unrelated trolls open at the same time. Once a player has progressed past a challenge, it should be generally clear that they have made the right decision. Any of the other options that had been available should either be obviously wrong once the player has progressed, or should be set up simply enough that they would immediately show themselves to be wrong should a player get curious and try them.

3. Make sure the decision a player makes feels like a choice rather than chance.

https://pixabay.com/vectors/confusion-left-right-straight-311388/

The player should never feel like all options available to them are equally likely to troll them. For example, giving a player choice between two paths through one-ways, either of which may contain a troll, will generally feel like pick-a-path. If, however, the player is scared or enticed into one path over the other, they will feel like they had reason to make the decision they made and will be satisfied that the troll effectively predicted their decision and violated their expectations rather than just having picked one and hoped for the best.

This happens a lot with game-state related things, such as whether to lose a powerup, or wait for a p switch to wear off, or which switch block state to put the level in before going through a door or pipe. The player, having no way to predict what will happen on the other side, will feel cheated if they go through and had no reason to think one or the other. These kinds of choices, like any, can be set up in a way in which they are part of a larger troll that makes the player think they know what to do, but just a straight “haha you didn’t lose your mushroom get trolled” is generally frowned upon.

4. Fence the player into the path you want them to follow.

https://pixabay.com/photos/fence-wire-mesh-sunset-background-72864/

Leaving the level too open presents the same problem as having branching paths does, and makes the player wonder if there's some tricky solution somewhere else in the level. Use walls, floors, and ceilings to ensure that the player knows exactly where they can and cannot go. Use one-ways, bumpers, semisolids, and saws to make sure your setups work without letting players escape into the “behind the scenes” areas. Let the trolling be in your setups, rather than in not knowing where to go next.

5. Limit the amount of backtracking a player has to do.

https://pixabay.com/vectors/arrow-one-way-right-sign-road-759223/

Solutions to trolls should be relatively accessible. The wrong path should either make itself obvious that it's the wrong path immediately, or else provide a clear and impossible to miss indication of where the real path was. If the actual solution is too far back, a player will have to search the entire level for it and lose their sense of progression. Additionally, allowing a player to backtrack too far may open up cheese possibilities, for example, by giving them access to items that were meant for an earlier section and allowing them to be taken to a later one.

4. Deceptive yet Docile Design

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

1. Fight the instinct to make traps that are easy for a player who already knows about the trap to fall into accidentally.

https://pixabay.com/vectors/slippery-wet-floor-danger-warning-98821/

When creating levels, a creator will often have a choice of arranging things in a way which makes a troll, trap, or death easier or harder to end up in accidentally. Your first instinct will be "well, this is a troll level. Falling into this accidentally will be trolly." However, falling into the same trap repeatedly while trying to make progress will raise the frustration meter, generally beyond the point where the level is funny and into the realm where it becomes frustrating. Fight the instinct, and make it hard for the player to screw up by accident.

2. No softlocks. No, not even if “the player would have to be trying to softlock themselves to end up there”. No, not even before CP1.

Self-created using elements from Super Mario Maker, Copyright Nintendo

This should go without saying, but I'll say it anyway. Please don't softlock the player. Never underestimate a player’s ability to blindly stumble into a complicated series of events that leads to a softlock. I’ve seen it happen too many times to believe there’s such a thing as a softlock that a player can’t possibly do by accident.

As for before CP1, where a softlock does not actually present a choice between losing all progress or timing out, most players still consider it bad etiquette. You don't want to force the player to get into the habit of hitting start over, and for a lot of people, hitting start over breaks up the flow of gameplay. Though many players WILL use the start over button before the CP, others consider it courteous to those who don't want to get into that habit to have the option not to. Frankly, I blame Nintendo for this, for making the start over button destroy CP progress, but even were that not the case, it’s generally a good idea for communication’s sake to let a player know they’ve reached the end of the line by always including some form of death in any potential softlock.

