Community Submissions

Attacking Potential – Next Six Gameweeks

The objective of this exercise is to look at recent form to help predict the attacking potential of each team over the next six Gameweeks. This aims to build on research already carried out by Footballoholic in his article  “Best chance of goals over the next six” which looked at the goals conceded by each team’s next six opponents.

Firstly, I’ll take a look at teams and players’ performances over the last six Gameweeks in terms of what would have been expected of them and what they actually achieved. This will give us a list of under performing, over performing and consistent teams and players.

These calculations are based on the average goals scored (GS) home and away in the past six matches, as well as the average goals conceded. I’ll then use this to predict performance over the next six Gameweeks and highlight key players to consider. Please note all decimal points in the tables are written in the European style as a comma.

Teams – Last Six Gameweeks

teamlast6

Players – Last Six Gameweeks

The expected attacking points calculation for each player is based on their involvement in goals scored over the past six Gameweeks, applied to the expected goals scored by the team over that period.

playerslast6image1

playerlast6image2

Team’s Attacking Potential – Next Six Gameweeks

The following uses form over the last six Gameweeks to predict expected points and goals over the next six Gameweeks.

teamsnext6

Players’ Attacking Potential – Next Six Gameweeks

Here are the top 20 players in terms of predicted attacking potential over the next six Gameweeks (please click on the image and use your server’s enlarge function to get a clearer image and see more details).

playersnext6

Conclusions

This research suggests teams are fairly consistent in their attacking behaviour (with 15 teams delivering as expected in the past six Gameweeks.

On the basis of opposition and attacking form, Alexis Sanchez and Sergio Aguero are still set to be involved in more than a goal a game over the next six. Newcastle, Southampton and Manchester City look the best bets for goals over the coming weeks, despite Southampton and Newcastle’s poor short-term fixtures. Aguero looks the safest option for the armband among players from these three sides.

Based on this analysis key advice is to:

  • Keep or invest in consistent players from consistent teams such as: Cesc Fabregas, Nacer Chadli, Diafra Sakho, Angel Di Maria
  • Keep your big impact players who can deliver as expected: Alexis Sanchez, Sergio Aguero
  • Tread carefully with overperforming players such as Christian Eriksen and Wilfried Bony.
  • Add a Newcastle forward to your watchlist. Their strong attacking form has the ability to confound their tricky fixture schedule.
  • Keep an eye on underperforming players over the last six games with good fixtures, such as Diego Costa and Steven Gerrard.

Gilles I am not spécial I love Ireland !!

61 Comments Post a Comment
  1. J0E
    • Fantasy Football Scout Member
    • Has Moderation Rights
    • 14 Years
    9 years, 3 months ago

    This is superb stuff. Please do click on the last table related to player potential to see it more clearly. Some really great information there.

    Thanks Gillesdegoult for submitting.

    1. Gilles
      • Fantasy Football Scout Member
      • 10 Years
      9 years, 3 months ago

      Thanks Jonty for your help on this.
      To give more details on the results (more players actually), how can I add a table in the comment section ?

  2. tm245
    • 12 Years
    9 years, 3 months ago

    Cool article. What is the methodology for expected points?

    1. J0E
      • Fantasy Football Scout Member
      • Has Moderation Rights
      • 14 Years
      9 years, 3 months ago

      I'm sure Gillesdegoult can provide that in the comment section. There was a lot to get in so we decided to stick to the results. Its quite complex.

    2. Gilles
      • Fantasy Football Scout Member
      • 10 Years
      9 years, 3 months ago

      "The expected attacking points calculation for each player is based on their involvement in goals scored over the past six Gameweeks, applied to the expected goals scored by the team over that period."
      Meaning I take each player's number of goals and assists over the past 6. This combined with the total number of goals scored by their team over the same period of time gives me an involvement per type of attacking threat in term of percentage (goals and asists).
      These percentages are then applied to their team's expected number of goals over the next 6.
      That gives an expected number of goals and assists on which I apply, depending on their playing position (goals = 4, 5 or 6 depending) to find an attacking points potential per player.
      I obviously do this for every player in the game...

