I am looking to get more serious in the sports betting world and I want to discuss some theory and some maths with you about middle betting.
I am still what you would consider a beginner in the betting world but I have a good understanding of gambling and odds in general as I played poker professionally for almost 2 years.
For the purpose of this example I will use a signal given by Oddstorm free service for middle alerts.
Game: Arsenal - Liverpool in Premier League
Under 4.5 @Bet365 for 1.175
Over 3.5 @Williamhill for 3
71.86% of our risk must go to the Bet365 28.14% of our risk must go to Williamhill
So basically if the event Amount of goals = [3 or less and 5 of more] happens, we lose 15.57% of our total bet
if the event Amount of goals = [4 goals] happens, we win 68.86% of our total bet
Now in order to calculate our expected value in this middle, we need to evaluate what % of the time this game will finish with 4 goals.
To do that I will use the odds given by Bet365 on Exact number of goals: 4 for this game.
Exact Total goals [4]: 5.00 @ Bet365 which means that Bet365 thinks that this game will finish with 4 goals about 20% of the time (actually it's less but just for the sake of the example we'll use this fugure).
So basically 80% of the time we lose 15.57% of our bet and 20% of the time we win 68.86% of our bet.
Expected value of this middle: 15.57% x 80% + 68.86% x 20% = 1.32%
So this middle has a 1.32% expected value over the long run. Am I correct?
What method could we use to get a sharper idea of the % of the time that the event: [4 goals] will occur?
Here is a screenshot of my excel calculation:

If I made any mistake in my reasoning or calculation, feel free to participate. Also, what are good middle alert service providers you could recommend?