prediction

Prediction operates as a forecasting practice by constructing claims from evidence. Remove evidence from the picture and prediction stops operating as a forecasting practice — it has nothing to build claims upon. The removal test passes: prediction needs evidence running to function.

Prediction is a specific kind of inference — one that infers future states from current evidence and patterns. Per Law 9, a competent speaker would call prediction 'a form of inference.' The filer leans on this sense.

Causality (epoch 0.01) predates prediction (epoch 0.76). The concept of prediction requires causality — understanding cause-effect relationships is what prediction builds on. Which came first test: causality existed first and fed into the development of prediction.

Probability theory existed first and provided the mathematical foundation for prediction intervals.

Optimism bias is the tendency to overestimate positive outcomes of future events. Removal test: remove prediction and optimism bias ceases to operate — you cannot have an optimism bias about something unless you are making a prediction about it. Object-level dependency.

Prophecy IS a specific kind of prediction: a claim about future events asserted to originate from divine or supernatural insight rather than ordinary inference. The 'a Y' test passes — a competent speaker calls prophecy 'a kind of prediction'. Epoch note: prophecy predates prediction as a general category (0.15 vs 0.73), confirming prophecy is the specific form that later enabled the general concept.

Supervised learning is built or maintained for the sake of prediction — its designed purpose is to learn mappings from inputs to outputs that enable accurate predictions on new data.

A probability forecast IS a specific kind of prediction: it predicts outcomes but expresses uncertainty as a probability distribution over possible values rather than a point estimate. A competent speaker would call a probability forecast 'a type of prediction'.

TESTED DEPENDS_ON: Remove predictions — residuals cease to exist. A residual IS the difference between observed and predicted values; without predictions there is no residual. The removal test passes: X stops OPERATING without Y.

Surprise is the psychological response when an outcome deviates from a prediction. Remove prediction and surprise ceases to operate — there is no deviation from expectation to register. This is object-level: surprise's mechanism involves comparing outcomes to expectations, not merely needing the word 'surprise' to be said.