Rust & Revenue started with one bot. It was straightforward, predictable, and got the job done. Then came two bots, both running the same script. The game state became a little more dynamic, but they still felt like echoes of each other.

The real shift happened when I added the random bot as the investor. That little bit of chaos introduced friction and tension. For the retail version, I went further: the bots can switch scripts mid-game. It’s not about deep strategic simulation. It’s about creating the effect of different players making different choices.

The real magic came when I added a random shuffle of scripts. At the start of each round, the bots trade scripts for a random 1 to 3 turns before returning to their usual patterns. That simple shuffle gives the illusion of different personalities at the table. It opens up the game in ways that a single, fixed script never could. Situations emerge that I couldn’t have predicted, and that’s exactly what solo play often lacks.

Just as importantly, the random/scripted dance keeps solo maintenance manageable. The scripted side only needs to be called once per action, while the random side is simply checking a box. The random bot also tends to take up valuable shares later in the game, and it can quietly win by attrition if you’re not paying attention. Providing both behaviors together created a more cohesive solo experience: structured enough to be believable, unpredictable enough to stay interesting.

Train games are built on numbers, values, and expected outcomes. That kind of math can bury a solo player fast. From the beginning, one of my design goals was to make the bots easy to run and believable at the table. I didn’t want a wall of flowcharts. I didn’t want to simulate every calculation. I just wanted opponents that created pressure and got out of the way.

That balance - believable behavior with minimal bookkeeping - is the core of why multiple bots work in Rust & Revenue. They don’t replicate human play, and they don’t need to. What they do is create the shape of competition. And in a train game, if there’s no tension, no emerging conflicts, no shifting pressure, it isn’t a game. It’s just math.