Φ-value analysis has proved to be a powerful technique in protein folding studies. It has been used to study the transition state structures, to infer microscopic folding pathways, and to identify key residues in the protein folding process. However, its effectiveness in protein binding reaction has not been tested, especially when a fly-casting mechanism is involved. In this article we attempt to answer this question through a coarse-grained study of the binding reaction of Arc repressor dimer. Our simulations show that its binding process proceeds through a fly-casting mechanism, consistent with previous results. We then estimate the importance of the residues for the fly-casting binding by computationally mutating each of them into glycine and calculating their respective effects on the capture radius. It is found that (1) the residues with high Φ values may not be responsible for the large capture radius in the fly-casting binding; (2) mutation of residues with low Φ values may destabilize the denatured states and then increase the capture radius, presumably further increasing the binding rate. Based on our simulations, we conclude that Φ values do not reflect well the importance of residues in the fly-casting binding reaction and need to be combined with other techniques to provide a complete picture.