I use systems (no surprise there) and make a good profit every year.
It can be done. Well researched selections with a high probability of winning are the best ways to make a profit in my opinion. The only issue is most peoples systems are not well researched or aimed at high probability selections.

For instance I can bet overlays that are $100 when they should be $50 but thats just a difference of 1% in the chance of winning. If my selection technique has an error rate great then 1% I am going to lose long term.
On the other hand if my selection is $4 and I classed it as a $2 chance (same as above double the money to what was estimated) then I have estimated a 50% chance of winning and the market has given me odds of 25%. Thats a 25% difference. My technique on ratings can now be 25% off before I would lose long term. Also my run of outs on a $4 selection overlay vs $100 overlay is a lot less and won't hit me pschologically.

Things to think about.