Predictions have limits — Data, 11–13 years
Past data can guide a prediction, but it cannot promise what will happen next. Data, 11–13 years.
Use patterns carefully
A prediction uses a pattern in known data to make a sensible guess about an unknown value or future event. It works best when the new case is like the cases used to build the pattern. A prediction is not a promise: new conditions or random events can break the pattern.
Why predictions can fail
Data describes what happened under particular conditions, not every possible future. A model may miss a new factor, such as a heatwave, a broken machine or a sudden change in behaviour. Knowing the limits matters because a neat calculation can still be based on an incomplete picture.
Predicting a queue
A café serves 24 people in 2 hours, so its recent rate is 12 people per hour. If the next hour is equally busy, predicting about 12 customers is reasonable. But a school trip arriving, rain keeping people indoors or a broken till could change the result. The number is a forecast from the old conditions, not a guarantee.
The line must continue
When points on a graph rise neatly, it is tempting to extend the line far into the future. That feels reasonable because the next few values may follow the same direction. Far from the data, however, the prediction depends on assumptions we cannot check. A nearby estimate is usually safer than a distant one.
Where it helps
Weather forecasts, delivery times, energy planning and sports tactics all use past data to prepare for what may happen. People compare a forecast with its uncertainty and the conditions behind it. Checking those limits helps them plan without treating a prediction as certain.
Keep exploring
Other languages
Loading MyLeoNes™…