FPL Gameweek 14: Abandon Ship

Tom Thomas
Fantasy Tactics and Football Analytics
7 min readDec 25, 2020
Photo: Scott McTominay celebrates scoring against Leeds United; source: EuroSport

I’ve only got one thing to say. Worst week ever.

First, let’s recap the week. We saw a return to goals galore. Both Manchester United and Liverpool inflict heavy damages on their opponents after they themselves got humiliated by much weaker opponents. It was odd to see that it was the very same two teams that bounced back this week. What a strange season it has been.

Anyone who held onto United or Liverpool players were in luck. 13 goals from these two teams definitely carrying a lot of points. Unfortunately, we were not so lucky.

Gameweek 14 Results

Gameweek 14 FPL Data Team; source: FPL
Gameweek 14 FPL Data Team Points; source: FPL

As far as points go, this isn’t our worst result. We’ve had worse weeks. Our lowest tally coming in Gameweek 1 at 39 points. Yet, when we look at our averages, this one has to hurt the most. In fact, we scored 44 points last week as well.

However, we were only 2 points below the average. This week, its a whopping 16. Gotta pat ourselves on the back for failing so miserably. Or have we failed? One can also consider that with the league average being so high, one observation we can make is that our team selections are outliers compared to the players the rest of FPL managers seem to hold. Hence, the huge differential.

We may be putting out a team that can only perform on some occasions. It could also be just a skewed result from the United and Liverpool results. We will have to wait and see how the trend progresses.

Vardy continues to impress and Kane had a poor outing. We were worried last week with much of our team players facing each other and the results show that. Bringing in John Stones seemed to be a great idea. Unfortunately, we kept in on the subs. It was unsure whether Guardiola would play him or not. So we had to pick someone who was a sure starter. This was a mistake, I admit. I could have put him in anyway and invoked an automatic sub. We will have to make more riskier selections from now on. Many of our stellar performers faltered this week either to tougher opposition or just a tiny blip. It was Grealish who bounced back after a few bad weeks with a better performance.

One positive again is that we finally have a clean sheet again from our Goalkeeper Mendy. Our goalkeepers have been struggling lately and this is nice to see despite the results

Changes incoming

Finally, we have a week where we can make some transfers that are not an immediate response to injury or bans.
I have two free transfers available this week and it would be very important to see who to let go and bring in given our poor performance over the past two weeks. We don’t have to use both but given that Ben Chilwell got injured, we can make way for another in-form player or two.

Updated Fixture Difficulty (Attack vs Defense)

Taking a look at the data for the past 14 Gameweeks, let’s focus on the Teams that scored the most amount of goals and the Teams that let in the least amount of goals. Let’s also add in the Teams that are in the most form at the moment.

Form, Goals put Forward, Goals scored against; source: FPL

If we combine these stats for attack and defense, we can form two new tables that would help us see the teams most capable of scoring goals/keeping clean sheets at the moment. We will face these teams against their opponent’s defensive/offensive strength for the next few Gameweeks.

ATTACK VS DEFENSE & DEFENSE VS ATTACK — These two tables are interconnected. The numbers presented are opponents. On the left, the most attacking teams are put against the strength of their opponent’s defense (which is taken from the table on the right). On the right, the opposite of this is done so.

Based on these two new tables, we can crunch out two new rankings.

(Left) Best attacking teams vs. Worst Defenses, (Right) Best Defenses vs Worst Attacking teams |Based on Schedule GW 15–20 | Capital cases represent Home games

We omitted Gameweek 19 as that is Double Gameweek for all with more 10+ fixtures.
What does all this mean for us? Great question. You may have guessed a bit of it by now but what we did helps us to differentiate between the attacking players and defensive players we want to add to the team. We can see that the easiest way to score attacking points may come from teams such as Manchester City, Southampton and so on. For defensive points, it would be from Liverpool, Manchester United and so on.

We do see a lot of similar teams on both, our aim is to look at the schedule as well and make a concrete decision.

One thing to note is that only a few teams play Gameweek 18. We want to add that into consideration as well.

Let’s also consider how our team is doing at the moment. We already have a Manchester City player in John Stones who seems to be doing well so it would be good to keep him. He isn’t guaranteed minutes but he seems to be getting more minutes with his recent form so it might be good to take this risk.
The worst performers seem to be Konsa, Coufal and Chilwell of all people. Considering that Chilwell is a doubt for the next two games, he will be the first to go out.

The next question is who follows him.
It could be any one of Bowen, Konsa, Coufal, or Ward-Prowse. To solve this dilemma, we’ll take a look at the table we created above. We want to get rid of the player with the worst upcoming opponents.
Bowen and Coufal face Brighton, Southampton, Everton and Crystal Palace. All are tricky games but West Ham aren’t a bad team either.
Konsa faces Crystal Palace, Chelsea, Manchester United, Tottenham and Burnley. That’s a tough stretch of games. Konsa seems like the first one out so far.
Ward-Prowse faces Fulham, West Ham, Liverpool and Arsenal. Two tough games and two relatively easy games. Given Arsenal’s poor form, that maybe an easier game as well for them.

However, there is one more thing to consider, do we want to add two defenders or do we want to add one defender and one midfielder? The diversity maybe of value to us.

After hours of staring at the screen and blanking into oblivion, this is what we have decided.

We will lose Konsa and not another midfielder.
Why?
1. There are no valuable players available for sale at the price point of Bowen or Ward-Prowse.
2. The schedule seems to favor Ward-Prowse so we may give him a chance just yet.
Lets take a look at the best available defenders.

Most Valuable Defenders, source: FPL

Observations

  1. From our defense table, the teams to consider are City, Leicester, Liverpool, United and Chelsea.
  2. We would love to take Kurt Zouma but there wont be a valuable defender to add with him. Another great player is Andy Robertson. Both cost a lot. After selling Chilwell and Konsa, we would only have £10.5 left.
  3. Considering all scenarios, the two value picks we see are Victor Lindelöf and James Justin. Adding to our ever growing list of James’ in the team!

OUT: Ben Chilwell, Ezri Konsa
IN: Victor Lindelöf, James Justin

And there we have it, our team for Gameweek 15.

Gameweek 15 FPL Data Team, source: FPL

Thank you for taking your time to read through this. If you’d like to follow along in this series, here is a timeline of the FPL data analytics project.
1. Introduction
2. Gameweek 1
3. Gameweek 2
4. Gameweek 3
5. Gameweek 4
6. Gameweek 5–8
7. Gameweek 9–10
8. Gameweek 12–13

— — — — — — — — — — — — — — — — — — — — — — — — —

And there we go! Hope you enjoyed reading this piece as much as I did writing it. Feel free to write down any thoughts or suggestions. I’m always working on improving my analysis and my articles as time goes on so I appreciate all comments! Collaborating is fun so if you’ve got any interesting projects in mind, feel free to reach out to me personally. If you would like to follow me to keep up with updates in this series, follow me here or — — @__tomthomas

— — — — — — — — — — — — — — — — — — — — — — — — —

--

--

Tom Thomas
Fantasy Tactics and Football Analytics

I geek on stats • Certified IBM Data Scientist • Industrial Engineering Background • Exploring the power of data • tomthomas.github.io