Views: 771,866,887
33 users online: 7 Up, Brutapode89, chickaDEE Magazine, Darolac, debug001, ECS.98,  Eevee, GbreezeSunset, Green Jerry, Hinalyte,  Hobz,  idol, izaguirrefermin28,  KKevinM, LouisDoucet, lx5, Magmatic,  Manofer, Maxodex, MercuryPenny, Minish Yoshi, nesta_, Nicoke42,  Ninja Boy, Pink Gold Peach, placeholdertest,  Samuel,  Super Maks 64,  Tahixham, Turbo86, VLSkoot, zacmario, Zapkaboom - Guests: 57 - Bots: 227Users: 40,544 (1,850 active)
Latest: Barry Jack
Tip: $5.27Not logged in.
How can I improve my level design?
Forum Index - SMW Hacking - SMW Hacking Discussion - How can I improve my level design?
Pages: « 1 »
I released my hack in this C3 and the reception wasn't that great, it was heavily criticized for having simple, boring and uninspired levels as well as overly frustrating levels, and overall being a hack with a bigger focus on having good looking aesthetics and having a ton of worlds than in having good and fun level design.

Because of all that I have a few questions:
-What makes a level boring or what make it stand out?
-How I can make a good but simple level?
-How can I make a hard level that is fun rather than frustrating?
-What kind of levels are considered "great" by today's standards?

I really want some level design tips because I want to improve, I want the third hack to be a major improvement over the first two which are considered awful hacks from what I can see. Not only that but making a better hack would give off better impression of me because I don't want to be know as "the guy who goes for quantity over quality" or "the guy who puts more effort in aesthetics than in level design". I also want my next hack to have an overall positive reception, rather than being blindly overpraised (being called the best hack ever when it's objectively not the best) by noobs and heavily criticized by experienced users, the later ones I have to thank for the feedback even if they were too harsh in some points.
My birthday is getting close
First of all, some points I think I should make before starting:

-I haven't played your recently released hack, only seen parts of it from youtube, so I can't say much regarding that part, altough I think the overpraise by people not that familiar or new with ROM hacks is normal, as the first thing you see in a hack is its aestethics and for sure those people may be to a certain degree blinded by that, while experienced hackers dig deeper and might certanly see the flaws of your hack.

-It's incredibly difficult to define or explain "good level design" in my opinion, especially considering that different people may have different tastes.

That said, I will try to answer every question with my opinions regarding them.

"What makes a level boring or what make it stand out?"
"What kind of levels are considered "great" by today's standards?"
"How I can make a good but simple level?"

It all revolves, in my opinion, in the development of the level. Tipically you want your level to be centered around an idea/s. That idea might be just using a certain sprite in a certain way, or a custom uberasm, etc. The development of that idea is what makes or breaks the level. If you repeat the same idea for several levels on your hack, it will likely be boring for the player to play through. The same can be said in a smaller perspective in a level: for the development of the idea, you should vary up things (mixing the use of sprites and obstacles) while mantaining the idea or gimmick in mind; if you just use the same setups over an over, it can be boring even inside the same level. So you want to mix things up, but also restricting the variety of sprites/obstacles to a certain degree to avoid diluting the main ideas and to keep things simpler. Another important thing to keep in mind is the obstacle/sprite positions. If you just place your sprites more or less mindlessly around the screens you wont get anywhere. You should try to think: "how can I put the player in danger in his position?" "Is that enemy/setup really dangerous for Mario or he can just jump it over without thinking at all?". That last question is my rule of thumbs for sprite placement.

"How can I make a hard level that is fun rather than frustrating?"

There are obvious cases (blind jumps, kaizo traps...), but in general it might be tricky to discover if a certain section of a hack feels frustrating or unfair (inconsistent spawning, for example). The best way to discover them is just going ahead and getting a dedicated playtester to play your hack and give you feedback. Also, getting LMSW and testing on the fly your hack is also very useful.

I might also suggest not even structuring the hack in worlds? They force you to design levels in a certain way. I like to design my levels the way they come to me and in the end, if the fit within world themes, great, but if not, it's not that big of a deal. That's more of an opinion though.
Disclaimer: I haven't played your hack.. yet. Im also not level design expert.

Good level flows through well, has good pacing and has stable difficulty all the way through or alternatively gets progressively harder. Also try to focus on quality instead of quantity, I at least rather play 10 good, well thought levels instead of 100 poor quality levels(this is also something I have learned over the time).