3. No hot garbage.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

Pick-a-path, enemy spam, random kaizo blocks, offscreen falling enemies, and find-the-hidden-block, insta-death-door garbo are generally more frustrating than funny, and are a lazy way of increasing the frustration meter. You’re trying to be clever, not lazy. The occasional kaizo block or easy-to-find hidden block or well-placed falling enemy snipe can be good, but too much of it will make your level feel chaotic and frustrating. Leave the super expert garbo to super expert.

4. Excessive use of sound effects can be grating.

Modified from:

https://pixabay.com/vectors/bang-explosion-noise-onomatopoeia-148261/

https://pixabay.com/vectors/suffering-comics-hopeless-noise-154637/

https://pixabay.com/vectors/pain-suffering-comics-hopeless-154638/

https://pixabay.com/illustrations/comic-noise-paleness-explosion-2011646/

https://pixabay.com/vectors/broken-noise-cracked-comic-cartoon-147959/

https://pixabay.com/vectors/music-noise-onomatopoeia-sound-148238/

https://pixabay.com/vectors/wouf-comic-noise-onomatopoeia-147858/

https://pixabay.com/vectors/red-green-blue-yellow-fight-41991/

https://pixabay.com/vectors/explosion-detonation-blast-burst-147909/

https://pixabay.com/illustrations/pow-comic-comic-book-fight-1601674/

https://pixabay.com/illustrations/zap-comic-comic-book-fight-1601678/

https://pixabay.com/illustrations/kapow-comic-comic-book-fight-1601675/

https://pixabay.com/vectors/pop-loud-sound-sounds-audio-word-41988/

https://pixabay.com/vectors/explosion-noise-fulmination-loud-147931/

https://pixabay.com/illustrations/lol-laugh-loud-funny-happy-706869/

https://pixabay.com/vectors/rooster-crow-morning-proud-loud-310266/

https://unsplash.com/photos/fgLoM6NeMvc

See how intrusive that was?

Sound effects are like the spice of a good level. A well-placed one can enhance the humorous flavor of a troll, but overspicing a level with constant sound effects will make it unpalatable.

5. Never...

Copyright Rick Astley or something

...gonna give you up, never gonna let you down, never gonna run around and desert you. Never gonna make you cry, never gonna say goodbye, never gonna tell a lie and hurt you. We're no strangers to love, you know the rules and so do I. A full commitment's what I'm thinking of. You wouldn't get this from any other guy. I just wanna tell you how I'm feeling, gotta make you understand. Never gonna give you up, never gonna let you down, never gonna run around and desert you. Never gonna make you cry, never gonna say goodbye, never gonna tell a lie and hurt you. We've known each other for so long. Your heart's been aching but you're too shy to say it. Inside we both know what's been going on. We know the game and we're gonna play it. And if you ask me how I'm feeling, don't tell me you're too blind to see. Never gonna give you up, never gonna let you down, never gonna run around and desert you. Never gonna make you cry, never gonna say goodbye, never gonna tell a lie and hurt you. To whoever actually read through all of this and is seeing this message, hit me up on discord and tell me you found this message hidden inside the wall of rickroll. Curious to see how many people actually find this. Never gonna give you up, never gonna let you down, never gonna run around and desert you. Never gonna make you cry, never gonna say goodbye, never gonna tell a lie and hurt you. Never gonna give, never gonna give. Never gonna give, never gonna give. We've known each other for so long. Your heart's been aching but you're too shy to say it. Inside we both know what's been going on. We know the game and we're gonna play it. I just wanna tell you how I'm feeling, gotta make you understand. Never gonna give you up, never gonna let you down, never gonna run around and desert you. Never gonna make you cry, never gonna say goodbye, never gonna tell a lie and hurt you. Never gonna give you up, never gonna let you down, never gonna run around and desert you. Never gonna make you cry, never gonna say goodbye, never gonna tell a lie and hurt you.

6. Not every troll needs to be lethal.

Self-created using elements from Super Mario Maker, Copyright Nintendo

Sometimes it’s funnier for a player to anticipate death and not receive it than to anticipate safety and receive death. Sometimes, a troll is just as funny even if it doesn’t actually kill you. Change it up. Don’t force the player back to the most recent checkpoint for every troll you pull on them.

