So you've decided to write an Idle Game, because you wanted to have something fun run in the background or watch numbers go up. As it turns out, it's not as simple - especially if you don't want to make Cookie Clicker Clone #1256.
Idle games can be pretty liberal in terms of design, with the genre having plenty of potential for evolution. It's very much gameplay-centric, nobody's going to worry about graphics too much if they do their job or even stick to text, and chances are that if you put music in the game people will turn it off and listen to something else while doing their thing. You can leave a lot of aspects to the players' imagination.
Necessary Tropes
It needs to be fun and want you to keep playing. Don't worry about accuracy if it ruins gameplay.Adam Smith Hates Your Guts almost always must be in effect. For example, every farm you have produces 10 carrots, but the farm's base cost is multiplied by 1.05^farms. In essence, their effects go up multiplicatively, but their cost grows exponentially. You can try making their effects go up exponentially of course, you'll just need a stronger exponent for the cost, or potentially even tetration. Just make sure the growth of profits of one thing scales slower than its cost.
There must be some kind of Gameplay Automation. The most basic is just producing resources or dealing damage every second. More complex games will enable production of lower-tier producers, often called polynomial growth.
To add more progression and longevity, you can add Pinball Scoring. Sure, millions are nice, but how about 2e30 (2*10^30)? Or make the power its own million at 1e200,000,000? Or outright use a higher level notation, like ee100 to show something like 10^(10^(100))? Or count the number of times you're raising the power of 10 by itself with the F notation? Or try something even bigger? Just make sure that it's actually satisfying and not just big for the sake of making them big, the gameplay actually supports the idle aspect, you can balance it around the enormous numbers, and have the necessary programming abilities for the respective scripts.
A New Game Plus mechanic where you reset the game or all previous resources to gain a new one entirely can extend playtime positively. Remember that the players want to see the benefits, but generally getting back to the point you originally were in the first playthrough should be noticeably faster. You can add multiple layers of prestige to add even more content (so you lose coins and coin-based buildings upon resetting for diamonds, but you lose diamonds and diamond-based buildings in addition to coin stuff if you reset for mythos).
Another nice way to add longevity is a Challenge Run system, where the player has to reach a certain goal with a disadvantage (like increasing costs of a producer if you buy others or limiting the use of a certain multiplier). The player should be expected to beat a challenge either when they have made a decent amount of progress without it getting tedious or figured out a strategy. Once it's completed, they should get a decent reward to reach a further point in the main game. It's also possible to make challenges beatable multiple times, increase either difficulty or the goal with each subsequent completion, as well as the reward.
If you get to the point where you have multiple layers of content, you'll want to add bulk buy options or autobuyers, because no-one will want to buy the 10,234th mine manually, especially when you have 60 superfactories. You can even make lower NG+ layers not reset anything or have them give you a percentage of their respective resource per second, treating them as just another smaller part of the game.
If something becomes too overpowered or progresses too quickly for your intended vision, it may be wise to implement Diminishing Returns for Balance or a softcap. For example, you can softcap a value above 100 by turning 100+x into 100+sqrt(x), which results in 136 becoming 106. Remember that a softcap can make things more difficult for the player to follow, so it may be worth trying a different formula first. However, you can also try to implement a mechanic that weakens its effects later on, for example if an upgrade makes it start later by 20, the calculation uses 120+sqrt(x) instead.
To ensure each session goes smoothly and you can quit at almost any time, make sure to implement a reliable method of saving progress. An Autosave system is almost expected to be present with a save at least once every few minutes and some subtle notification when the game is saved/was last saved, but let the players use a manual save button in case they want to quit at any time. To carry over their progress on another device or back it up just in case, consider adding an Export Save system so they can paste it into another file or download a file from the game or cloud saving for easy access.
Pitfalls
Long timewalls often irk people due to not providing anything to accomodate for an active playstyle and slowing down progress. Yes, I need to wait 20 minutes to get the necessary bills for the big gun and have nothing to really spend them on.An overly cluttered interface with Billions of Buttons can be an absolute nightmare to navigate, which is all too easy to do if you have at least three complex NG+ layers. Even worse is if you show them all right off the gate, potentially confusing new players and not letting them know where to go. Make sure that you only show the necessary stuff for the beginning of the game, with a teaser of the next mechanic or two, and check if the interface is intuitive enough to navigate later on.
Microtransactions tend to interfere with the balancing, potentially making the game too slow if you try to play for free. Ironically, it can technically making people pay to get done with the game faster, which can suggest it's not enjoyable on its own. If you intend to make money, consider something like "double producer efficiency and no ads for 5 dollars" (cost can range, but don't go above 10 bucks), or just make the game itself something to purchase for a low price (anything above 10$ will be seen as greedy, unless you make something really good and worth the cash).
Potential Subversions
You don't have to make an Endless Game or rely on No Plot? No Problem!. Feel free to put an end screen at the point where the players have experienced about everything there is to offer. You may try adding Post-End Game Content to let them keep playing and put some extra content that would only truly interest the hardcore people (see Idle Mine Remix, which has the main story end shortly after the e200$ mark, but the game challenges you to reach e1233$, and breaking e3003$ and farther is certainly possible).Some games actually manage to avert Adam Smith Hates Your Guts. The Unscaled Incremental never raises its dimension prices. However, it requires the use of hyper-E operators to handle the notation and still makes upgrades increase in price, so it's not possible to really have a game with no cost increases.
The Greats
- Antimatter Dimensions and its remake of sorts FE000000 have well-made layers with plenty of content, the former in particular making the concept of polynomial growth more fleshed out.
- Universal Paperclips manages to avoid the common No Plot? No Problem! issues of the genre, as well as deconstructing the endless number-go-up cycle that define idle games, with an Artificial Intelligence going from simply an assistant at a paperclip factory to a world leader in the industry to turning the entire Earth, and eventually the entire universe into paperclips.
- Time Clickers and Time Warpers combine idling with shooting blocks with a cool futuristic aesthetic and a satisfying Tech Tree accesible after performing a Time Warp. The latter also has a world to explore, even if it repeats.
- Synergism has tons of mechanics revealed as the player goes along and great balancing.
- Idle Mine Remix manages to take Idle Mine and remake it with more modern sensibilities, with four upgrades types for different currencies, progression that makes it difficult to actually run into a wall and a story.
- Prestige Tree and some of its more fleshed-out mods like The Prestreestuck use lots of prestige layers to add lots of variety during gameplay and have an intuitive tree-styled interface.
The Epic Fails
- Reinhardt's House has a cool atmosphere and some unique ideas, but it ultimately is poor thanks to its combination of false advertising, repetitiveness, and effectively turning unplayable at the second nightmare layer.
