How many levels should I make?

Question

One issue I grapple with across many of my games is trying to decide how many levels to put in my game. This is across genres and platforms.

In general, constraints that may decide -- although they are not applicable in my case -- can include:

• Pre-determined schedule or release dates
• Fixed budget
• The game's story finishes/concludes
• (PCG games): difficulty becomes "impossible"

These are all great, but not sufficiently constraining as to tell me how many levels to make.

Another constraint that sounds good in theory but is hard to implement is game time -- for example, Mario or Super Meat Boy, I can guesstimate x minutes per level, and target y minutes of intended total gameplay, yielding y/x levels.

But none of these seem quite right to me. It seems like there should be a better way to decide when to add more levels and when to stop.

Answer 1

Asking how many levels to include in a game is sort of like asking how many paragraphs to include in your essay, or footsteps to take in your journey. Once you have a clear idea of your destination and your path, the question answers itself.

But it is a question worth considering because it leads to some interesting observations about game design.

What is a level?

Let's take a step back and look at what a level is, what it does, and what is its purpose.

• It introduces something new to the game, such as:
  • Higher difficulty e.g. space invaders, pacman
  • A new mechanic or puzzle e.g. donkey kong
  • Progressing the story
• It's used to control pacing, giving a temporary lull in the action. Sometimes entire levels are devoted to this purpose, e.g. bonus stages
• It rewards the player, giving them a sense of progress e.g. the castle fireworks in Super Mario Bros
• It is often used as a checkpoint, somewhere players can resume if they fail in the game
• Historically, there were technical limitations to how much content you can have all together, so the game was split up into levels

Usually levels are used for many of the above reasons, but you may have noticed a common thread: levels encompass a unit of content, a self-packaged experience meant to be consumed as a whole. That isn't to say you must consume one unit of content at a time, but usually you (are encouraged or compelled to) finish playing after the end of a unit.

Units of content occur in many places, outside video games too, and there are many ways artists/authors encourage audiences to consume these in units. Books have chapters, and readers often - if they have a choice - stop and start reading between chapters. TV series have episodes, where each episode covers a self-contained story, sometimes within a greater story arc. Some media have only one unit of content, for instance movies are best enjoyed in a single sitting.

So it is with games; some have lots of small units of content, e.g. Angry Birds. Some have a single large unit of content, e.g. Civilization, where you are never encouraged to take a break until you've finally conquered the world, some 20-30 hours later. (Some days I think Sid Meier isn't aware that his players need to sleep.)

I believe the question "how many levels should I make" is better answered if you split it into two:

• How long should my "units of content" be?
• How long should my game be?

And divide the second by the first to get your answer.

How long should my levels be?

There are many factors that will affect the length of your levels:

How long do I want my players to play the game, at a time?

This goes right to the heart of your game design. If I'm standing in a line, I'd play Angry Birds or Candy Crush; if I were at home on a lazy weekend, I have more time and attention available - I could watch a movie.

How much time is necessary to teach the player a new mechanic?

This depends on the complexity of your game. A particularly complex strategy game might require long levels.

How much content do I need to meaningfully progress the plot?

For story-driven games, the appeal comes from seeing the plot unfold, and if your levels are too short there is simply not enough time to meaningfully move forward with your plot.

How long should my game be?

The answer to this one depends on quite a few things I've touched earlier. Your game should be:

• Long enough to tell the story, but your levels should be meaningful in terms of plot
• Long enough to teach the player about the game, but your levels should avoid being too repetitive
• Long enough to keep players interested, but keep in mind if it's worth the budget!

Sometimes there is no upper limit to the length of your game; you can keep making it longer so long as the cash is flowing in and the players are there. Just as they keep making episodes of The Simpsons by adding characters or even rehashing plots, you can add levels, new mechanics, rehash old mechanics... and sometimes the players love it!

Answer 2

Based on my experience (mainly on the action/puzzle genre), I could only say to you that it depends on the features.

For example when I had a new set of puzzle logic, I needed 1-2 levels to introduce the feature, then 3-12 levels to build it up and complicate it / fix the difficulty curve, and then maybe in those 3-12 or in some extras I would mix this new feature with previous ones that appeared.

A great example of how level design works is Cut the Rope! They really nailed this.

Hope you find my experience useful.

Answer 3

A somewhat viable choice is to have a constant of levels you want to make that aims at some % of the budget( and or time) like 70 or 80. Once you have achieved that you have some freedom for the remaining time/money.

The options are almost limitless: you can start adding achievements (if you haven't already) You can add extra levels, challenges, bonus content like skins, rewards add extra rooms to some levels.

tl;dr: Finish the game with not the optimal amount of levels for the time/resources add extra content.

Edit: The X minutes per level is impossible to make, since how fast the player will go through the level depends entirely on his play-style experience. For example an FPS veteran will speed through CoD missions (let's consider that they will be played) where as an RTS player will slowly and methodically go through each part of the level.

Answer 4

I like this question because it is something most developers have to face.

The only thing I would like to add, is that you NEED to have your game played by impartial, virgin (they haven't played it before) testers.

You could put in 100 levels to get X hours of gameplay, but if it is you who is testing (you the programmer who knows the game's algorithms and quirks, you the gamer who is already an expert with the controls, you the designer who knows the level design and shortcuts like the back of your hand) then your calculations may be way off. When it comes to new players, you may only need 20 levels to get the same X hours of fun.

I speak from experience as a lone indie developer.

Answer 5

Here is a formula I use for iOS specifically. In iOS you're allowed to create about 25 leaderboards / gamecenter boards. So depending on what you want to show as leading metrics such as time or score you can have 25 levels. If you have leaderboards with both metrics, you can only have 12 levels a piece. 12 leaderboards for time, and 12 for score. This is what I generally use for iOS only, not sure what it matriculates to on other platforms.

Of course you don't have to have game center integration and you can create as many levels as you want! Or you can aggregate multiple times/scores into a single leaderboard(s). Just some food for thought.