7. Less is often more.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

The temptation to pack a level full of trolls is great. However, you don’t need to have too many trolls to make a good troll level. I have found that generally, you only need between 6 and 10 trolls per CP section to make a good troll level. Too many more than that tends to make the section too long and raise the frustration meter too high before resetting it with the CP. The bonus of doing this is that using fewer trolls per level means you have more ideas for additional levels.

8. Try to make trolls as easy as possible to avoid once a player knows they're there.

https://unsplash.com/photos/Pvclb-iHHYY

Avoid difficult platforming or other difficult tricks, as spaghetti tends to raise the frustration meter very quickly in a troll level. A player dying over and over on a section they've already seen will not feel as though they are making steady progress through the level.

9. Puzzley or harder to figure out sections should either be placed right after a spawn point or made nonlethal.

https://pixabay.com/vectors/rubik-s-cube-cube-puzzle-colors-157058/

No one wants to play through a whole section just to get a glimpse at a screen and have absolutely no idea what to do the next time, go back there, get another glimpse, and still have no idea, etc. If you have a segment that's hard to figure out, either give the player the ability to retry, or put it close to the beginning of a section (at the start of the level or right after a CP) so that they can at least get back there quickly.

10. Long waits and moves that are annoying to do over and over or difficult to execute should be right before a progress save.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

Having to repeatedly wait for things such as P switch timers, long track contraptions, contraptions involving several bomb timers, ! block resets, or anything that involves the player having to stand and do nothing for more than about half a second is annoying. Dying to the same trick over and over is annoying. And that annoyance multiplies the more times you have to do it. Remember that the farther before a CP or the goal something is, the more times it will need to be done successfully. The thing right before a CP will need to be done only once (or perhaps twice twice ), the thing before that might need to be done twice or three times, the thing before that three or four times, etc. Generally, the more difficult and/or annoying something is, the closer to the end of a section and the ability to save your progress it should be. The exception being that a long wait or annoying to perform trick can have a SINGLE, extra-clever troll after it for humorous effect. Forcing someone to do something extra annoying more than twice before saving their progress, however, tends to become frustrating extremely fast. Also, be aware that this guideline should be used to balance the level’s pacing for a trick that is harder than others, and not as a free pass to give someone a ridiculously impossible trick before gracing them with the end of the section.

11. Fix ALL cheese, even if it's a harder solution than intended.

https://pixabay.com/photos/cheeses-cheese-gouda-truffle-gouda-389687/

This is one of the most important pieces of advice I have, and one of the least intuitive. You may think to yourself "why should I fix it if it's harder? It won't break the level or let the player avoid any trolls." The important thing to realize, however, is that the moment the player finds a solution that works, they will stop looking for an alternate solution, no matter how janky what they found is. They will therefore become frustrated VERY quickly with any solution that doesn't flow well. This is one of the things that took me the longest to learn, and one of the most crucial.

A specific example of this to pay special attention to is jank that may provide the player a way through sometimes but not others in a seemingly random fashion (be it based on RNG, subpixel movement, subtle timing differences, or anything else not within easy control for the player).

12. It's okay if people miss a troll the first time and then fall for it the second.

Emote belonging to CarlSagan42, used with permission

Sometimes, it can be even funnier this way. People may not see that they have dodged a troll the first time and then walk face-first into it the second time.

13. Don't use obscure tech the player has to put together or perform as the solution to a troll. DO use obscure tech that can be set up to happen automatically as the trigger for a troll.

Self-created parody of an Ikea instruction pamphlet, Copyright Ikea

So you just found out that dropping a pow on a piranha next to a wiggler booty makes a thwomp do cartwheels. That's great. It's also something a player would never think will make any difference, so they're not going to try it. If you can come up with a setup that automatically drops a pow on the piranha and makes the thwomp cartwheel into the player, that's a really top-tier trolling setup, because the player is not going to expect a cartwheeling thwomp. If, on the other hand, your setup requires the player to drop the pow there themselves to cartwheel-thwomp their way to victory, then the setup is better put in an uno-mas level and not used for trolling. What’s considered “obscure tech” is, of course, somewhat subjective, and depends a lot on your target audience, so use your best judgement.