      1. Kalix
        • Fantasy Football Scout Member
        • 12 Years
        9 years, 3 months ago

        Only taking 6 games is as your basis for predicting future results is going to lead to huge issues of sample size.

        Case in point, the gross underprediction of Alexis Sanchez.

        Why not just take all games for the season when looking at past data?

        1. Gilles
          • Fantasy Football Scout Member
          • 10 Years
          9 years, 3 months ago

          Did this last week and been asked to do with form. LoL
          Either way will be wrong as it is only predictions.
          And like every productions, they are wrongs 🙂

          1. tm245
            • 12 Years
            9 years, 3 months ago

            I think the timeframe can work since it might be the closest you can get to assigning statistical weight to the dark magic that is form.

            That is a lot of number crunching so thanks for the work.

  3. internal error
    • Fantasy Football Scout Member
    • 11 Years
    9 years, 3 months ago

    How did Obertan make his way into the table given his game time?

    1. Gilles
      • Fantasy Football Scout Member
      • 10 Years
      9 years, 3 months ago

      Newcastle has strong potential given the analysis.
      And he's been involved in 2 goals in the past 6 GWs.

      1. internal error
        • Fantasy Football Scout Member
        • 11 Years
        9 years, 3 months ago

        he's also out injured

        1. Gilles
          • Fantasy Football Scout Member
          • 10 Years
          9 years, 3 months ago

          Well... I will add a comment next week then 🙂
          Thanks for the info.

  4. rickardinho
    • 9 Years
    9 years, 3 months ago

    Cool article. Love these sorts of analyses!

    Which Newcastle forward can we invest in? Cabella? Not too keen on any of their strikers due to rotation issues...

    1. Gilles
      • Fantasy Football Scout Member
      • 10 Years
      9 years, 3 months ago

      Thanks Rickardinho.
      Not too keene on Cabella.
      I would definitely go for Cisse if he was nailed... But he's not.

      1. rickardinho
        • 9 Years
        9 years, 3 months ago

        Yeah definitely. That's why I was puzzled by the tip that we should get a Newcastle attacker... there's no clear player to get, unless Cisse (at 5.6iirc) is fit again.

      2. rickardinho
        • 9 Years
        9 years, 3 months ago

        I don't mean to judge you on your FPL rank, but could you post a link to your team? I'd love to see what team someone with such a refined statistical prediction model has! 😀

        1. Gilles
          • Fantasy Football Scout Member
          • 10 Years
          9 years, 3 months ago

          Can you please explain how to do that ?
          For instance, my current rank is really bad. Tried to be smart in the past 3 weeks and break my 'template team'. Taken 3 hits that costed me a lot (around 25 points more than the 12...).
          A good new advice to myself : don't be too impatient, take it easy 🙂

          1. rickardinho
            • 9 Years
            9 years, 3 months ago

            Haha, same here. I was at 17k but just had a 30 point GW. Breaking away from the template with a Siggy>Oscar move really messed up my team somewhat. Need to get back to basics!

            On your profile page you have a row called "website" where you can copy and paste the URL for your team history page. Think there's instructions somewhere on this site on how to do it best.

            1. Gilles
              • Fantasy Football Scout Member
              • 10 Years
              9 years, 3 months ago

              Done 🙂
              Hope that worked...

              1. rickardinho
                • 9 Years
                9 years, 3 months ago

                Nice looking team mate, apart from all those injuries! Damn, that is some really bad luck with baines, downing, chadli, sakho, wilson! Fortunately i only have two of those!

                1. Gilles
                  • Fantasy Football Scout Member
                  • 10 Years
                  9 years, 3 months ago

                  Unfortunately, I have the 5 of them and, for once, did 2 early transfers to avoid price rises 🙂

  5. shortymak
    • Fantasy Football Scout Member
    • 9 Years
    9 years, 3 months ago

    great article make transfers easy!!!

