See the last bit in the second definition? Or just the first definition in its entirety? Having faith means you don't need proof. Scientific argument can't convince anyone subscribing to a science-less universe. You're basically telling them that they shouldn't believe because ... they shouldn't believe.
But a scientific argument doesn't need to convince the believer that there is no god. It just needs to keep their religious beliefs the hell away from the things that allow a modern society to advance. If they were trying to build a functioning airplane supported by the power of faith, you could do the physics calculations using Newton's laws to show them why they should trust science instead. If they still refuse to change their stance, the ensuing plane crash will do the job.
Whatever the case, at some point, science is winning that argument.
However, if they're using religion to support themselves emotionally, well, Isaac Newton would have supported that idea. He was extremely religious. He thought the beautiful laws of reality were the masterwork of a divine creator. Science and religion work fine together as long as they stick to their own roles. It's only when religious ideals intrude on scientific fields like particle physics or basic medical care that things get scary.