14. Don’t use a dev exit unless absolutely necessary. Even then, ensure that the dev exit is much harder to perform, takes longer, skips only what it needs to, and is complex enough that it can’t be found by accident.

https://unsplash.com/photos/HK8IoD-5zpg

No one wants to play a level that the creator didn’t upload fairly. If you can’t beat it legitimately, chances are that your level is too hard. The exception to this is that certain checkpoint shenanigans may require a dev exit to allow one of the 3 clear checks (for example, if clearing from a CP depends on the existence of blocks that get destroyed while playing an earlier section, or some kind of red coin level trickery). In such cases, make the dev exit as fair as possible. If you can, ensure that the dev exit can only be used from the necessary clear check. Use the dev exit to skip only the part of the level that’s problematic, but make the dev exit still require completing everything else in the level. Ensure that all elements involved in the dev exit either have some other purpose or cannot be discovered by accident (e.g. a hidden block that needs a spring drop to get to in a place where no one would ever try a spring drop naturally).

5. Test That Trash

Image elements belong to CarlSagan42

1. Have your levels playtested thoroughly. Keep revising until the level flows well. There's no shame in a 52nd revision.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

You will make mistakes. You will make a lot of mistakes. You will make a lot of mistakes with every level you make. No matter how experienced a creator you are, you will never be perfect. The only way to find the mistakes that you will inevitably make is to test your level thoroughly. Find a group of players you trust to give accurate feedback, and accept constructive criticism graciously. It can hurt to hear sometimes, but being able to accept it, learn from it, and apply it beyond just the specific situation is one of the most important skills to develop.

Finding viewer level streams on Twitch so you can actually see your level being played live can also help. I can’t tell you how many times I’ve had a nonstreaming playtester find cheese and not even realize it was cheese until after I had updated the level and they asked why I had changed it, thus making their playtesting somewhat useless. Seeing it played live will also help give you a better sense of what players might be thinking at every point in the level. If ever you see something wrong with the level during such playtesting, fix it and reupload.

Of course, please ask the streamer if they enjoy this style of troll level and don’t just submit it to a stream, as troll levels are a contentious genre and there are many streamers who won't enjoy them. You're going to want to find streamers who will have fun with it in the process of helping you test and not get annoyed with you for submitting it. Our dear friend B0X Gaming maintains a list of streamers who are willing to test troll levels, but searching Twitch’s Super Mario Maker 2 tab for others not listed here may also get you to some streamers willing to help. You may even make a friend or two in the process.

2. Before uploading your first draft, hit "play" in coursebot and play through it to completion while intentionally dying to every troll at least once. Go slow and pretend to be exploring. Time yourself doing so.

https://pixabay.com/photos/time-stopwatch-clock-hour-minute-731110/

This sounds crazy and unnecessary at first, but bear with me. The idea here is to generally get a gauge of what the level will be like from the player's perspective. By dying to every troll at least once, you'll see the level as the player will see it, you will more or less be progressing through the level the way they will, and if you find yourself being frustrated at any point, you'll immediately know what parts to change. You'll see the limited frame of reference that the level gives the player rather than the overall view you get when scrolling freely in the editor. Timing how long this takes you will also give you something of an estimate of how long it will take a real player to complete (though always estimate that a real player unfamiliar with the level will take somewhat longer than your time). If it takes more than an hour, your level is probably too long. I do this with every level I create, and believe me, it helps.

6. Mess with the Meta

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

1. Don't feel compelled to include a back-to-CP1 troll unless it's a really clever new setup or a new twist on the concept.

