Module 4 – Lesson 4: Risky Predictions

A strong scientific explanation does more than fit what we already know. It makes predictions that could be wrong. The more specific the prediction, the greater the risk. This lesson explains why risky predictions are a key feature of severe testing and why Design Biology treats them as a standard requirement for serious claims.

A risky prediction is an expectation that follows from a claim and would be unlikely under competing explanations. If the prediction fails, the claim is weakened. If the prediction succeeds, the claim gains credibility. Risky predictions are valuable because they limit interpretation. They reduce the freedom to adjust a story after the fact.

Many explanations in biology and origins research are flexible. They can accommodate many outcomes. If the result is A, the story explains A. If the outcome is B, the story explains B. That kind of explanation may sound persuasive, but it is not severe. A severe approach demands predictions that do not automatically fit every possible result.

Risky predictions have three qualities.

First, they are specific. They identify what should be observed, under what conditions, and what measurable outcomes to expect.

Second, they are exclusive. They distinguish one explanation from others by predicting something the alternatives would not strongly expect.

Third, they are testable. They can be checked through observation or experiment, and failure is meaningful.

Design Biology applies this principle to questions about information, control, integration, and robustness. If a system truly depends on coded information and regulated control, then we should be able to predict where those features must appear and what patterns they should produce.

For example, if a biological system is highly integrated, we can predict that disruption at specific control points will cause system-wide effects rather than isolated changes. If a system uses error correction, we can expect that removing the repair will increase certain types of failure. If regulation is real, we can predict measurable differences in outcomes when regulatory signals are altered.

Risky predictions also help avoid confirmation bias. Instead of selecting evidence that already fits, you define expectations ahead of time and then test whether nature matches them. This makes evaluation cleaner and more disciplined.

In Design Biology audits, risky predictions work like a filter. They separate claims that only describe patterns from claims that explain mechanisms. A claim that explains mechanisms should be able to forecast what will be found when you examine systems more closely.

A helpful habit is to write predictions in plain language. Not in vague terms like “we expect complexity.” Write what should be found and what would count against the claim. If you cannot state what would count against the claim, then it is not a risky prediction. It is a flexible narrative.

Risky predictions also force precision about alternative explanations. You must ask what they would predict as well. If all explanations predict the same thing, then the observation does not discriminate between them. But if one explanation predicts a distinct signature, then testing becomes meaningful.

When you write your audits, include a risky predictions section with three items.

What the claim predicts.
What a null or alternative would predict.
What observation would count as a failure?

This turns evaluation into a disciplined process instead of a debate.

In the next lesson, we will address the final question severe testing demands: what would falsify a claim? Once you can state falsifiers, you can design tests that are genuinely severe rather than merely supportive.

Lesson Summary

A strong scientific explanation goes beyond fitting existing knowledge by making risky predictions that could potentially be proven wrong. These predictions are key to severe testing and are standard in Design Biology for evaluating serious claims.

Definition and Importance of Risky Predictions:

• A risky prediction is an expectation derived from a claim that would be unlikely under competing explanations.
• If the prediction fails, the claim weakens; if it succeeds, the claim gains credibility.
• They limit interpretation by reducing post hoc adjustments to fit any result.
• Risky predictions enforce specificity and exclusivity, distinguishing one explanation from others.

Qualities of Risky Predictions:

• Specific: Clearly identify what should be observed, under what conditions, and the expected measurable outcomes.
• Exclusive: Predict outcomes that alternatives would not strongly expect, thereby differentiating contenders.
• Testable: Subject to observation or experiment; failure provides meaningful evidence against the claim.

Application in Design Biology:

• Focuses on information, control, integration, and robustness in biological systems.
• Predicts where coded information and regulated control must appear, and what patterns they should create.
• Examples include expecting system-wide effects from disruption at control points, increased failures after removing error corrections, and measurable changes from altered regulatory signals.

Benefits and Best Practices:

• Avoids confirmation bias by setting expectations before testing, leading to cleaner, more disciplined evaluation.
• Helps separate claims that merely describe patterns from those explaining mechanisms capable of forecasting results under closer examination.
• Encourages writing predictions plainly and precisely, including what would count against the claim.
• Requires considering alternative hypotheses and what they predict, enabling meaningful discrimination between explanations.
• When auditing, include three key prediction elements:
  • What the claim predicts
  • What the null or alternative hypothesis predicts
  • What observation would constitute failure

The next lesson builds on this by exploring how to state falsifiers of claims, allowing for genuinely severe tests rather than tests that merely support hypotheses.