In key decision T 36/82 (OJ 1983, 269), the board stated that inventive step was not considered to be constituted by efforts directed at the concurrent optimisation of two parameters of a particular device by the simultaneous solution of two equations which were known per se and respectively expressed those parameters as functions of certain dimensions of the device. The fact that it had proved possible to find a range of values for the dimensions in question which provided an acceptable compromise between the two parameters could not be considered surprising where there were indications in the prior art suggesting that favourable results might be obtained by the method of calculation applied.

In T 263/86 the invention related to a spectacle lens with an astigmatic effect. The board pointed out that the relationship between residual astigmatism, focussing error and frequency response could be assumed to be known by a spectacles expert. The board therefore saw the quality formula as merely the result of simultaneous optimisation of a number of lens properties which led to a compromise lying within the skilled person's discretion. However, such compromises in the case of a parameter optimisation were not deemed to be surprising and their discovery was thus not considered to involve an inventive step.

In a number of other decisions, all of which referred to T 36/82 (OJ 1983, 269) the subject-matter was found not to involve an inventive step, particularly when the problem addressed was to find a suitable compromise between different parameters (T 38/87, T 54/87, T 655/93, T 118/94). In T 410/87 the board stated that it was part of the activities deemed normal for the skilled person to optimise a physical dimension in such a way as to reach an acceptable compromise, serving the intended purpose, between two effects which were contingent in opposing ways on this dimension (see also T 409/90, OJ 1993, 40; T 660/91; T 218/96; T 395/96; T 660/00).

In T 73/85 the board stated that the very fact that the problem of improving the property in question was solved not – as was normal – by means of a specific change in structural parameters, but by amending process parameters, had in fact to be considered surprising. In this case it did not matter that the individual reaction conditions claimed in the disputed patent were known per se; more important was whether the skilled person, in expectation of the sought-after optimisation had suggested, or – in the absence of possible predictions – had tried as a matter of priority, the combination of measures known per se claimed.

In T 500/89 the board established that the fact that individual parameter areas taken per se were known did not imply that it was obvious to combine them specifically to solve the problem according to the contested patent. The combination of the individual parameter areas was not the result of merely routine optimisation of the process according to document 1, as there was nothing in said document to suggest this combination.