-What makes a level boring or what make it stand out?
What makes level stand out are good palette and tileset, combined with interesting gimmick or focus on some specif sprite for example. "boring" would be just typical 105-level, its still fun but overdone so it bores most people.
-How can I make a hard level that is fun rather than frustrating?
(multiple) Checkpoints and not being too stingy with powerups, maybe even consider using some sort of health system or smb3-powerdown. Sometimes RNG can make some parts more difficult than they actually are, same with projectile enemies. Also try not to make levels/parts too compact or cramped.
-What kind of levels are considered "great" by today's standards?
This depends on who you ask, I dont mind generic levels personally, so idk.


I've not played your hack personally but I'm well aware of the poor reception. Before I continue I just want to say that while striving for improvement is always a good thing, it sucks to have a project you worked so hard on be regarded as not very good by a lot of people, so I'm sorry that the reception was as bad as it was.

-What makes a level boring or what make it stand out?
I think time and time again this community has established that good levels are the ones that are fun to play have absolutely nothing to do with aesthetics. I can't point out any specific things but I'd say look at the popular levels in hacks like JUMP and JUMPhalf for fun level design. Top 20 VLDC entries (any contest after VLDC7 really) are also great references.

-How I can make a good but simple level?
Incorporate some challenges the player has to get past, but nothing overly difficult. Hidden areas are also a plus.

-How can I make a hard level that is fun rather than frustrating?
Come up with unique ways to challenge the player that aren't overly tedious like muncher jumping, hidden blocks or enemy spam. If you want an example of a hard level that's fun, this right here is a masterclass.

-What kind of levels are considered "great" by today's standards?
That question is kind of subjective but I as I mentioned earlier, the higher rated VLDC entries and the popular levels from the JUMP games are good places to start. I've been out of touch with the hacking side of SMW for awhile now (mainly been focusing on music) so I'm sure other people will have better answers than me.

Something I'd also recommend is design your levels first and add aesthetics later. Aesthetics ARE important, but they shouldn't be the only appeal. The hack has to be fun at the very least.

I hope my and others' feedback helps you improve your levels in the future!
Layout by Erik, art by Meirdent
I've played both of your hacks. The answer to all your questions is variety.

The levels in NSMW2 would be no different if you changed the background and music in each one. There's nothing separating each level from the next one. The level aesthetics are nice and you should keep them as good as they are, but be sure the level design is unique in each level.

There are many ways you can do this and ultimately it's up to you for how you go about it. You can make use of custom ASM and stuff to make each level require some puzzle solving, or maybe one level requires an item from another level in order to complete. I dunno, think of some stuff. NSMW1 was one of my favorite hacks because it had this variety. Like using fireflowers to light up lanterns, and the minecart, and the level where everything's reversed, and the house of illusions, and the custom powerups.

NSMW2 had some things like that, like turning into a bat in a couple levels, or the going upside down thingy in 14-2. But the hack didn't do well because it didn't have enough things like that, so it got boring fast. Also House of Illusions 2 was probably some kind of medieval torture. And the new worlds were just remixed worlds from your old game.

If you add variety to NSMW3 and don't try to rush 16 worlds, then it'll probably be the best hack of Super Mario World yet.
e: Your willingness to improve is really appreciable. Most people after pumping out two huge hacks like that would probably just retire after people told them they weren't too good. I hope you can make something great eventually. But take my suggestion: when trying a new formula, try it on a smaller scale [first].

Originally posted by Pink Gold Peach
-What makes a level boring or what make it stand out?
-How I can make a good but simple level?

One answer for both: give your levels gimmicks and develop! them well. This is key: in general, a gimmick can't stand up on its own, so it's preferable to have more than one gimmick per stage, so you can integrate them and mix them to make setups the original gimmicks wouldn't render possible on their own. Don't just use a gimmick in predictable ways; no one wants to see the same old uses for darkness or the propeller powerup again. Also, avoid stage filler!

-How can I make a hard level that is fun rather than frustrating?

That's not hard to figure out; you'll get that answer when you play your levels yourself. If it gets to a point it feels like it might be too much, it probably is. Here, gathering stage-specific feedback helps a lot, so consider having people ready to test out your levels as you make them, and point out sections that are too long or frustrating. Be careful with setups that may look clever, as they may be way too finicky to execute for them to be worth keeping.
e: Also don't be powerup stingy! I just noticed this after playing one level of yours. Giving zero powerups through an entire level is ridiculous.

