What is the "excessively mathematical" way of valuing a loss?

I will not take the "excessively mathematical" approach criticized by the court in Rosvold v. Dunlop. Nor will I simply use a minimal lost projected future earnings as encouraged by the defendant. Neither is correct; as the court emphasized, what is being compensated is the loss or impairment of earning capacity as a capital asset. In this particular case, I can usefully consider projections from past and current earnings as a factor in valuing the loss, but only as one factor. I must also address the unknowable by considering real and substantial possibilities, and give them weight according to the percentage chance they will happen. Speculation must be avoided.

It is not disputed that the well-known factors approved again in Rosvold v. Dunlop apply to this plaintiff: 1. he has been rendered less capable overall from earning income from all types of employment; 2. he is less marketable or attractive as an employee to potential employers; 3. he has lost the ability to take advantage of all job opportunities which might otherwise have been open to him, had he not been injured; and 4. the plaintiff is less valuable to himself as a person capable of earning income in a competitive labour market.