Extrapolation

Extrapolation continues the pattern beyond the available data; it is less reliable because the real relationship may change outside the observed interval.

An estimate made outside the observed range is extrapolation. This is more dangerous because there is no observed evidence in that region: the relationship may weaken, level off, curve or stop altogether. The further the prediction lies beyond the data, the less confidence it usually deserves.

A prediction made outside the recorded range is extrapolation. Estimating a score for 12 hours would require extending the trend beyond the data. This is less reliable because the relationship might change outside the observed range: scores cannot rise without limit, for example.