ABSTRACT
For on looking around at previous predictions by geographers, failure is the most likely outcome (Haggett 1994). This can occur in two ways. Underprediction or failure to predict the scale and nature of change (what the statistician calls Type I errors) and overprediction or the prediction of events that failed to occur (Type II errors). An earlier review of geographical predictions (Haggett 1972) suggested Type II errors were rare, implying that our past attempts to look forward have been too cautious.