-What kind of levels are considered "great" by today's standards?

i think a problem your hack had was the effort you put into it. don't get me wrong - putting together a hack like this takes time, but its what that time is used for that matters. for example, you yourself stated numerous times when showing levels off that they were average, bland / boring, generic and safe, amongst other words. the rest of your level showoffs seemed to have very little info to actually interest anyone: you'd just say you used x sprite and called it a day. were you making a hack where each level was filled with love and care? or were you just filling in levels into an idea of how long you wanted this hack to be?

to answer your first question, what makes a level stand out is the care you put into it and the knowledge you have of level design. you want to make sure things flow effectively, and that you're both creative and fair to the player (unless you're making a troll hack or a kaizo with trolls in it). i'd argue even a great kaizo hack is fair to the player because they take into consideration how the player will maneuver the level and deal with obstacles.

to make a hard level that is fun, you need to still be fair to the player. lolyoshi or lazy can be good examples of designers who make hard levels while still remaining fair to the player. often, RNG elements would be considered unfair to the player: fish generators and bullet generators are two examples of unfair ways to pad out a level. a player should still be able to be rewarded in a hard level with powerups or ways to deal with the level easier if they do well - so powerup starvation or very sloppy powerup placement can make a hard experience unnecessarily harder.

i'd suggest if you're looking to make levels that are hard you should play hacks that are hard and well designed first. in fact, playing other great hacks is a good way to become a better designer yourself. try looking at top vldc levels of the past few years, or highly rated / recently featured hacks, and then play through. try to play it critically: think about what the designer was doing when they made it - why they placed enemies where they did, what functions those enemies had, how they moved with the landscape around them, etc. try to avoid just placing things for the sake of placing them, and try to avoid placing enemies that are either completely out of the players way (enemies that fall off ledges before you get to them, etc) or interact with the level jankily (maybe something like jumping koopas getting stuck in recesses in the level).

i also think you won't be able to achieve any of this until you move on from your last hack and feel genuinely motivated to become a better designer. i feel this would also include ceasing to complain about your hack and your skills on the various Discords you frequent, and your second most recent cry for help signature that said you're the worst smw hacker out there. if you don't take yourself seriously and be genuine to yourself, you're only going to get more frustrated with your skills.

finally, i second koopsters notion that if you're wanting to learn level design, don't use it on a monolithic hack. in fact, i'd argue that designing a hack where you draw out how many levels are in each world and work on it linearly usually ends up in an unenjoyable experience anyways. just come up with a rough hack idea, and then start making levels. make them in whatever order, make them creative, introduce mechanics through playing rather than telling and then expand upon those mechanics in a level. but most of all, actually care about what you make. think about each setup, think about each enemy placement and think about the level as a whole. only finish a level when you feel happy with it, and then get playtesting or insight gradually as you go.
I played seven levels of your hack, and so far, I think I am beginning to see that the problem is not that the design is poor. On the contrary, I've been very satisfied with the challenge and gameplay. The levels are linear but not dull. What I've played so far has been really good overall. It seems to be later levels that get complaints, so I am going to guess that the design issue has more to do with poor evolution rather than a lack of talent.

Pay attention to specific examples given to you of issues people mention and play those parts yourself. Gauge how much fun you're having as you do. If you recognize these parts where you aren't enjoying yourself, focus your attention there and try to make it fun. If you don't enjoy a part of your level, likely other players won't either.

SMWC members who have helped make Ganymede the hack I want it to be
From my point of view, the hacks is good, though some levels are linear, I think it's being too underrated. Maybe the difficulty of the latest worlds was too high, but it's a good hack.
Anyway, if you're going to make a possible sequel, you can use LX5's powerups you used in NSMW1, you could make some interesting gimmicks from that. Keep using LM'S big levels, they were great to play (except for the Ghost House from Abstract World, which was a bit unfair). You can also use more graphics which are not from SMW.
1. A level gets interested when you have a genuine thene or idea for a level. Don't use too many sprites, but still keep the level playing in a different and more interesting. Repetitive and unfair design will make a level bad.

2. Before making a level, you should think how the level will be and how it plays. Having a strech of level design or imagination makes the level way better then making a random level for the win.

3. Avoid Sprite Spam. When the game causes slowdown, there are definitly too many sprited. You should reduce sprites during a generator. And don't make it too hard. When the player dies, they have to play the same level over and over again which gets fruststing. Test your levels always without savestates or other tools.

