Average : 72%


Remarks :

** I am very pleased with people explaining more their reasoning.

** A lot of people say "x>(1/epsilon+1)^2>1/epsilon^2 so take R=1/epsilon^2", taking a R too small might not give us the right thing, otherwise we'd just take delta=1<1/epsilon^2 for epsilon small enough


** Some people still divide by 0 (in the last question)


** People are starting to make lots of mistakes in calculations, I did not take away points this time but far too many +'es became -'es and others in that style.


** There was a small confusion, when you asked to explain without delta-epsilon, I still expected *some* maths, I encountered drawings of the graph, which is too little.



Again, all around it was pretty good.