I'm not sure how you could tell, visually, as common high-current contact platings are microinches of various silver alloys; silver-cadmium-oxide (becoming less common thanks to European RoHS requirements), and silver-tin-oxide are common and would appear no different than zinc plated bolts (which used to be cad plated).
I was only responding to your comment about welded contacts which are typically caused by excessive switched current due to undersizing/improper application of the relay, or a failure in the circuit being switched....or, they just plain wear out! (Loss of plating with wear eventually results in excessive resistance which causes excessive heating leading to failure of the contacts).
I am not aware of how coil winding failures would induce contact welding, nor do I have failure mode/rate data to offer a factual basis to establish whether coil or contacts are the most prevalent relay failure mode. My seasoned opinion based on my personal experience as an EE is that moving mechanical parts, such as contacts, are much less reliable that coils of wire in transformers, inductors, and relay coils which have no moving parts.