Self-created using elements from Super Mario Maker, Copyright Nintendo

A lot of people seem to think that a CP1 is something that should always exist in a level. Granted, they can be funny, but just like any other tired old troll, they shouldn't be done over and over without a new twist. Additionally, if you’re going to CP1 the player, don’t do it more than once or maybe twice if you have extremely clever setups for it, as the more times you do it, the more tedious and frustrating it will be. Especially if one setup is less creative and another setup is more creative, the less creative one will set the tone and make even the creative one feel frustrating.

2. If you do CP1 the player, make the middle section relatively short and straightforward.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

CP1 can be funny if done well, but if the middle section is too long, it will feel like a chore to get back to CP2.

3. Come up with some clever, original, innovative trolls instead of just reusing 100% old trolls.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

Reusing the same trolls over and over becomes boring. You may get someone with it, but it will not be particularly amusing. Be creative. Do something new!

4. Disguise your clever trolls to look like they're simpler, older trolls.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

Making your trolls LOOK like classic trolls and then switching it up on the player, however, is very entertaining.

5. Mix in simple, older trolls along with your clever, original trolls.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

It's okay to include classic trolls alongside your new ones. It makes things interesting and keeps the player on their toes. Mix the old meta with the new meta. Put in the quadruple-reverse troll. Send the player on an emotional rollercoaster. Disguise the setups for your new trolls as setups for classic trolls. Make the player think that the sideways spring that's really part of a brilliant new contraption was only there to bounce that mole into their face. Wreck the player unexpectedly with your genius.

7. Cautious Creating

Self-created using popular meme and Builder Mario hat, Copyright Nintendo

1. Don’t get too attached to anything in your level.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

As you test your level, you may discover that certain setups simply aren’t working. Don’t be afraid to scrap whole sections (or save them for a level where the section’s placement can be more optimal). Sometimes things don’t work out, and that’s okay. Not being able to improve a setup any further is not an excuse for keeping something confusing, janky, misleading, overly difficult, annoying, or that otherwise doesn’t work in your level. If it can’t be improved to a point where it works, then means that it just doesn’t work. Don’t cling to a sinking ship.

2. Start out as a new creator by making shorter levels.

Mini-Me, Copyright New Line Cinema, Warner Bros.

Level creation is a skill that needs to be learned like any other. The last thing you want to do is to pour hours into a level only to find out you've done everything wrong. Make short proof-of-concept levels and get feedback on those. Refine, rearrange, and reevaluate what works until you've developed your own style and you’re comfortable making a full-length level. It'll be much more pleasant for both you and anyone testing your levels if the levels you create as part of your learning process are shorter.

3. Don't expect Carl (or whatever popular player) to play your first level.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

Carl played my first levels. Looking back, I wish he hadn’t. While there are a couple of sections in each that still hold up, I can say that I'm mostly embarrassed to look back on my first couple of levels now. It's easy to get excited about your creation, but you shouldn’t expect greatness from yourself your first time out. Guaranteed you will make better and better levels as you practice more. Experiment. Get feedback on what works and what doesn't. Get comfortable as a creator. Then, when you're ready, create the masterpiece you were always meant to make and send it to Carl.

In the same vein, don't name your first levels for Carl. The fact that there are so many garbage levels named for him that he never actually plays gives him a reputation in some circles for "ruining mario maker with awful troll levels". We all know that's not true, that he encourages the creation of good troll levels, not awful ones, but the fact that so many new creators who don't yet know what works and what doesn't are naming their levels for him makes those who are less familiar with him think that those are the levels he's encouraging. Please don't add fuel to that fire.

8. Tips for Top-Tier Trolling

Modified from screenshot from Super Mario Maker 2, Copyright Nintendo

1. Mix and match creative techniques.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

Okay, I’m going to break the fourth wall for a minute here. Up until now, everything listed have more or less been things the troll creation community has discovered and generally agreed upon to be objective best practice the majority of the time. This section, however, is more about presenting some different approaches and techniques you can use to get the creative juices flowing and up your troll game, and the tips that follow are subjective things that may or may not work for any particular creator. Try them out and see what works for you personally.

