Randomisation is the technique that researchers use to try to prevent the bias in results that could potentially happen if a researcher could choose which treatment group to assign a volunteer to.
While there may be hope that one treatment is better than another, running a trial is only ethical if there is true uncertainty about which treatment arm is better. But just to be sure, it adds scientific rigour, and reduces the risk of bias to randomly assign volunteers who are theoretically similar (based on the eligibility criteria) to the different treatment groups. It helps save researchers from inadvertently influencing the results by subconsciously (or consciously) putting people into the groups they think they will help create the result that they want, thereby skewing the research.
In simple terms, the chance you will be randomised to one treatment arm or another is like the chance of a women falling pregnant with a boy or girl. There is basically the same chance of both options, and neither option is better than the other in theory (although you might think have a preference).