- The forecast error is calculated by a number of universally accepted methods.
- After going over them, we will question if these methods are effective.
The question of how can forecast error be calculated is a common question in forecasting.
See our references for this article and related articles at this link.
How can forecast error be calculated is a common question, and it is done by applying any one of the standard forecast error calculation methods.
The calculation of these methods is widely known, but not as well understood as generally thought.
This article is written from the vantage point of having worked on forecasting projects for over a decade and a half, and explaining forecast error calculation and measurement to many people of different levels in organizations as well as people in many different departments. I will describe the reality of forecast error calculation.
How Can Calculate Forecast Error be Calculated
The most commonly used forecast error calculation methods are listed below.
We have not conducted a poll to determine the forecast error calculation method frequency, but having worked in the field for a while, this is our rough estimate of the relative frequency of use from most to least used.
To see an explanation of each, just click the link.
- MAPE: Mean Absolute Percentage Error
- MAD: Mean Absolute Deviation
- MAE: Mean Absolute Error
- RMSE: Root Mean Square Error
- MASE: Mean Absolute Scaled Error
- sMAPE: Symmetrical MAPE
Something to notice is that as one goes down the list, the lesser-used forecast error calculations are lesser used. A significant issue is forecast error measurements that are not proportional. Forecast error calculation methods that are not proportional are unintuitive and, hence difficult to understand.
When the topic of forecast error is discussed in most cases, the topic does not move beyond the forecast error of the item, generally at a location, what we call the product location combination.
In many cases, forecasts are created at a much more aggregated level, such as with sales forecasting. However, when the forecast error is created at the product location, a significant element of the forecast error measurement is how the error is reported outside of the individuals actually performing the forecasting. And this gets into the topic of both forecast error reporting and aggregation.
The topic of forecast error tends to focus overwhelmingly on the forecast error measurement, when in fact, this is only one dimension of the forecast error measurement as we cover in the article How is Forecast Error Measured in All of the Dimensions in Reality?
A Better Approach
Observing ineffective forecast error measurements at so many companies, we developed the Brightwork Explorer to, in part, have a purpose-built application that can measure any forecast. The application has a very straightforward file format where your company’s data can be entered, and the forecast error calculation is exceptionally straightforward. Any forecast can be measured against the baseline statistical forecast — and then the product location combinations can be sorted to show which product locations lost or gain forecast error from other forecasts.
This is the fastest and most accurate way of measuring multiple forecasts that we have seen.