    1. Gilles
      • Fantasy Football Scout Member
      • 10 Years
      9 years, 3 months ago

      Don't bring Overran in though 🙂
      I'm gonna need a wildcard to come close to what the analysis suggests...

      1. Gilles
        • Fantasy Football Scout Member
        • 10 Years
        9 years, 3 months ago

        Obertan of course. Damn it !

  6. FPL P0ker PlAyer
    • 9 Years
    9 years, 3 months ago

    I'm loving this analysis and the final table in particular as I've got 5 out of the top 7 players suggested there, and I brought in Bony this week in preference to Costa. Fingers crossed that life mimics art in this instance...
    Great work. Thanks for finishing the job Footballoholic started.

  7. Woy of the Wovers
    • 13 Years
    9 years, 3 months ago

    The figures for expected points over last six. Where do they come from?

    1. Gilles
      • Fantasy Football Scout Member
      • 10 Years
      9 years, 3 months ago

      See above. Question asked by TM45. And answered.
      If it's not the case, feel free to come back to me.

      1. Woy of the Wovers
        • 13 Years
        9 years, 3 months ago

        Partially. I understand the team -> player link but not how team expected goals are calculated. While I'm here, let me say that this is top notch analysis. I may sound pedantic but I know that flaws can creep in that distort the picture.

        1. Gilles
          • Fantasy Football Scout Member
          • 10 Years
          9 years, 3 months ago

          Not easy on the iPad 🙂
          And I'm with my wife 🙂
          You can find explañations in footbalcoholic's article. Link is in the article above.
          Cheers.

          1. Woy of the Wovers
            • 13 Years
            9 years, 3 months ago

            Ok. I wanted to comment there but I was too late in joining the conversation. The flaw, as I saw it, was that team defences were ranked according to the goals conceded in the last six. It used this to determine easy fixtures over the next six but failed to recognise that historic fixtures would be distorted by the same idea of difficulty. If a team had tougher fixtures, in which they conceded more goals, it would rate them as easy targets in future fixtures. In short, the defensive strength of a team was not corrected, as I see, for the strength of their opponents.

            1. Woy of the Wovers
              • 13 Years
              9 years, 3 months ago

              Too add more detail, team defensive strength over a short time frame like 6 GW is better measured by Shots in the Box rather than goals conceded. Still needs to be corrected for opposition but it's a better measure of the defensive strength since goals are more elusive - and lucky - events.

              1. Gilles
                • Fantasy Football Scout Member
                • 10 Years
                9 years, 3 months ago

                Understand your point.
                I'll however keep it that way as I don't have access to shots in the box data 🙂

                1. Woy of the Wovers
                  • 13 Years
                  9 years, 3 months ago

                  It's so easy though. I can go look at the data in 15 seconds in one of my ready-made tables.

                  1. Gilles
                    • Fantasy Football Scout Member
                    • 10 Years
                    9 years, 3 months ago

                    And shots in the box doesn't give any fantasy points...

                  2. FPL P0ker PlAyer
                    • 9 Years
                    9 years, 3 months ago

                    I think you're right about the SITB% being a useful stat albeit probably best used alongside, rather than instead of, goals conceded, and it would certainly be interesting to see by how much the results differ. It could be a helpful model for making captain choices.

  8. Zasa
    • 9 Years
    9 years, 3 months ago

    What a great article, +1000 & Noble as 6th mid, cheers

    This will definitely affect which out of Bony & Rooney comes in this week for Pelle.

  9. rickardinho
    • 9 Years
    9 years, 3 months ago

    Very interesting that you have Southampton as one of the attacking teams to pick players from...

    They play MCI ars MUN next, which big games even though their defenses are a bit suspect, and then bur EVE cpl who are all pretty solid at the moment, especially at home in bur and cpl's case.

    Could you give some more details about why Saints are attractive as an attacking team still?

    The reason I ask is because I'm considering Pelle>Ings and Tadic>Hazard(-8) this week!

    1. Gilles
      • Fantasy Football Scout Member
      • 10 Years
      9 years, 3 months ago

      Can it wait until tomorrow as I don't have the information with me ?
      I will give you more details then.

