How to make a conflicting evidence decision before acting

A person standing at a fork between two softly lit corridors

To make a conflicting evidence decision before acting, name the disagreement, check each source's basis and freshness, and apply a decision rule set in advance. Record what would change your view so you can revisit the call when new evidence arrives.

What exactly conflicts in the evidence?

Start with the decision the evidence needs to inform. A forecast can look contradictory while each source answers a different question. One might describe orders received; another might estimate revenue expected to close. Before weighing either, write down the choice in front of you and the measure that matters to it.

Then restate each source's claim in comparable terms. Name the measure, unit, and time period. Mark what each source directly observed and what it inferred. If one source reports a past result and another predicts a future result, they are not direct contradictions. They may still matter to the same decision, but they need different treatment.

A disagreement about a recorded result is different from a disagreement about an assumption or an interpretation. Keep those apart. The distinction between a fact, an assumption, and a prediction helps clarify what each source can support; see how to distinguish facts, assumptions, and predictions.

How do you check each source and its freshness?

Check where each source came from and what it directly observed. A source that records completed sales supports a different claim from a forecast based on discussions with potential buyers. Neither is automatically more useful. The decision determines which evidence matters, and the source's basis shows what it can reasonably support.

Look for the date, the coverage, and the conditions behind the claim. Ask which customers, transactions, teams, or time periods are included. Check what has changed since the evidence was gathered. A result can be accurate for its original period and still be too old to guide a current choice.

Freshness is relative to the decision. A stable operating measure may remain useful longer than a measure affected by a recent change. When the source leaves out its date or coverage, record that gap instead of treating the claim as complete.

Keep a source's certainty separate from your confidence in it. Confident wording does not show how the evidence was collected, whether the sample is relevant, or whether the conditions still apply. Brainis maintains a live model of company state with sources, freshness, and confidence. You can read about company state in Brainis and what company state means.

If two sources still disagree, write down the specific difference you cannot resolve. That makes the remaining uncertainty visible rather than letting the stronger-sounding statement decide by default.

How do you make a conflicting evidence decision?

Set the decision rule before looking at which source supports which outcome. Otherwise, it is easy to move the threshold after seeing the evidence. State what level of support would justify acting, what would justify waiting, and what would make you choose another path.

Name the evidence that would be strong enough to support each option. For example, you might require a confirmed result before making a costly commitment, while allowing a reversible step on weaker evidence. The rule should match the cost of being wrong and how easily you can change course. That is a choice about how to act under uncertainty, not proof that one source is correct.

If the evidence points to different possible outcomes, preserve the range rather than forcing it into one certain number. Note the assumptions behind the range and what the prediction depends on. When reviewing a prediction, Brainis carries ranges and a model version, and preserves the prediction unchanged when reality arrives. Read about the world model and why ranges can be more useful than certainty.

Sometimes the evidence does not meet the rule for either path. Say so, then name the smallest useful check that could change the decision: which observation you need, who can obtain it, and when you will review it. If waiting carries a cost, include that in the decision too. You may choose a limited, reversible action while the check runs, but state what would stop or change it.

Two colleagues comparing options laid out side by side on a long table

What would change your view after the decision?

Record the choice alongside the evidence considered and the reason for choosing it. Include the source, the date, and the limits you noticed. Then write the result you expect and the observation that would challenge that expectation. This gives the team a basis for reviewing the call rather than reconstructing what people thought after the outcome is known.

Example: A fictional small company has two sales forecasts for the same coming month: one based on signed orders suggests a quieter period, while another includes several likely renewals and suggests a stronger one. The team chooses a limited hiring delay, records the evidence and the assumptions behind each forecast, and expects the signed orders to rise after renewal decisions. It will review the choice when those decisions arrive; if renewals do not move forward, that result challenges the stronger forecast and triggers a fresh decision.

Changing an old prediction after the result is known makes it hard to see what the team expected at decision time. Compare the expected outcome with what actually happened, then record what the difference teaches you. Brainis records predicted and actual outcomes in the Decision Ledger, where they improve the next recommendation. The memory page and guide to writing predictions down provide related reading.

What should you do before acting?

Use a short sequence before committing. First, write the decision and the competing claims in one place. Make sure each claim addresses the same measure and period, or state clearly why it does not.

Next, check the source, date, coverage, and confidence each claim supports. Separate observed results from assumptions and predictions. Apply the decision rule you set before weighing which outcome each source favours. If the evidence does not meet the rule, name the unresolved question and the check that could answer it.

Finally, record the choice, its reason, the expected result, and the evidence that would reopen it. Set a point to review the decision when that evidence is available. If you act while uncertainty remains, state what would cause you to pause, change course, or continue.

This process does not remove uncertainty. It makes the basis for acting clear and leaves a record you can test against what happens. If the decision needs to become assigned work, see missions in Brainis.