Play the WebGL version here: https://swivelmaster.itch.io/troll-farm
Mobile links on the itch page too.
Troll Farm was initially released in 2018 as a solo project with UI/UX design help from Nathaniel Fong, and then was updated in 2020 with some additional writing by Mandy J. Watson.
I’ve always had a soft spot for idle games and wanted to build a ‘narrative idle game’ that used idle mechanics to convey a fairly linear narrative. The story of the game is based on (and a reaction to) the many revelations about who and what was driving online engagement during the 2016 election, including Russia’s meddling as well as folks like Steve Bannon and Rupert Murdoch.
In the game, the player takes the role of an anonymous political player who runs a ‘troll farm’ to drive online engagement in exchange for money (and eventually political influence). The content escalates somewhat realistically from basic trolling to Gamergate-style harassment campaigns to actually running TV networks… and more.
One of the key design decisions that drives the game is that all of the narration is written in what I would call the “sociopathic second-person” – essentially, any tragic real-world consequences of the player’s actions are conveyed as either completely matter-of-fact, or framed as negative only when it impacts the player’s operation. This is meant to reinforce the idea that someone who would do these things truly has no regard for the negative impact they’re having on the world, focusing only on some combination of their personal mission and the fortune and infamy that it brings about.
The game can be completed in 60-90 minutes and has a deliberately unsettling ending in which the player has basically taken over the entire country, but the troll machine they’ve built just keeps going until they manually reset the game.
Some technical and design notes:
- The game’s data is mostly driven by CSV exported from Google Sheets, with just a few hard coded items.
- The key innovation of the game (and its primary method of storytelling) is Random Events. These are basically short news articles that show the negative outcomes of the user’s actions, with some kind of currency penalty to match. These scale in currency penalty and dramatic severity, but not quite at the same rate – by the time the player is buying politicians, most of the negative events (significant as they are) barely make a dent in the player’s finances.
- For balance, I checked out some formulas from Clicker Heroes for inspiration, but ultimately built my own spreadsheet from scratch. The important part was that each ‘generator’ have a ‘Base Efficiency’ that decreased the further players got in the same, but that the increase in price with each subsequent unit purchased would also drive that efficiency down. This way, the most efficient strategy is (as is common in Idle games) to buy N of the first generator, then wait to by Q of the next generator, rather than just continuing to buy the first generator over and over again.
- Upgrades increase the output of a class of generators, thus putting a multiplier on their base efficiency and making it worth buying a few more.
You can view the spreadsheet that drives the design of the game here:
Some of my favorite feedback on the game from Reddit:
And from itch: