succesful systems will rarely be simple IMO - otherwise everyone would do them and the advantage would disappear
IMO, the first thing with any system is to develop a concept, e.g. the 'why' a system might possibly work. Then you try to find a way of formalising this - i.e. gathering and processing whatever data you require. Then you go about back-testing it, e.g. football-data.co.uk
have a look around PP and you'll find lots of successful systems. If you have specific questions, ask. I may be able to give you the answer