[simpleTruth]

IMO a reasonable test is: If a patent can be used to sue one company it may be patentable. If it can be used to sue 100 companies working on separate products it's probably obvious.

Basically, you can't prove 1 person did not read it. But, you can be positive 100 people did not read it. Now clearly this depends on the number of people working on the problem. So, while 100 is excessive in most fields, somewhere around 3-10 is probably a good lower bound.

[petegrif]

That is clearly not a reasonable test. In fact it is almost certainly a great test for someone having innovated in a valuable way. If you file a worthless patent then no-one will be interested in your innovation and there will be no-one to sue. If however you file something valuable you would expect to potentially have lots of people to sue. In your example the question arises, why have so many people trespassed on this patent? We know they have an incentive - it is valuable. But why in this hypothetical case did they trespass? Did they think the patent was breakable? the patent holder weak? the market opportunity so great that they would build now pay later? Perhaps they didn't find the patent? So many questions. But as a cursory study of hindsight bias will make plain it is extremely common for something that is not at all obvious at the time of its invention of appear so later.

http://en.wikipedia.org/wiki/Hindsight_bias