Arguments to overcome obviousness rejection, particularly prior art combinations -

1. At least one claimed element is not “familiar”/known/shown in the art.
2. The claimed invention is not a combination of discrete elements at all.
3. The proposed combination of elements does not yield the claimed invention.
4. At least one claimed element functions differently from that in the prior art.
5. At least one claimed element was known only in substantially unrelated area.
6. It would take ingenuity, beyond one of "ordinary skill in the art," to combine elements. [Remembering that KSR renders one of ordinary skill as omnificent, as expert as any inventor.]
7. The number of alternative solutions for one claimed element was large or infinite. [This attempts to overcome a rejection on 'obvious to try.']
8. The result of the claimed combination would not have been predictable, or would not have had a reasonable expectation of success. [An unlikely sell without some evidence to point to.]
9. Secondary considerations: long-felt but unsolved need; commercial success; prior art expressions of doubt. Note that KSR reaffirmed Graham v. Deere.

Examiner guidelines -

1. Analysis must be made explicit; some “articulated reasoning”; note that KSR reaffirms Kahn.
2. Analysis must contain a rationale (“some rational underpinning”) and not be merely conclusory – during prosecution, note Rule 104(c)(2).
3. Facts must be supported with evidence – refer to Rule 104(d)(2); in the instance where an examiner makes an unsupported assertion of fact.
4. Analysis must be judged as of time of the invention, without application of hindsight bias.
5. KSR doesn’t change the standards for PCT application examination.

The lists above are taken from discussion at the NAPP convention, with added comment in brackets [].

Kindly thanks to Louis Hoffman.