Fantasy Premier League’s popularity has increased dramatically over the past 10 year. This has led to an increase in the number of resources that can be used to assist us in making smart decisions when playing the games. There are different types of managers — some rely on the eye test, some use stats, and there are others out there who just like to play on pure gut.
You don’t have to play FPL in a certain way if it is just for fun. But if you want the best rank, you will need to know how to use stats.
Expected data, also known as xG and xAG, is the most precise and useful tool we have. They correlate with FPL points better than any other metric. Of course, they can’t tell you everything, but they are two of the best tools we have.
What does xG mean and xAG?
xG stands to indicate expected goals. This statistic measures the chance of a goal scoring goal and its likelihood.
xG can be measured for both individual players and teams as a whole — this ultimately tells us how well they Should be performing in front of goal.
Expected assisted goals (xAG) — you may have also seen this metric called expected assists — measures the expected goals value of a shot that is assisted and gives a clearer idea of how many assists a player should have, given the quality of their attacking output.
How are xG, xAG calculated
xG is calculated using many factors, such as type of assistance, shot taken (head, foot, etc), angle, distance and chance. It’s the most sophisticated stat we have as it takes multiple data points and rolls them into one, giving you a predicted number.
Let’s say, for example, that Liverpool’s Mohamed Salah (£12.8million) has a shot at 0.5 xG — this means that the average player would be expected to score that chance 50 per cent of the time. Therefore, it’s a very good chance. xG can be measured on a scale from 0 up to 1.
xAG can also be calculated by a number factors like the type of pass and pass end-point as well as the length of pass.
Go Deeper
The Athletic’s football analytics glossary: explaining xG, PPDA, field tilt and how to use them
What are the benefits of expected stats for FPL?
First, looking at stats and — especially expected stats — is much more efficient and less time-consuming than watching 10 full games or even just match highlights to make an assessment of a player. Most people won’t be able watch more that two to three full matches per week.
Therefore, we need to look at stats to get the overall picture — and this allows us to assess players or teams quickly.
Stats can also be objective and factual and offer a neutral view of the situation. Stats take emotion out of the equation. While simply looking at a player to determine their suitability to your team is easy, it can become complicated by opinions and differing views.
Expected data includes several stats, making it a more accurate predictor. It’s much easier to compare players using xG, xAG, rather than multiple stats, such as shots, shots at target, and third passes.
We also need goals and assists from attacking players in FPL. This gives us the information we need without becoming overwhelmed.
The best place to get expected stats
The best place for freely available expected statistics is FBref.com.
By selecting “clubs”, you can click on your Premier League team of choice and check the full team stats from each player in that team — below is what you’ll see when viewing Liverpool’s page.
This provides a comprehensive overview of all you need including goals, assists, minutes played, and many other details. This provides the xG, xAG and player values. In the above example, I have filtered on “xG per 90 minutes”. It’s important to use both per 90 stats and the standard “expected” numbers.
We can also find the xAG of players by going clicking on the player’s name and then clicking on “Passing”.
That is what we can see. Darwin Nunez (£8.8m) is getting the most chances for Liverpool right now, but he won’t get 90 minutes most weeks — we have to add context to these stats, which is where your football knowledge, garnered through watching games and reading articles, comes to the fore.
By clicking “matches”, we can see a player’s stats in each match they play, which can also provide useful context. With a larger sample of matches, for example, we can see whether a player has higher xG scores against stronger teams, or if they are played in different positions. We can also see which position he gets the highest xG and/or xAG figures.
Scouting for new players in the league
It is easy to analyze new players entering the league by looking at expected data. Highlight reels don’t give the full picture as you are only seeing what the broadcaster wants you to see.
Let’s look at one of the new FPL assets this season, West Ham’s Gianluca Scamacca (£6.8m).
You can search for players in the search bar and also find them by going to their club’s page. This gives an individual outlook on a player’s performance over the last year compared to similar players in the “big five” leagues.
The above shows that Scamacca can score a lot of goals. His non-penalty goal totals are in the top 88% of his position. Additionally, he has high shot volume and above-average G per 90.
We can immediately see that while the Italian is a skilled goal scorer, his assist and passing threat are low. We can see that he has seven assists and 61 goals in his career, which is a total of 189 games.
Let’s compare him to Gabriel Jesus (£8.0m). We know that ArsenalMan is a versatile player who can provide assist and goal threat threats. His overall game is also much better. Jesus has scored 121 goals and assisted 57 times in his 302 career games.
This comparison takes less than five seconds and allows you to determine which player is better at scoring goals/playing FPL assets.
Why stats are better than the eye test and playing with your gut
Overall, I think that looking at stats gives you a more fact-based view. This can help reduce biases that can build up from simply looking at matches and outcomes. For example, if a player hasn’t scored in four games, without looking at stats, you would probably look to transfer them out as soon as possible.
However, the expected stats will help you decide if they are lucky or not. If their stats are good over the period they haven’t scored — ie, their xG and/or xAG scores are high — then you know to keep them and that attacking returns are likely not far away.
Also, it is more useful to look at expected stats rather than just looking at numbers such as shots on goal, shots into the box, big chances, etc. These advanced stats are better at predicting future returns and FPL points.
The eye test remains useful.
It is not to suggest that your judgment isn’t useful. While it can certainly be useful, it should not be used as the primary source of information when making decisions about which players to purchase or sell. Using solely what you’ve seen or only stats won’t provide you with the whole picture.
Both will provide you with all the information you need to make good FPL decisions.
Check out the FPL General. The Athletic FPL PodcastHe will be on hand twice a week throughout the season to share his latest tips.
(Photo by Mike Hewitt/Getty Images).
