Coad's Methodology ?

In Coad's methodolgy we also identify objects based upon classification i.e. actors, participants, places etc...this methodology can also result into making un-necessary classes; then how it is different than Abbots methodology? Which is better?

Coad methodology is step-wise process which involves identification of actors, participant, places, transactions, container objects and tangible things. There is no chance of identifying unnecessary classes because if you identify some unnecessary thing in one step. In next step, you would be able to identify that unnecessary thing and can remove it. For example, if you identify an actor which is not part of system, there would be no participant for that actor, so you can remove it. Both methodologies are equally good and their usage depends upon requirement/scenario.