Happy ending problem (polygons)

269 Views Asked by At

I finished reading a book about Paul Erdös "The Man Who Loved Only Numbers" and I came to couple of questions which I need to check if I am understanding them well or not, one of them about what has become known as Happy Ending Problem.

What I understand that according to the formula

$$ 2^{n-2} + 1 $$

you can know the least numbers of vertices to be able to construct n sided-polygon. If that is right then e.g. let's take $$n = 5 \to 2^{5-2}+1 = 9$$ which means 9 vertices needed to guarantee a pentagon (5 sided-polygon). However, if you have even less vertices you are still able to construct a pentagon (it is not mentioned if it should be regular or irregular) so according to the illustration below we need just as much vertices as the needed-polygon requires.

enter image description here

Please correct me if I misunderstood the problem and feel free to post any link that may clarify the subject and add more details.

1

There are 1 best solutions below

1
On

You should carefully read the Wikipedia entry you linked to.

Note that a pentagon here is a closed simple, i.e., not selfintersecting, circuit consisting of five segments and five vertices. It is neither assumed regular nor even convex.

It is a theorem that among $\geq9$ different points in the plane, assumed in general position, i.e., no three in a line, you can always select $5$ that form the vertices of a convex pentagon. On the other hand the devil can place $8$ points in the plane such that no convex pentagon can be selected from them. This is the case for the $8$ points in your figure. It follows that $9$ is the "Ramsey number" for this little geometric problem.

Erdös and Szekeres 1935 conjectured the following generalization: Given $\geq2^{n-2}+1$ points in the plane, assumed in general position, you can always select $n$ of them forming the vertices of a convex $n$-gon. In its full generality this conjecture is still open.