2. Think of ways to reuse elements from trolls you’ve already placed in your level.

https://pixabay.com/vectors/black-and-white-environment-recycle-1297353/

Often, having the same element troll you in multiple ways can add to the humor. You run into a troll, see how to dodge it, and the next thing you know it comes back around and gets you from another angle. Even elements that are reused for a completely different section can help foster creativity and will be impressive when people realize that it was the same setup used in multiple places.

3. If you have a point where you need to give or take away an item or powerup to set up the next troll, come up with a funny way to do it.

https://unsplash.com/photos/hQ4BQ3wdHsQ

So the player was big in the last section and you need them to be small again. Sure, you can make them jump through a saw to lose their powerup, and that’s perfectly fine. However, in many cases this may be an opportunity for additional humor. If you can come up with a clever new troll that does something funny to them if they’re big and will let them through if small, you’ve just added value to your level. Likewise if, for example, you need to give the player a POW block. If you come up with a funny way that they could potentially miss the POW block, such as blowing it up automatically if they wait too long, it can also add humor.

4. Harness your constraints, don’t fight them.

https://pixabay.com/photos/cowboy-saddle-saddle-arizona-2345137/

For example, spawn mechanics are a very big constraint. Sprites in Super Mario Maker 1 only spawn when the screen is scrolled to where they’re about 2 and a half blocks offscreen. It took me a while to get used to that mechanic. Many setups I had in mind before I got the game didn’t pan out because of spawn mechanics. Launchers I was depending on to stop the movement of enemies weren’t spawned yet, and the enemies would just walk offscreen and despawn themselves. Enemies that I expected to come in at a certain time didn’t because they weren’t loaded. Things worked differently when I loaded a section in editor than when I scrolled in the section normally during gameplay. The whole thing started out feeling like jank. Eventually, however, I learned how spawn mechanics work and how to use them to my advantage. For example, spawn mechanics can be very useful for timing setups to trigger when the player has scrolled the screen by a particular amount, thus turning that constraint into an advantage.

5. Trolls that focus on a certain theme or item or style can be interesting, but be sure you’re experienced enough at creation before adding constraints.

Self-created parody of Clippy, Copyright Microsoft, using popular meme and Builder Mario hat, Copyright Nintendo

Water levels, darkness levels, autoscrolling levels, speedruns, and kaizo are all examples of styles of troll level that CAN be done well, but due to the additional constraints inherent to each, can also be done extremely poorly. Make sure you know what you’re getting into before adding those kinds of constraints. Experience with making both “regular” troll levels and non-trolly levels of whichever type will help when eventually you want to combine them. Note also that while it’s possible to create a level which combines trolls with pretty much any other style of level, it generally does not work to combine more than one of those styles. For example, a kaizo-puzzle-troll level is almost certainly doomed to turn out frustrating.

6. Find new tech by starting with a particular item and experimenting.

Modified from https://pixabay.com/photos/lab-research-chemistry-test-217041/ with elements from Super Mario Maker, Copyright Nintendo

Try out different movement patterns, combine it with other elements, etc. and see what kinds of interesting interactions you can find. Try different angles. When you find something interesting, think about how it can be used for trolling.

7. Find new tech by playing levels or watching other people play.

https://unsplash.com/photos/bUgaIaZysH0

The majority of the tech that I use in my levels are things I’ve seen happen in other levels, usually by accident. Obviously, if you see a particular setup used for trolling, don’t just reuse that setup, as that’s basically just stealing someone else’s idea. However, very often a weird interaction that’s unrelated to the level being played will happen by accident, or I’ll see a particular tech used in a non-troll level and think about how that can be turned to trolling. Any time I’m playing SMM or watching someone else play on Twitch or YouTube, I’m always on the lookout for weird tech, and always thinking “how can I use this for trolling?”

8. Figure out a funny way to kill Mario. Then disguise it with additional trolls to mislead the player into thinking the setup is something else.

https://unsplash.com/photos/zjQZgslA0M0

The way I personally tend to build levels is that in each segment of the level, I take one interesting and impressive piece of tech, create a setup that uses it, and then add layers of trolls that either disguise the setup by using elements of it for simpler trolls, or which bait the player into walking face-first into the setup. Often, I’ll have several layers of trolls disguising the troll that disguises the troll that is the disguise for the main troll.

