Sports Betting Pre-game profile - soccer

Hello, this is an open question for ideas around how to predict over 1.5 goals market or higher looking at stats/form.

For example, looking at a particular league and watch how many games the home team finish with games over 1.5 goals. Let's say at least four out of five previous.
Then you do the same for the away team.

Now if you get 8 matches from both teams over 1.5 goals, at least four each or higher, then we could conclude a strong trend.
That is 80% or 8 points.
But still, I feel this is a week indication, even if the trend is present.

Here is two more parameters I am considering.

We look at the last meeting between both teams with the H2H function.
If there last meeting ended over 1.5 goals we add one point to 9 or 90% with my example above.
If the last meeting ended under 1.5 goals we reduce one point to 7 or 70% with my example above.

Then the last parameter is about the ranking for each team and not sure how to solve this into the estimation. Let's say we have the top four/five ranking teams into category one.
Let's call them the super good teams where one will probably win the whole league.

Now it both teams are members in categorie one, then all is fine.

Let's get lower into the ranking system for the league and pick the four/five second-best teams.
How would we estimate them if they meet a team at ranking category one?
Depending on the pre-game profile and percentage or points we could consider finding strong underdogs and still predict over 1.5 goals.

Around 78% of the bigger leagues end over 1.5 goals.

But when we reach the ranking category at level three, four and five, average performance, weak performance, and poor performance. Do we at this level predict under 1.5 goals market depending on pre-game profiling and stats?

Do we only use the team's top two or three categories when we predict games to end over 1.5 goals?
And when we reach below category two, three we considering finding strong underdogs or aim for under the 1.5 goal market depending on pre-game profiling.

Cheers