4. Great levels are levels with an interesting concept and good gameplay.
Don't make it too easy or too hard. Consider that many SMWC users are skilled players.

Also while there are some flaws, I don't agree with the comment section.
I recently played this hack and I had lot of fun. The Autosave patch is really great and the levels were amazing. Some of the levels were really impressive and the hack has a valid difficult curve (most hacks fail at that)

Hacks always get backlash no matter how good or bad it is. People don't know what effort means and attempt to create problems for simple mistakes or "it's not their style of hack". There isn't really a way go around it, especially when you are a big creator on the site.

Listen to Feedback and consider the facts for your next hack (If there will be one).
As idol already said, play other hard hacks with good design.

Figure out what makes them good designed and take that konwledge into your design.

Some rules i give myself when desinging a level.

1. If it has a gimmmick, 'show' the gimmick in a area the player has time to figure out what the gimmick will be.

2. Use the gimmick in OTHER ways than in the shown area.

3. Maybe put in a second gimmick. Show that second gimmick also.

4. Mix them togheter

5. Never think 'Oh, that unfair hit here is okay, that will not happen to anyone else'. If you see a section that can lead into an unavoidable hit, you need to change it.

Sometimes i build a section, test it, see that it's unfair, rebuild it, see that it's still unfair and i just build a complete new section.
Levels that are considered great by today's standards undergo enough testing so you and subsequently other players are capable of deeming them fun. You don't slap gimmicks for the sake of existing and calling it a day, you have to find a way so they aren't too stingy [I didn't play your hack yet so I cannot say for sure]. Don't move forward to your next hack without fixing the issues the current one has first.
Also give priority to not rushing just to release it on a certain date. You admitted that yourself. That's why the level design was largely criticized the way it did.
I'll happily give some more in-depth feedback seeing as my review of your hack was short and sweet as well as been to the point.

Looking back and playing through the first three levels in World 1 I saw the immediate pattern take effect.

World 1-1

Pretty standard and safe for a first level which is too be expected. Nothing really wrong with it other than it been generic.
It doesn't help that it's Grassland themed as that will be another issue we get onto later on.

Honestly this level is only bad as the Tutorial Level prior introduces a small walljump required section which is longer than anything you can walljump on in the first main level.
I'll dive on this topic later as well.

World 1-2

You introduce a new enemy (Sleeping Goomba) in a rather impractical way. You go past the first screen and bam it starts chasing you.
Having it in an area you can't access or having more time to react to it would of been a better introduction.

Once introduced correctly you can then use it in surprising ways to keep the players on their toes as they'll already feel comfortable getting around said enemy.
Once again there are no real emphasis on the walljump and we are in the second level. Another point for a more in depth section.

World 1-3

Mushroom Platforms had some interesting layouts however the Hidden Passages do seem sorta random granted nothing too bad.
In the Switch Palace Area you do have a Sleeping Goomba hidden behind one in which the player can easily hit due to having no prior warning to it been there.

That's all for the levels I decided to thoroughly look into again as these three alone paint a pretty good picture for the hack.

Before I continue I just want to say don't think I am talking with an ego or some idea that I am the best SMW hacker ever.
I look back at my old work constantly and it could of been waaaaaay better. My Super "Mario" World 2 hack really goes insane with bullshit hard level design in the latter levels which just is not fun.

Over the past 10 years I have been trying to "perfect" my level design philosophies and how I tackle building structured levels.
I want to pass on some advice from what I have learnt so you can maybe go away and see what you can improve seeing as you are asking for advice.
Anyways lets move onto some of the biggest points I'd like to discuss about your level design and hack in general.

Generic Isn't Bad It's Just Not Interesting

Something I find to be odd is you have really distinct levels later in the hack yet for many of your world and level themes you keep the visuals and overall style very safe.

Your first three levels all blend together bar different palettes. It also does not help that you used a Grassland for the first theme.
It's not bad as everyone has done it. It's just I feel we need to stray away from the Standard Mario Formula and be more creative.

Why can't World 1 be a hot Savannah themed world with dry plains and small overgrown areas.
Heck throw in some lava as well why just save that for castles and later worlds? (looking at you every World 8 in 90% of hacks...)

If you have not seen of played Donkey Kong Tropical Freeze I urge you go look at game for level aesthetics. Each level in World 1 has its own thing to set it a part and that theme follows throughout the entire game.
Don't limit yourself to the common norms we see in most Mario games. Break the mould and think outside of the box.
Having a really interesting set of levels early one with a distinct theme will keep players interested in continuing to play your hack as well as making it stand out and be generic.

