Chelation is an entropy driven reaction, i.e. complexes with polydentate ligands are more stable.
Consider what happens when you dissolve copper sulfate in water:
\begin{align}
\ce{CuSO4 + 6H2O <=> [Cu(H2O)]^2+ + SO4^2- (aq)}\\
\end{align}
This complex forms and with it comes some kind of ordering of the water molecules around it. Dissolution is often favoured because of the solvation energy. Now consider a bidentate ligand like ethylene diamine, $\ce{H2N-(CH2)2-NH2}$, short $\ce{en}$. Adding this ligand to a copper sulfate solution will result in a ligand exchange:
\begin{align}
\ce{[Cu(H2O)]^2+ + 3en <=> [Cu(en)3]^2+ + 6H2O}\\
\end{align}
In this reaction you increase the number of particles from four on the left side to seven on the right side. Taking into account that you also disturb the order of the surround water molecules, this reaction is very much favoured by entropy. In an attempt of zeroth order approximation, one could assume that entropy roughly doubles. Since the Gibbs energy is inversely correlated to the entropy, this means, that the reaction will be spontaneous. This is usually referred to as the chelate effect.$%edit$