Hello everyone,

My name is Dominic Webster and I am currently in my last 5 months of the BS degree in game development and programming. My biggest interest in games has always been the worlds inside them. When watching movies or TV shows you get to see incredible worlds, but inside video games not only do you see them, but you experience them. For my first post I want to know what makes a good level to you? When looking at a large open world and a small more traditional level, does the criteria of a good level change? Or are there specific criteria that need to be in any level to make it good? Finally, what are your favorite levels and worlds in video games?

For me, my favorite worlds are Red Dead Redemption, Zelda Breath of the Wild, Inside, and Celeste. These worlds are all so unique and handle level design in their own way. Inside created one level. If you don’t die in the game, the game has no loading screens, and you can walk through the whole game. Celeste has unique short levels with loads of character, and design. The precision needed to complete a level is incredible, not to mention the precision needed to complete a B-side level or to obtain all the strawberries. RDR and BOTW are vast worlds, and you know what part of the map you are in based on your surroundings. RDR feels lived in and has dynamic weather, whereas in BOTW you can go anywhere you can see.

One final question. If you are currently a level designer, what advice do you have to me and any other person looking to get into this specialization?

I'm not personally a level designer, but I do have some tips.  Make sure you're worlds are playable.  It can be linear, but don't force people to follow one single path with no deviation, like enemies in a tower defense game.  Think destiny.  You still have to go to predetermined locations, but you can also explore the environment to a pretty solid degree.  It's not open world, but it still feels open.  Another tip is to not make your worlds look like a game.  The more natural and realistically built they are, the better.  Placing tons of obvious cover that's just there to be cover is a bad idea.  Place props that would logically be there, such as a table in the middle of a room surrounded by chairs.  If a person or nature would place it there, it's good.  If you're developing open-world, use procedural generation to build it, then handcraft the details after.  It will save days and still look great.  So in summary, don't keep the world too linear, and think: "why is the prop there?"

Adding some tips on top of what Pepsidog said, I think that Level Design should focus on reinforcing the gameplay, adding challenge and guiding the player at the same time. You should aim to give the players interesting things to do and not let them feel neither lost nor unfairly defeated. This is not always simple, and some mistakes can be made even in large AAA productions, but its a nice goal to have, irrespect of the game style (the ones that require level design, of course).

How you do that depends on what kind of game you're working. A FPS style level can be very different from a RPG or platformer. Playing these games and paying attention to these things, as you seem to be doing, is a good start, but you need to go beyond, asking why things were made that way, and thinking what could have been done differently.

There are tons of material that explain the level 1-1 from Super Mario Bros, a relatively simple level, that teaches the player many things that will be used through the whole game, and provide a nice initial challenge. Check this video from Game Makers Tool Kit (and other videos while you're there) to get an idea on how to build a Mario level. From there, you can check some GDC videos and other talks, and look for articles and books on level design.

Also, you can look at what to avoid. For example, many Mario Maker levels rely on extreme gimmicks and absurd difficult (example), just because they can (and memes and Youtube/Twitch views). They're not necessarily bad levels, but not very interesting for the average Mario player. Nintendo probably would never release levels this kind of levels on a standard Mario game.

21 hours ago, dawebster said:

One final question. If you are currently a level designer, what advice do you have to me and any other person looking to get into this specialization?

As any other area in game development (and life in general), practice and more practice (studying in parallel, of course). Being more specific, get a game with a level editor and create some levels. Not necessarily for all types of games, but at least the kind of levels you would wish to do on a day to day job.

For example, you can download Unity or Unreal, some free assets or even entire projects (this one or this one, for example), and try to create your own levels from it. Games with level editor (such as Mario Maker or Little Big Planet) are also a good start. But don't go blindly. Research first, draft some things, discover what you want the level to achieve. Then start building.

I agree with @TerraSkilll.  The quality of a level depends on how efficiently it serves the gameplay.  Levels that served Quake 3 wouldn't have benefited Half-Life nearly as well, and vice versa.

For a portfolio, I recommend making levels that each serve a different purpose (a level for a platformer, a level for an arena shooter, a level for a linear shooter, and etc.).  You don't know what kinds of levels you'll be making when you get hired, and it's not unlikely that you'll need to adapt after you get hired.