My Sky and Telegraph teams will continue through the season to be selected using The Algorithm and various analytical evidence.
Following my original draft of the Sky team, I found out that Rudiger was injured meaning the process needed to be re-worked. I’m also slightly relieved as Chelsea have a new manager so am not sure if their style will change (eg frequency of passing bonus).
Before starting with selecting the team, I hear a lot of chatter about Guilbert of Villa priced at £4.5m (being a good enabler). Based on Villa’s defensive rating (albeit based on last season) and their opponents in the first four games, the algorithm forecasts defenders to score 7.65 points (excluding bonus, goals etc). This on its own makes him a reject. However, he is meant to have a lot of tackles and gets forward a lot. If I assume his attacking threat 90% of Doherty and bonus as 90% of Wan-Bissaka the project increases to 15.76. Though, this coverts to a very good 3.5 pts per £m over the first four Gameweeks. This is good but not enough to move money away from defence and comes with risk. As such won’t be included at start of the season (though may select him in my second team).
To select my team, I have gone through the same process as my initial draft, eg. Setting the prediction model for weeks 1 to 4 and selecting the best performing eleven players. Obviously, this is over budget, so then swap out the players that offer the least value for money with cheaper higher value players. I skip players that I have doubts over, Pogba (due to transfer issues), Rudiger (maybe injured), etc. In my initial draft, I ignored players like Stones and Matip due to doubts over playing time though following pre-season am more willing to select them. Based on the data I have selected:
There are risks with game time with several of these players, mainly Stones and Matip. I see Liverpool’s team sheet on Friday and so will need a contingency for if Matip doesn’t start. Probably Salah to Sterling and Matip to Robertson. I’ll be listening to Pep’s press conference for any clues on Stones and Laporte. Assuming no press conference news or injuries suggesting any absences, then this team will not change.
My strategy is to save as many transfers as possible for later in the season and as such have no plans to make any transfers before the over-haul. The above covers all captaincy days.
This season I’m going to monitor closely the performance of the algorithm predictions. Of course, the assumption is a player starts the match, if he comes on as sub, then I will halve the prediction and if he doesn’t play then the prediction will be set to zero. For the first week the prediction is:
There are no changes from my initial draft when going through the sane analysis. However, I have been alerted to concerns over Gross minutes so have swapped him out for Dendoncker:
There are two red flags, though both are expected back for the weekend.
As with Sky, I’ll be monitoring the algorithm through the season. The predictions for this week are:
That’s all my main teams in the various games. It’s just a case of listening for any injury news and follow press conferences in case somebody doesn’t play. My main hope is that my players play and no forced transfers are needed going into week 2. Can’t wait until kick-off. Good luck all for the season ahead.
Join Carl in our community TFF and Sky Slack Channels