We’re delighted to reveal that the first set of Pre-Season Projections are now available to our Members. These can be found under the Projections section in the Members Area – we’ve supplied a season-long table, split by each position and, for a short-term view, have also provided a table which plots the Projections for just the opening six Gameweeks.
The Player Projections are, in many ways, our player predictions, offering a view of “projected” goals, assists, clean sheets etc for the season ahead. We also include Fantasy Premier League (FPL) points and bonus in the Projections, although no matter what Fantasy game you play, these will hopefully provide an invaluable guide.
We will be updating these Projections frequently: there will doubtless be tweaks required, particularly to the playing probabilities – in some cases almost immediately as the new signings alter our thinking – but we’ve decided to present our first set of data.
In case you missed our previous article, this season’s Projections are being created using the statistical model behind the Rate My Team website. This site was run by Chris Atkinson for the past two seasons, during which he achieved some superb results with the model. Chris is a statistician who specifically built his model to simulate and predict the returns of players in Fantasy Football and, particularly, the FPL. Having brought Chris on board to help us with our Projections, we’re pretty delighted with the end result and the power and flexibility of the model behind them.
The Projections are, of course, part of our Members offering. For more details on the benefits of Membership and how to join, click here.
Here’s a short FAQ with more details….
How do I use the Projections?
Crucially, these tables can be used as another opinion-former when comparing players. Every player is rated for goals, assists, clean sheets etc based on values created by the “Rate My Team” statistical model – the values returned for each player are the expected values in accordance with its calculations.
A good way of thinking about them is to consider them average values for each player. If the season was played out 500 times, the values you’ll find in these projection tables are the average that the model would expect the player to score.
One of the most useful aspects of the Projections is the value column. You can sort by this to find the best value (Points Per FPL Million) delivered by players in each position.
With tables for the season and the opening six Gameweeks, you can get a view on long-term and short-term returns and value.
What are these values based on?
The “Rate My Team” statistical model is a complex beast – it uses historical data stretching back a number of seasons and, of course, specific player and team data. When generating the player values the model looks at the specific fixture in each Gameweek, assessing the strength of both teams before calculating the likely team goals scored and how these goals will translate to specific player stats. The probability of a player taking part in the match is also a key ingredient. Factors such as rotation, injuries, suspensions and absence due to the African Cup of Nations are also simulated by the model.
Why do the stats show decimal places?
The values returned by the model are not exact values – they are not figures that state what a player will score for a particular stat but a value which indicates their expected return. As such, the model can be more accurate when producing data with decimal places.
Why has Van Persie got fewer goals etc?
There’s no denying Robin Van Persie enjoyed an incredible season last term. For starters, he avoided injury and was almost ever present – the first time he has achieved or even came close to this in his entire Premier League career. Secondly, Van Persie was, without doubt, the key attacking force in the Arsenal lineup – the summer acquisitions of Olivier Giroud and Lukas Podolski suggest this may not be the case any longer. Added to this, of course, there is uncertainty surrounding the Dutchman’s future at the club.
Why are there fewer goals across the board?
The model used historical data to create an average goals per game value for the season. Although we’ve seen a trend that has seen the goal count soar in recent campaigns, the model has been more cautious and is currently expecting less goals for teams, with fewer big scorelines and teams that are likely to leak large amounts of goals. In a nutshell, the model doesn’t think that there will be a Blackpool or Wolves around as whipping boys next season.
When will these Projections be updated?
We’ll most likely update these every day up until the season kick-off. As we get more data on likely first choice lineups and as the transfer market moves to cast doubt over starters, we will adapt and tweak the model accordingly. It’s highly likely that, over the space of a few days, you’ll notice significant movement within the data for certain players.
There are still lots of issues to be resolved that will have a marked affect on the Projections – notably the striker situations at Arsenal and Spurs.
We’ll keep you posted on any major changes and movements in the Projections as and when they are made. Full Projection analysis articles for members will be following next week.
