Forecasting is not possible although assessments based on similar past events can be made. It is a factor in the choice of technique. Forecast accuracy has to do with the closeness of a forecasting model’s predictions to the actual data.
The forecasting accuracy increases with the number of observation as well as the level of aggregation.
Forecast of customer demand aggregated over many customer are more accurate since their is a high probability that increased demand from one customer will be compensated by decreased demand from another customer during the same unit tie period.
The larger the number of items involved - while all other things being equal, the more accurate the forecast.
Because of the statistical law of large numbers, the size of forecasting errors and therefore the accuracy decreases as the number of items being forecast increases and vice versa.
Thus, the accuracy of phone calls arriving at a switching station during five minutes interval than the number of HP computers sold in a certain day.
It means that the larger amount of an item’s volume are forecasting, the lower the unexplained variance in comparison to the actual volume being forecast, thus lower the error.
Forecast accuracy : number of item factor
