Even before the World Cup has finished, our heads are turned to the fantasy football season ahead.
I play fantasy football by the numbers using various analytics to aid my selections and captain choices. The first step for the new season is to gather information on players, positions, and points scored the previous season and use this to understand how I should be spending the budget.
Data summary
The summary data for FPL and Sky is in the following links:
<a href="https://1drv.ms/x/s!AkYcP_eLvYMVj0BcpIoqL-Swnl1G">Sky summary data</a>
<a href="https://1drv.ms/x/s!AkYcP_eLvYMVjz-hraN1mUaCQPph">FPL summary data</a>
Both spreadsheets have a similar structure with pages:
Summary &#8211; contains all the players, their position, price and points scored.
Season top 20 – list the top 20 by position based on total points, points per £m and points per game per £m
Team – Select a team from cell C3 and it lists that players teams.
VFM – Shows a scatter plot for each position of points per game by cost of player. A line of best fit is shown and the top eight players compared to the line are labelled.
Metrics and initial view
It is quite difficult to know the best metric to use to compare players, but my preference is the number of points per game per value of player.  The points per game, makes adjustments for players that have missed a game through injury or rested etc.  The chart below comes from the Sky spreadsheet above, that plays striker who have played at least 7 games.
<img loading="lazy" decoding="async" class="alignnone size-full wp-image-5813" src="https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/Capture.png" alt="" width="708" height="486" srcset="https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/Capture.png 708w, https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/Capture-300x206.png 300w, https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/Capture-610x419.png 610w" sizes="(max-width: 708px) 100vw, 708px" />
The solid blue line is a linear regression line.  The r2 figure (bottom right hand side) indicates how much of the variation is explained by the line.  The closer to 1 means a good fit and 0 means there is no relationship.  Here the value is 0.7276 which a very good fit.  Salah points have been adjusted as he is now classed as a striker and still looks like an outlier, I expect this is most due to him converting a high rate of chances compared to expected.
Choosing formation and where to invest budget
Once we have a regression line for each position we can plot them on the same graph as follows:
<img loading="lazy" decoding="async" class="alignnone size-full wp-image-5814" src="https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/fig2.png" alt="" width="670" height="428" srcset="https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/fig2.png 670w, https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/fig2-300x192.png 300w, https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/fig2-610x390.png 610w" sizes="(max-width: 670px) 100vw, 670px" />
The first step is to assume you only had £55m, where would the money be invested. You’d choose the keeper (blue line), then the defenders, followed by midfielders and finally strikers – leading to a 5-3-2 formation.
The next question is if you had £1m extra where is best to invest in an upgrade. Here you look at the slop of the line. It is clear that midfielders have a low gradient and hence should only be invested in if you are maxed out elsewhere.
Of course, an expert fantasy manager could go through the data making adjustments for expected points scored and ignore all the poor performers. As a proxy, I have just selected the top 15 performers for each position and produced the same analysis. The coefficients of the regression line are:
<img loading="lazy" decoding="async" class="size-full wp-image-5815 aligncenter" src="https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/fig3.png" alt="" width="298" height="159" />
Using either data from all positions or only using the top 15 players from each position lead us to a 5-3-2 formation. However, if you can identify the best performing strikers it is better to invest additional money in them before defenders. Midfielders are an area of low investment.
Does this apply to FPL?
I believe to a degree it does. The squad is fixed so there is no need to look at the constant for choosing a formation. The real benefit is knowing where to invest. In the FPL summary spreadsheet we have the regression lines of:
<img loading="lazy" decoding="async" class="alignnone size-full wp-image-5816" src="https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/fig4.png" alt="" width="651" height="431" srcset="https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/fig4.png 651w, https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/fig4-300x199.png 300w, https://cms.fantasyfootballhub.co.uk/wp-content/uploads/2018/07/fig4-610x404.png 610w" sizes="(max-width: 651px) 100vw, 651px" />
This does tend to indicate that there is value in investing in expensive defenders.
&nbsp;
Is this the solution for team selection?
The short answer is no!!!
The above doesn’t take into account players scores from last season being above or below expectation. For example, if a player had ten shots and scored all ten, then this is unlikely to be repeated. He may only be expected to score four times so the data overestimates his likely output.
We have new players joining the league and they aren’t included in the data and new managers will change formations and style of play.
More importantly, there are features in the fantasy game which aren’t accounted for here, eg captains worth double points, rotation of squad in FPL, etc.
Conclusion
This article doesn’t give you the ideal starting team for your fantasy team, but hopefully, give you thought on formation and where to achieve the best value investing the budget.  For both games, it seems to suggest investing in defenders and going for cheap midfielders.  You need to keep an eye on captain options.
I hope this has been helpful and good luck for the season ahead.
Join Carl in our new <a href="https://cms.fantasyfootballhub.co.uk/community/sky-fantasy-football-forum">Sky Sports Fantasy Football discussion forum</a>. Please also consider supporting the site and its contributors by becoming a member via <a href="https://cms.fantasyfootballhub.co.uk/member-registration/">this link</a>.

Should we be playing with 5 at the back?

7 Day Free Membership!

Community & tools

Live rank

Fixture ticker

Vote