I went to lengths because gambling is what I do (sportsbetting & poker).

Sportz basically said how I approach it. Created a database in excel and use a formula to crunch the numbers and pump out the spreads. I will bet on the picks that don't seem like the 'sharp' side, simply because overall the picks have shown to be very profitable over a good sample.

And I don't aim for 60% at $1.95 sportz, 57.5% at $1.90 is what I'm hoping for (it's done better than this so far). I'd really be happy with 55% at $1.90 (enough to double the bankroll of an aggressive bettor).

It is pretty difficult to just create your own ratings/system. Look at things like home advantage, travel, form, injuries, motivation, matchups etc. You also have to consider things outside of the game itself like public opinion, who the money will come in on etc. All you have to do is be sharper than the bookies. To beat them you either need more info, or just be able to process the same info more effectively.

Look for things like an in form home dogs vs a popular public team.

Underdogs are generally a better bet than a fave.
Home dogs are usually the best bet (compared to away dogs, away faves and home faves).
Away faves are usually the worst option.

In NRL, underdogs have performed very well over the last few years.

In AFL, solid home faves and home dogs have performed well.

These aren't necessarily profitable trends, but this is the sort of stuff to look at.