9. Subtly train the player as they go to put them in the mindset you want for later trolls.

Self-created from https://imgflip.com/memegenerator/Expanding-Brain

When trolling, a lot of whether a troll will succeed or not and go over well or not will be based on the player’s mindset. A lot of it can be luck. A player already in a bad mood playing a troll level will have their frustration meter overload far more quickly and will tend to not enjoy is as much as a player in a good mood. A player who is tired may miss things they wouldn’t if they were more alert. Much of this is not controllable. However, there are things you can do within the level to affect the player’s mindset for later on in the level. An example of this phenomenon is that players will be a lot more receptive and quick to correctly interpret hints in an escape room level than they are in a troll level, because they’re expecting the hints to be in a certain format in the former and are not expecting that in the latter. This concept is too complex and too subtle to fully explain or even fully understand without a PhD in a VERY niche area of psychology, and a lot of it will be necessary to simply get a feel for as a creator. A lot of it will also naturally differ from player to player, but there are a few general techniques you can use. For instance, you can train the player to do a certain action or expect a certain thing to happen by including a similar setup earlier in the level. This can be done in a way that makes the player conscious of it, by giving them a Twice Twice, three-identical-action, or otherwise identically-looking setup and making them question whether to do the same thing or a different thing, or it can be done more subtly in a way that the player is not conscious of it, by giving them a challenge which is similar to an earlier setup and recalls it to mind, but not directly enough that they realize they’ve been trained.

10. Coming up with a grand concept for the level as a whole can sometimes be a good starting point.

Modified from https://pixabay.com/photos/super-mario-game-nintendo-super-1508624/

Many troll levels have a main troll around which the entire level is based. For example, “7 Troly peices” (sic) was a red coin level where each red coin section was completely playable and fair, but placed the disguised flagpole between the level hub and the key door with its safety coin. The troll being, when you finally went for the safety coin, you’d suddenly beat the level and realize that everything you’d done was unnecessary. Another example is “Larl Does Everything Twice Twice”, a level which gave the player two sections which looked the same but which changed up the trolls each time, after which, if the player failed to survive the final troll before the goal, they would be forced to do both sections again in order to have another shot at the goal. Coming up with a clever new overarching idea of this kind can be a creative starting point for a level.

11. Above all, stay creative. There’s no one “right” formula for making a troll level.

https://pixabay.com/photos/small-fresh-super-mario-color-1309442/

I wish making a good troll level were as easy as a spaghetti recipe. It’d be great if I could just tell you exactly what elements to use and where to place them to make the world’s perfect troll level. But you can’t make a good troll level by copying what someone else has done. Ultimately, troll levels are about pushing the envelope, violating expectations, and making something new and unique. I can only tell you my observations on what works and what doesn’t, but figuring out exactly what to create with those observations in mind is all up to you. Recognize that all the points discussed in this document are guidelines for best-practice, not hard-and-fast rules. Try to work within them, but be aware that there may be times when consciously ignoring some of them may make your level more entertaining. Experiment and find a style that works for you. Just because something hasn’t been done doesn’t mean it can’t be done. Be bold, try new things, and think outside the (Carl)box. Have fun, and I can't wait to see what you come up with!

Appendix A - Streamers Interested in Playtesting Troll Levels

As I no longer have time to maintain this list, B0X Gaming has taken up the task of keeping an up-to-date list of streamers interested in playing troll levels, some of whom are willing to playtest and will accept them as viewer levels. Check out the link for an extensive list of streamers:

https://docs.google.com/spreadsheets/d/1h4VbfZeyjcdV259BomxhukWjcrFDDRV1uBHE0hpTEec

This list is continually being updated and expanded, so be sure to check back often for updates.

If you are a streamer who would like to be added to the list, let me or B0X know. If you know a streamer you think would like to be on the list, tell them to contact me or B0X. If you are bored of reading, good news, this is the very end. Thanks for reading.