Premier League Surprises - where they are?

All my models are based on past data... and based on closing odds..
As I described before - I use a bot that place bets 1 minute before kick-off if it fit my model, so my plays are very-very close to "closing" odds..

According to the last eight years data - it was very profitable to play big home underdogs in EPL, but season 2018-19 shows a 0-9 run from the start of the season...
I was lucky enough to start my plays after the season start and own a 0-2 record, but still it's disappointing..
The favorites are 8-1-0 (there even no enough draws... for a plays with a hope for a win. Is the differences between the teams so big? The scores shows a big YES.
Liverpool, Man.City and Chelsea started the season in storm and play even the very weak opponents in full squads.. My model shows an average of 20 matches per season that good for a play, but after 10 rounds (out of 38) already 9 matches fit the strategy... and it's one more sign for a bigger gap between the teams...
Having a lot of a strategies in my portfolio - I will continue with those underdog plays, but (in this 3 leaders season) I'm very skeptic - it will have a success in this season.

Here is the data of the last 8 seasons (playing home dogs with odds around 7):

"Liiga" - Finland Hockey League trends

In one of my posts (here or on twitter) I mentioned that currently working on Finland Hockey League model.. The work is now over and I found nothing interesting for me... but..
European hockey lines are including the possibility to bet on "Draw" after 60 minutes and not a lazy bettor can quit easily find draw trends (with a nice profit).

The average odds for a "Draw" in Finland are 4.46 (in last 8 seasons) and betting them blindly shows + 34 units (poor ROI of 1% out of 3446 selections)...

One filter can change a lot... You can bet only on "close matches" (where the difference between the teams is up to 10%) and You already getting only 890 selections and 89 units profit (nice ROI of 10%).

 I'm personally don't want it in my portfolio, because it not shows profit every year (or at least 7 out of 8 last checked years), but those who want should know that in last 4 seasons this easy to make system made a ROI of 31%

What's next?
I will take it easy and will work on more soccer leagues to add to my model.. (Right now 10 in my list:
Argentina, Brazil, Japan, USA, S.Korea, Norway, Sweden (both), Finland, Ireland.
Those leagues are close to finish their seasons and I'll be ready for the next season..).

WNBA, NRL, AFL and MLB are also on my wish list..

See Ya:)