An exciting announcement...

Fantasy Gameweek to work with Impect as data partner

Hello and welcome to the latest edition of Fantasy Gameweek. Today we want to share exciting news with you about the next stage of our growth.

Since we launched Fantasy Gameweek last summer we have been blown away by the interest from subscribers. There are more than 11 million people playing Fantasy Premier League this season and it is clear that interest is only growing, not just among new players, but existing players.

Our ambition is to help you win your mini-league, or just be more competitive if that is all you want. To do this we have combined 30 years of experience in finance, journalism, sport and fantasy sports, which we believe gives us a unique insight into how to succeed at FPL, a game that has many similarities to playing the financial markets.

We have been sending two editions to subscribers every week - one looking at what stood out from the latest Premier League matches - Tom’s “what we learned” column - and another focused on ranking the players and looking for who offers the best value - Graham’s FGPE rankings.

You will continue to get two editions a week during the rest of this season. However, these are going to change thanks to an exciting and significant improvement we want to share with you.

We are delighted to say that we will be working with Impect as our data partner, using their analysis to explain and understand what is actually happening in the Premier League and give you an edge in Fantasy Premier League.

We have been studying the football data industry looking for an ideal data partner. Impect was founded in 2015 by former professional footballers Stefan Reinartz and Jens Hegeler, who felt that the existing data available in the sport was unable to properly evaluate the performance of a player or team. What they came up with takes football data to the next level, beyond the rudimentary indicators like xG that are being widely used. At the heart of Impect is a concept called Packing - a collection of key performance indicators that show how effective a player is at increasing the chances of scoring a goal or stopping a goal from being conceded. This data includes tracking how many opposition players a player takes out of the game with a pass or dribble, and how effective a forward is at finding space infront of goal. Impect also produces the standard data like xG and when you combine this with its unique data you get a far more rounded understanding of what is actually going on in matches and who is truly influencing them.

Data is a vital and rapidly expanding part of analysing football and has been part of the Fantasy Premier League landscape for some time. However, we believe that some of the basic stats - such as xG (expected goals, which judges the quality of a chance for a player) - are being overused and presented by analysts as showing more than they actually do.

xG is a useful stat and has been pivotal in taking data in football mainstream, as explained in Rory Smith’s superb book Expected Goals: The Story of How Data Conquered Football and Changed the Game Forever. But it has significant limitations, ones that often aren’t recognised when it is presented as some sort of gold standard to praise, criticise or rank players and teams.

Firstly, it is subjective. A whole load of judgements are needed to be made by the data provider to get an xG number for a single chance. This is why the leading data providers produce different xG numbers for each chance and match, often differing so widely that they don’t agree on which team had the highest xG in a single game. Given the variation for a single chance, imagine what the variation could be for a whole season.

Secondly, xG only measures the quality of the chance and takes no account of the player who has it or the goalkeeper they are facing.

Thirdly, most xG measures only track shots, so opportunities that are ruined by a poor final pass do not register.

Fourthly, and obviously, it is backward-looking. Just because a player or team registered a certain xG in their last game, what does that mean for this game?

Bill James, the baseball statistician and godfather of the Moneyball system popularised by Michael Lewis’s 2003 book on the Oakland Athletics, has criticised the oversimplification of data in sport, particularly where a collection of subjective judgements are put together to give a single data point. “The world is vastly more complicated than the human mind,” he has said.

This is why xG is only truly powerful when used alongside other data and analysis. Wouldn’t you like to understand why a player’s actual goals and FPL points are under or overperforming their expected goals? Or how a player’s xG in one game may translate to the next game? Or whether a player’s xG is connected to their own ability or simply the role they play for their team, ie as a forward in a team that creates alot of chances? xG shows the past opportunities for a player, but that is only part of the story. What about their talent? What about how their opportunities may change in the next match? Understanding this is truly powerful for FPL players.

Working with Impect’s data, we will help you get a more rounded understanding of what is going on in the Premier League and FPL - and therefore what may happen next. We will make it interesting and fun, identifying what are signals and what is just noise in an easy-to-understand way. Howard Marks, the renowned American investor, used this quote from Albert Einstein in a recent memo to his clients on the perils of trying to understand what is important and what is not: “Not everything that counts can be counted, and not everything that can be counted counts.” Working with Impect, we will help you understand what counts.

We will be using Impect’s data in all our newsletters, and, we are excited to say, a new podcast which will be coming soon.

We can’t wait to share these improvements with you.

Best

Graham and Tom