ICT Index Quadrants

Introduction

This season I have been using the ICT index as provided by the Official FPL site to assess players and assist my transfer decisions. I have done this for a couple of reasons. Unlike Opta stats, it is freely available and permission is not required in order to publish articles which reference the metric. Secondly, like many managers I had previously ignored the metric and now wondered if it was actually useful? At the end of this article there are a few examples of transfers where I have used the metric.

I have not just used the raw metric but have create Points vs ICT Index Quadrants spreadsheet. Screen shots of these have been published on Twitter and now I want to make the spreadsheet available to everyone.

I would also like to acknowledge the help given to me by Ross Kenny @deadlegswamp who gave me feedback on the spreadsheet and made suggestions, which I have implemented.

What is it good for?

The idea is that it can be used to spot players who are underperforming and overperforming to their ICT Index.

Figure 1 Basic Points vs ICT Index Quadrants

The basic quadrant plots Points on the Y-axis against ICT Index on the X-axis and at each intersection the %TSB is displayed. The axes are divided at the mean of the metric. The four quadrants are as follows:

Performing (Green) Compared to the other players these ones have returned the most points and they are returning inline with their ICT Index. You should already own some of them!

Overperforming (Upper Yellow) Compared to the other players these ones are overperforming to their ICT Index. Their points return may drop off especially, if they have difficult fixtures.

Avoid (Red) Compared with the other players these ones have a poor ICT Index and points return. If you own them, then maybe you should look at moving them on.

Watch (Lower Yellow) The players in the bottom yellow quadrant have underperformed to their ICT Index and should be considered for inclusion in your team. Especially, if they have good fixtures, as shown by their Elo FDR rating. The Elo Fixture Difficulty Rating is based on team Elo rating. A link to a fuller explanation is here: Elo FDR

Instructions

At the top of the spreadsheet there are a number of parameters which can be selected.

Figure 2 Selectable Parameters

Range:

Two ranges can be selected either the full season or the last 4 game weeks. Note, when a change is made to the range, the user will have to manually sort Column N: Points Pts/Last4GW in descending order.

Quadrant split:

Two options are available either Mean or Median. Whichever one is chosen will be applied to both the X and Y axis.

X:

The variable plotted along the X-axis The options available are dependent on the RANGE selection

Y:

The variable plotted along the y-axis. The options are either Points or Points/Last4GWs. They are automatically selected when the range is selected.

Z:

The parameter plotted at the intersection of X and Y axis and is fixed as %TSB. Note this is not live but the value at the end of the previous game week, when the player data was downloaded from the FPL site.

Focus:

How many players to focus on, either 10,20,30 or 40. Set this parameter and then filter the Focus (column H) on YES to generate the Quadrants.

Slightly more players may be displayed if there are players tied in 10th, 20th, 30th or 40th places.

Column Headings

The filters on these are used to create the quadrants and the titles for the most of them are self-explanatory. However, some of them need to be clarified.

Figure 3 Column Headings

Focus

This is the most important one and should be the last one that is filter. Depending on the value selected for Focus, in the top section, that number of players has a YES against their name.

ICT

Influence, Creativity and Threat. The following is the ICT Index description as from the Official website:

Elo

This is the Elo FDR rating for the next 6 game weeks and is calculated by my Elo FDR spreadsheet. More information on Elo ratings can be found in this link: Elo FDR

Process

When the spreadsheet is first opened a full list of players is shown. The top parameters should be set, and the columns filtered to generate the desired quadrants.

1. Set parameters in upper section.

2. May need to sort column N ‘Points’ in descending order.

3. Set the column filters. Usually, this will be column F ‘Pos’ so either GKP, DEF, MID or FW.

4. To create another set of Quadrants it is recommended that all the column filters are reset.

Anomalies

Switching from Season to Last 4GWs and vice versa.

When this is done the X: parameter will be highlight in yellow/red. This is to indicate it needs to be updated by the user.

Figure 4 X Parameter Anomaly

Also, if the quadrants look jumbled as shown in figure 5, then column N needs to be sorted in descending order.

Figure 5 Column N Sorting Anomaly

Players with the same ICT Index

Sometimes two or more players will have the same x-axis value. This can lead to duplicate columns and can have a slight affect on the x-axis Quadrant Split.

Update

At the end of every game week I cut and paste data from the Official site. This means that Prices and %TSB are not live and can be slightly out of date. Points and ICT Index are not affected and are always up to date.

Examples

I have been using the spreadsheet for a number of weeks and now follows an few examples where it has helped with my transfer decisions.

Silva GW4

Figure 6 shows the post GW4 Midfielder ICT Index Quadrant.

Figure 6 Gameweek 4 Midfielder Quadrants

I used this to support the decision to swap Mkhitaryan for D Silva who had the joint lowest points, but the 4th highest ICT index. He then scored 11 points in game week 5.

Mane to Hazard GW4

When I played my wild card, I wanted to compare Mane with Hazard. I knew Hazard had not started any of the first 4 games so to make a fair comparison I plotted Points vs ICT/90.

Figure 7 Game week 4 Points vs ICT/90

He had the highest ICT/90 of all MIDs at 17.8 compared with Mane’s 11.3. Figure 7 also confirmed Silva had the second highest ICT/90 at 15.8.

Game Week 12 Anderson

In Game week 12 I replaced Maddison with Anderson.

Figure 8 Post Game Week 11 Sub £7.5m Midfielders Last 4 game weeks Points vs ICT Index

Figure 8 showed that Felipe Anderson had the 4th highest ICT index over the previous 4 game weeks however, he had only scored 18 points. I brought him in for a one-week punt, as WHU had a good away fixture to Hud and he returned 9 points.

Finally

The above examples demonstrate the ICT Index can be a useful metric when assessing players. It should not be used on its own, but in conjunction with other metrics and the eye test.

I have used it ahead of game week 13 to assess my Mendy replacement.

Figure 9 Post Game Week 12 DEF Last4GWs Points vs ICT Index Quadrants

Digne over the last 4 game weeks has the highest ICT Index of all defenders and is firmly in the Watch Quadrant. On the other hand, Laporte is firmly in the Avoid Quadrant.

The spreadsheet that I used to create the quadrants is available for download here: ICT Quadrant File Post GW12

Please give it a go and let me know what you think.