      1. rickardinho
        • 9 Years
        9 years, 3 months ago

        Yeah of course, thanks so much!

        Please just reply to one of my posts here so I get an email notification 🙂

        1. Gilles
          • Fantasy Football Scout Member
          • 10 Years
          9 years, 3 months ago

          In the past 6, City has conceded 1.3 goals average per game away.
          Same 1.3 for Arsenal at home, Man United away and Everton away...
          Crystal Palace and Burnley conceded 2 goals avge at home.

          Southampton tends to be less than an average scoring team away (67% rate vs average = 100%), which means they would score 1 goal per game away.
          Their appeal at home is messed up with their game vs Sunderland (8-0) which gives them a 239% rate at home. That suggests they would score 3.2 and 4.8 goals in their 2 home games in the next 6... A bit too high maybe 🙂

          Not sure correcting those would provide help as we would try correcting every little things... Analysis is the key 🙂

          1. rickardinho
            • 9 Years
            9 years, 3 months ago

            Nice analysis!

            I think I'd rather go with Hazard and maybe Ings even if it costs me a -8. If Pelle can't get a shot on target against STO hul LEI avl, then there isn't much hope for him against better teams, despite what the stats say.

  10. Redhulkster
    • 10 Years
    9 years, 3 months ago

    Great article mate...keep up the good work.....

    Really helpful 😉

  11. Rolls-Royce
    • 10 Years
    9 years, 3 months ago

    Where is my boy Charlie Austin??

    1. Gilles
      • Fantasy Football Scout Member
      • 10 Years
      9 years, 3 months ago

      Made actually a mistake on QPR. I gave them 4 Home games and they have really awful goal scoring at home in the past 6.
      See above corrected full player analysis.

  12. Christina.
    • 14 Years
    9 years, 3 months ago

    😕

  13. Gilles
    • Fantasy Football Scout Member
    • 10 Years
    9 years, 3 months ago

    Top 30 correct now (damn Excel spreadsheet !!) :
    Name / Team / Actual attacking points / Expected attacking points / % involvement GS / % involvement Assists / Team GS next 6 / Potential attacking points
    Sánchez __________ARS_____33_____9,3_____66,7%_____11,1%_____8,7_____32
    Agüero ___________MCI_____32_____18,1_____66,7%_____0%_____11,3_____30,2
    Bony __________SWA_____20_____3,2_____62,5%_____0%_____10_____24,9
    Austin _____________QPR_____19_____5_____57,1%_____14,3%_____7,4_____20
    Sigurdsson ____________SWA_____14_____18_____12,5%_____37,5%_____10_____17,4
    Pellè _________SOT_____17_____16,5_____15,4%_____23,1%_____12,9_____16,9
    Tadic ___________SOT_____17_____7,8_____7,7%_____30,8%_____12,9_____16,9
    Fábregas ____________CHE_____17_____19_____9,1%_____36,4%_____10,8_____16,7
    Oscar ____________CHE_____16_____8,5_____18,2%_____18,2%_____10,8_____15,7
    Eriksen _________TOT_____18_____5_____37,5%_____12,5%_____6,9_____15,5
    Sakho ____________WHM_____15_____13,3_____30%_____10%_____10,4_____15,5
    Di María _________MUN_____17_____15,4_____12,5%_____50%_____7_____14,9
    Hazard ____________CHE_____15_____11,6_____27,3%_____0%_____10,8_____14,7
    Valencia ____________WHM_____14_____8,5_____20%_____20%_____10,4_____14,5
    Fletcher ____________SUN_____19_____0_____57,1%_____14,3%_____5,3_____14,3
    Ayoze ___________NEW_____12_____0_____33,3%_____0%_____10,7_____14,3
    Walters___________STO_____18_____0_____37,5%_____12,5%_____6,3_____14,1
    Baines _______EVE_____18_____11,8_____10%_____40%_____7,8_____14,1
    Berahino _________WBA_____12_____20,4_____50%_____0%_____6,9_____13,8
    Cissé ___________NEW_____11_____7,7_____22,2%_____11,1%_____10,7_____13,1
    Wanyama _________SOT_____13_____4,3_____15,4%_____7,7%_____12,9_____12,9
    Ings_________BUR_____12_____0_____42,9%_____0%_____7,2_____12,4
    Yaya Toure___________MCI_____13_____0_____16,7%_____8,3%_____11,3_____12,3
    Ameobi________NEW_____10_____0_____11,1%_____22,2%_____10,7_____11,9
    Diego Costa ________CHE_____12_____33,8_____27,3%_____0%_____10,8_____11,8
    Vargas__________QPR_____11_____0_____28,6%_____14,3%_____7,4_____11,6
    Jedinak___________CPL_____13_____13,1_____28,6%_____14,3%_____6,2_____11,6
    Downing ___________WHM_____11_____13,3_____10%_____20%_____10,4_____11,4
    Chadli ______TOT_____13_____19,9_____25%_____12,5%_____6,9_____11,2

