6, and the chances of rain is 0.4. Historical data may be used for forecasting, but past patterns will not necessarily be reliable. To gain a more reliable assessment there are alternatives; the weather service makes advanced forecasts based on a large amount of data. However, to gain this information the business will have to pay a fee. This means that there is a cost associated with gaining the more accurate data. It may also be assumed that if data to be used in the decision is being gained from an external source, there may also be a time delay. If a fast decision is needed, the decision may need to be based only in the information at hand, trading off the potential accuracy to gain the speed. If the decision is not needed quickly, then the extra time may be well spent; accuracy will be gain at the cost of speed. If the forecast from the weather service is that there is only a 30% chance of rain over that weekend, this can be used in the probably calculation, showing a 0.7 probability that it will be fine, and a 0.3 probability it will rain.

Assessing the accurate probability that there will be competition is more difficult. The trader knows that only one competitor has been invited. It may also be known and that in the past some years they have attended the festival, but in other years they have attended an alternate festival; five years at each. As there is no approach that is likely to be accurate, so the most viable approach is to use the historical data, giving 0.5 probability of each potential outcome.

This may then be used to assess the probability of each outcome.

Table 1; Probability table

F/R probability

C/N probability

Overall probability

FN

0.7

0.5

0.35

FC

0.7

0.5

0.35

RN

0.3

0.5

However the business will also want to assess the potential financial impact. For this historical data and estimates may be used. If it is assumed that the stall will cost $5,000, and if it is fine net revenues on the goods will be $20,000, but if it is raining the net revenues will only be $8,000. If competition is present the overall level of net revenue will be halved with the trade shared between two traders.
Table 2; Revenues for each outcome

Net revenue

Less fee for stall

Net profit/loss

FN

20,000

5,000

15,000

FC

10,000

5,000

5,000

RN

8,000

5,000

3,000

RC

4,000

5,000

-1,000

From this it is apparent that there is only one outcome where there is a loss; that is where it rains and where the competition is present. This means there is only a 0.15 probability of the firm making a loss and a 0.85 probability that there will be a profit. Therefore, in this case it is recommended that the business takes the stall, as long as there is not a better opportunity available.