Controls provide a reference that scientists doing an experiment can compare their results to.
For example, you want to know whether eating blue food dye makes people turn blue. So you do an experiment in which you get a group of people, feed them 1000 drops of blue food dye each day for a month.
But without having a control group of people who weren't fed blue food dye each day, you have no way of knowing whether people turned blue because of their diets, or for some other reason.
For example, maybe you did this experiment in Alaska in January and they all turned blue - not because of their diets, but because they were just so cold.