    1. Twisted Saltergater
      • 14 Years
      9 years, 3 months ago

      Err.. no!

    2. FPL P0ker PlAyer
      • 9 Years
      9 years, 3 months ago

      Thanks for this immense work. I'm thinking that it might be a useful model for basing captain picks on. Your observation regarding Shots In The Box not scoring FPL points notwithstanding, it would certainly be interesting to see how much the results would differ if that stat was used rather than goals conceded.

  14. Gilles
    • Fantasy Football Scout Member
    • 10 Years
    9 years, 3 months ago

    Doest the link below works ?
    https://docs.google.com/spreadsheets/u/0/d/1wsdraW5E_Dekooi1HzVfl5pydu6g38a25_jtd1rp95M/edit#gid=0

    I wouldn't have to bother people with all the data 🙂

  15. Gilles
    • Fantasy Football Scout Member
    • 10 Years
    9 years, 3 months ago

    So next time, I will post these raw data on google...
    Seems to be better than directly on the post...
    Sorry again Christina !!!

  16. Slumdog Mignolet.
    • 11 Years
    9 years, 3 months ago

    Read this article when you wrote it and realy likedit but didnt follow any of your advice and lucky i didnt. Only been a couple games in to it so far I'll give you that, but..l

    "Keep or invest in consistent players from consistent teams such as: Cesc Fabregas, Nacer Chadli, Diafra Sakho, Angel Di Maria"

    Done crap so far

    "Tread carefully with overperforming players such as Christian Eriksen and Wilfried Bony."

    Done great so far

    "Add a Newcastle forward to your watchlist. Their strong attacking form has the ability to confound their tricky fixture schedule."

    Done crap since you wrote this lol.

    4 games left to prove me wrong that percentages and flashy excel tables can win you FPL 😛

    1. Gilles
      • Fantasy Football Scout Member
      • 10 Years
      9 years, 3 months ago

      Hi Slumdog,
      I can't say you're not right 🙂
      But as I said above, I made a mistake in the original spreadsheet which affects a lot final numbers... And the analysis behind it of course 
      I posted afterwards the link to the corrected table which should give you better insight.
      For example, you can see that Bony was actually second best forward to get after Aguero for the next 6 (24.9 attacking points expected). None of the Newcastle lads was in the top 15… Chadli was 29th only… Etc.
      Sorry for the misleading info on the article. It was work in progress. 
      By the way, it’s working fine for me as I got from 149K to 23K in 4 weeks with transfers based on this analysis : wildcarded GW13 with Costa, Downing, Sakho, Di Maria and Chadli out for Austin, Fabregas, Bony, Siggurdsson and Hazard, got Clichy for Dummett in GW15, and Aguero, Fabregas and Krul out for RVP, Yaya and Pantilimon in GW16.
      We’ll see over Christmas how it goes. I’ll keep you informed 

  17. John t penguin
    • 9 Years
    9 years, 3 months ago

    >:)