Bots and solo practice: how to play when your team isn't full

The practical problem bots solve

Coordinating five people who are free at the same time, with the same energy and the same enthusiasm to play, isn't always easy. Someone gets sick on session day, a team only has three people free that week, or someone just wants to get a feel for a game before inviting the rest of their team. That's exactly what bots are for: teammates that play on the server side, held to the same standard as a real person.

What bots are (and aren't)

Worth being upfront about this: Experiencia RPG's bots are heuristic policies — rules, memory and some reproducible randomness — not generative AI or machine learning. They don't "see" the full game state to solve it optimally; they decide using the same filtered view a real person would see from that seat, and they react to what the team says in chat, the same way a human teammate would react to the information available to them.

That's the whole point of making solo practice mean something: if a bot "cheated" by seeing information a real player wouldn't see, the practice wouldn't train anything real. By limiting the bot to exactly what a player in that role would see, playing with bots ends up feeling a lot like playing with people.

The three bot levels

Bots come in three levels: novice, competent and expert. The level affects how fast they react, how well they prioritize among several possible actions, and how much "noise" (occasional suboptimal decisions) shows up in their behavior. A lower level makes the session more forgiving while someone is still learning the mechanics; an expert level builds a tougher match for practicing under more pressure.

When it makes sense to play solo with bots

To understand a game before inviting your team. If you're weighing which of the six games to pick for your next session (our guide to choosing a game can help), playing a round solo with bots gives you a real feel for the mechanics in 20 or 30 minutes, without having to coordinate anyone's calendar.

To fill out an incomplete team. If you planned on five players and only three show up, bots cover the missing roles and the session goes ahead without losing the scheduled slot. That's especially valuable when the session was planned as part of a program — onboarding, for instance — where rescheduling carries a real coordination cost.

To practice before facilitating as a Game Master. If you're about to facilitate a session for the first time, playing a full round with bots lets you learn the game's flow end to end — including what hints to give and when — before doing it with your actual team.

How they behave in practice

Bots talk in the team chat just like a person would, and react to what's said there. If a human player gives a clear instruction in chat, a well-designed bot follows it with the same logic a real, attentive person would use. That means playing with a mix of people and bots — say, a team of four humans plus a bot covering the fifth role — feels natural, not like playing against some separate machine.

A concrete example: filling out a Blind Operations team

Imagine you set up a session of Blind Operations for five people, but on the day only three show up: Operator, Analyst and Coordinator. Without bots, the session gets cancelled, or plays out with empty roles that break the mechanics — nobody has the map or the procedures manual. With bots, an artificial Navigator and Specialist join the team, reading exactly the same limited information a real person in that role would read, and speaking up in chat whenever there's something they need to share. The human players still have to actively communicate with those two artificial teammates to resolve the five systems, so the session keeps its training value even with two people missing.

Why this is different from "playing against the computer"

In most games, "playing against the computer" means facing an opponent. Here it's the opposite: the bot is a teammate, not a rival. Its goal is the same as yours — getting the whole team to the game's objective — so the dynamic stays cooperative even when part of the team is artificial.

The honest limit of practicing with bots

Practicing with bots is a great way to learn a game's mechanics and avoid losing a scheduled session, but it doesn't fully replace playing with real people: part of the value of a cooperative game is genuine human communication, with all its nuance, hesitation and personality. Bots are a tool for practice and continuity, not a permanent substitute for the real team.

Bots learn from the chat, not the full match history

It's worth calling out a technical detail that reinforces just how honest this practice is: a bot decides its next move by looking at what recently happened in the team chat and its own filtered view of the game state — not the full decision history of the match, and not information from other roles. That means if a human player never communicates something relevant, the bot won't "know" it some other way either: explicit communication is still the only way information travels between roles, exactly as if every seat were filled by a person.

How to get started

You can try any of the six games in solo mode right now, choosing whichever bot level you like. When you're ready for the session with your full team, set it up from the platform.