<aside> 🌱 We're making a game MVP. In other words, Minimum Viable Fun: the game should be genuinely fun to play.
</aside>
POC: https://github.com/amethyst/shotcaller
In the way "MOBA" games such as DOTA2 or LoL are usually played, the captain of the team is the default shotcaller.
The shotcaller needs to be unbiased and not have tunnel vision. You need to be able to think in the future and discern what would happen if you make one call or the other. This becomes crucial when deciding to base-race or teleport back to defend.
In a traditional MOBA, everyone on the team can play the part of Shotcaller on occasion. The act of shotcalling is not the most prevalent activity of any player, even for a captain — after all, they also need to play their hero.
But in this game, all you do is shotcalling and big-picture strategizing. The game plays as if you were controlling the 6th-person-in-the-booth “coach” player, and your team (of AI-played bots) actually follows your instructions to the letter, within their designed constraints.
Reaction times do not matter in this game. While it plays out in real-time, only a select number of Action Per Minute are allowed, and instructions are performed in batches rather than instantly. This also means that the game is highly latency-resistant.
By being text-based first we ensure that the game will be machine-readable and can thus be parsed by various types of assistive technology, such as screen readers. It’s no fun to be the best at something unless everyone is invited to play.
Modifying the behavior of your individual bots is part of the game. By separating matchmaking into different tiers of customization, players will only be matched with other players who are modifying the game to the same degree, within the same constraints.