Plenty Of Gimmicks But No Twists

You do introduce new enemies and gimmicks regularly which is great however you don't add twists or new spins to them. They show up and then they'll show up again later on.

For Twists introduce a concept in a way the player will grasp quickly. Build upon and test their skills with it be it a challenging enemy or platforming section that focuses around that gimmick.
Then introduce the twist. It'll be the same idea but you have added in a new element.
This can be a new enemy or block or anything that changes the way you go around using the existing gimmick making the player think hard and thus leading to more rewarding experience.

I recommend looking at Mark Brown's video on Youtube about Super Mario 3D World and it's level design structure. It's essentially what I've just said plus it's a great watch as well.

Less Is More: "Quality Over Quantity"

Honestly a lot of this hack could of been condensed down into fewer levels and fewer worlds focusing more on stronger designed levels and more planned out gimmicks for each level to really bring about a greater sense of identity.
If you levels all blend together it's hard to pick one out or remember anything that was truly worth going "oh yeah I remember that level it was really cool".


If you have time to make all these cool Vanilla SMW styled GFX for levels then you can try harder and make each level stand out more. Not every level in a grassland should look identical.
I know you can do more but I do know how easy it is to just re-use assets to save time.

Some of your levels really do pop so I'd say it's like a Double Edged weakness for you. Nice Visuals but Weak Design. Get the balance down but focus on design first. It can look like shit but if its fun to play its fun to play.


Like most hacks Mario feels like SMW Mario. Adding in a walljump means you don't have to abide by the same rules and layout as standard Mario levels. In the New Series walljumping is a thing but its not heavily utilised.
Like in this hack there are only a few levels that heavily use it and it's usage is sparse in between those said levels.
Why have it if you won't use it? Why make levels that are simple go left to right? With LM3.0 get Vertical with your levels if you have the walljump as the level layouts expand tenfold.

Closing Points & Conclusion

If you have the time to go all out on a hack this size you are more than capable of spending time to focus on building on your hacking skills.
I'll say though all the criticism I have seen is fair. It must be striking a cord with you if it's enough to make you post this thread... and guess what that's a good thing. You are wanting to improve and well done.

You could easily made the third hack in this series and called it a day by mimicking the past two and playing it safe.
From this point going forward I'd say work on a small hack of a few levels working on your designing skills. Let people play it and get more feedback. Once you have it pretty much down you can then start building back up to a hack of this size again.

Granted you can do what you want at the end of the day but like you have said yourself if you don't wanna be known as someone who makes generic styled hacks you'll have to jump out your comfort zone and take some risks.
I'll always remember a distinct hack even if the levels are odd over something that looks like the same type of stuff that is constantly been pumped out. Once again this is to not insult or discourage people from making more generic styled hacks.
It's just if you want to stand out you gotta make your work actually STAND OUT.

I believe you can do it and I hope you read all of this.
If it's too long then please read at your own time and pace but when I see your next project I'll be looking to see what you are doing differently.

So yeah Good Luck. I know you can do it. #smrpg{y}
Feel free to PM me or hit me up on the Discord to talk more if needed.
I commend you for seeking feedback and advice, as that is one of the best ways to improve at not just level design, but really anything.

So you can learn all the design tips and tricks and read all the tutorials on design, but none of that stuff really helps that much until you begin to change the way you think about designing a level.

My advice to you would be to try making a couple levels just for fun that won't go into a big hack. Also, don't show them off on the forums either. You can always release these at a later date when they are finished, but for now, just design "by yourself, for yourself". I think that you'll be able to get the "creative juices flowing" better if you focus and put your attention on these one or two levels.

The best levels nowadays are really subjective and different for everyone. But what makes a level stand out usually is, like idol mentioned, that it has specific attention put into it. I think that if you focus on a "one level at a time" mentality, instead of trying to create the hack first and then make levels for it, you'll find that each level feels unique.
Pages: « 1 »
Forum Index - SMW Hacking - SMW Hacking Discussion - How can I improve my level design?

The purpose of this site is not to distribute copyrighted material, but to honor one of our favourite games.

Copyright © 2005 - 2019 - SMW Central
Legal Information - Privacy Policy - Link To Us

Total queries: 9


Follow Us On

  • YouTube
  • Twitch
  • Twitter


  • Talkhaus
  • SMBX Community
  • GTx0
  • Super Luigi Bros
  • MFGG
  • Gaming Reinvented