What decision training is, and why it matters

People are taught how to make decisions the way they’re taught how to swim — by being told, often clearly, what to do, and then dropped into the water to find out whether they got it. For day-to-day choices that’s fine. For life-defining ones, it’s terrible.

Decision training is the alternative: structured practice for the decisions a person will face in real life, built so the wrong answer is recoverable in the simulation, not in the world. This post explains what that means in concrete terms, why we think it matters most for high-stakes transitions, and how Next Move™ is building the category.

What “high-stakes” actually means

The transitions we focus on share a feature that most life decisions don’t: the consequences of getting it wrong are large, often irreversible, and they cluster in time.

Take reentry. In the first thirty days after release from incarceration, a person typically faces something like 38 distinct decisions — where to sleep, which supervision officer to talk to first, whether to take the first job offered or wait for a better one, when to disclose a record on housing applications, how to enroll in Medicaid, whether to pay a parole fine in cash or money order, which prior addresses to list, and on and on. Each of those decisions has paths that lead to stable rebuilding and paths that lead back inside. There’s almost no built-in margin for error.

A new veteran navigating the VA, a young person aging out of foster care, a survivor leaving a violent relationship — different specifics, same structural problem. A lot of consequential decisions, all at once, with rules that are not obvious to anyone who hasn’t seen them before.

Why training-by-being-told doesn’t work here

The default for these audiences is some combination of paper handouts, classroom sessions, peer mentors, and one-on-one case management. Each of those has value. None of them replicates the cognitive load of the actual moment, which is the moment people fail in.

You can know that a particular landlord will reject you with a felony conviction on the application and still flinch when you’re filling it out. You can be told that a parole officer will respond better to a calm explanation than a defensive one and still default to defensive when you’re actually in the room. Knowing is not the same as having practiced.

That gap — between knowing and having practiced — is what decision training is built to close.

How a decision-training scenario actually works

A scenario in Next Move™ is a branching simulation. The player is dropped into a moment — they’ve just been released, it’s 9:14 a.m., they’re standing outside the gate with a paper bag of belongings — and they’re given a decision: go to a friend’s couch (a known address but possibly a parole violation), go to a shelter (no known address, longer wait), or call the reentry agency that scheduled an intake for tomorrow.

Each choice routes to a different next moment with different rules in play. Each next moment has its own decisions. Some paths converge; some diverge; some lead to dead ends. The player can try a path, see what happens, back out, and try a different one.

Three things make this useful as training rather than as a game:

1. Each consequence is grounded in real-world rules. A scenario set in New York applies the actual supervision conditions in New York. A scenario about Medicaid enrollment uses the actual eligibility cliffs. We don’t make the rules feel real; we make them be real.
2. The scoring isn’t about reaching a single right answer. Many decisions have several acceptable paths and several catastrophic ones. The training is in seeing the difference, not in memorizing the script.
3. The replay is designed in, not bolted on. Every scenario can be played repeatedly, with different starting conditions, until the player has built intuition.

Why now

Two things changed in the last few years that make this work practical at scale.

The first is hardware: by 2025 the major correctional tablet vendors — Securus, ViaPath, Orijin — had deployed enough devices that you can actually reach people inside, and the major agencies had begun to permit non-vendor content. That removed the distribution problem.

The second is content modeling: branching scenarios are dramatically cheaper to build today than they were five years ago, and the cost of authoring a high-quality node graph has dropped to the point where regional and state-specific variants are economical. That removed the cost problem.

Neither of those changes is widely known outside the corrections-tech world. But together they mean that decision training, which used to be a thing you could do for a few hundred people in a workshop, is now something you can do for a few hundred thousand people on the devices they already use.

What we’re building first

Next Move™ is building decision-training scenarios for the most underserved audiences first, starting with reentry from incarceration. The first scenario module is live on correctional tablets in New York and is being expanded state by state. Veteran transition modules and a foster-care-aging-out module are next.

The thesis is simple: the people facing the highest-stakes transitions deserve the highest-fidelity preparation. We’re not the only company that should be working on this, and we won’t be the only one. But we’re trying to be the one that does it most carefully.