Probably Weekly

Welcome to Probably Weekly, where we tackle key questions about probabilistic modeling, risk reasoning, or general planning under uncertainty. Whether you're analyzing simulations, experimenting with models, or just wondering how to reason when things get muddy — this is your corner of the internet.

Each post aims to clarify complex ideas, highlight useful techniques, and showcase how probabilistic methods — can support better decisions in uncertain environments. Every decision in a new market is a wildcat well — PLANS helps you drill smart.

Got a question? Send it my way to cassie.garza@agr-software.com and we will (probably) feature it next!

What can you model when you have no historical offset data?

When most teams face a decision with no historical data, the default instinct is to delay, guess, or mimic someone else’s playbook. But in many high-stakes environments — waiting for data just isn’t an option.

So… what can you do when you're drilling into uncharted territory?

Probabilistic modeling gives you a way to formalize uncertainty. Even if you don’t have past data, you likely have:

• Expert intuition
• Reasonable ranges or constraints
• Similar-but-not-identical case studies
• Business logic or structural assumptions

These can all be encoded into a model as starting points that evolve as you learn.

With a tool like PLANS, you can:

• Define your parameters explicitly
• Simulate a range of possible outcomes
• Quantify the uncertainty around each decision
• Recalculate the model as soon as new trends appear

PLANS gives you clarity even without historical data.