Community

Is Salah the Most Expensive Player in FPL – A Theory on Effective Player Price

The key to success in FPL is to pick players that outperform their pricing in the game. A player who outperforms his pricing would then be considered good value. One way to look at player value is to calculate points per million (pts/m). If a player scores 200 points a season and costs on average 8m he is providing you with 25 pts/m. The starting budget is 100m so if you would average a 25 pts/m over the season with a budget fixed at 100m that would result in a 2500 points season.

But that’s not true, is it?

This calculation is ignoring the most important factor in FPL and that’s captaincy.

Each week FPL managers pick a captain who gets his points doubled for that particular gameweek (GW). Now let’s set up a typical scenario for a FPL manager who is picking his team for the 21/22 season. Let’s assume he plans to wildcard after the second international break. This means we are setting up a team for the first 7 GWs. Let’s agree on the consensus that Mohamed Salah will outscore all other players during that period so he picks Salah into his team. Bruno Fernandes is also tipped to do well so he puts him in the team as well. Salah and Bruno are both premium players priced at 12.5m and 12m respectively. On paper, Salah is more expensive but one could argue that Bruno has a higher Effective Player Price (EPP). Let’s explore how one could reach that conclusion.

The first 7 fixtures for Salah and Bruno are (captaincy in bold):

Salah: norBUR – CHE – lee – CRYbre – MCI

Bruno: LEE – sou – wolNEW – whu – AV – EVE

For simplicity let’s assume these two are the only two captaincy candidates in the team (although one could probably find a better pick in GW3 and even GW7). I think it’s fair to say that of those two Salah would be captained in at least 4 out of those 7 fixtures (GW1, GW2, GW5 and GW6), so let’s go with that scenario.

Now let’s define the term EPP as the player price multiplied with the Effective Points Multiplier (EPM).

The EPM is defined as

– 1 for players who you own (bench included)

– 2 for your captain (3 if triple captain)

Advancing with our scenario we can calculate the EPP for the two players over the 7 games period in the following way:

Salah: ((12.5/2 x 4) + (12.5 x 3)) / 7 = 8.9m

Bruno: ((12/2 x 3) + (12 x 4)) / 7 = 9.4m

These results show that although Salah is priced higher than Bruno he has a lower EPP than Bruno. This is because Salah is getting double the points for the price of one more often. This also illustrates that both players have a lower EPP than lower-priced players such as Heung-Min Son (10.0m) or Jadon Sancho (9.5m) if we would add them to the team but never captain them. Even if we would captain Son in GW3 vs WAT his EPP would only drop to 9.2m which is still higher than Salah’s 8.9m.

Back to the question in the headline. Is Salah the most expensive player in FPL? The answer is no if you like most others would captain him often enough that his EPP would drop lower than EPP of other players. In all likelihood the most expensive player is the highest-priced player that you would never captain.

Thank you for reading, if you have any questions or comments you can find me on Twitter @FPL_Disorder

32 Comments Post a Comment
  1. Rotation's Alter Ego
    • Fantasy Football Scout Member
    • Has Moderation Rights
    • 12 Years
    2 years, 9 months ago

    Thanks for writing Disorder!

    1. riot
      • Fantasy Football Scout Member
      • 13 Years
      2 years, 9 months ago

      Hey RAE, a feature I'd quite like to see added to community articles is a way to quickly rate them (similar to how the Athletic does it, https://i.imgur.com/VqFv77Y.png). As a side note, I liked this one - although short, it made me think!

      1. Rotation's Alter Ego
        • Fantasy Football Scout Member
        • Has Moderation Rights
        • 12 Years
        2 years, 9 months ago

        Interesting. I'll see what those in charge think - I guess for Athletic it's important because everything is behind a paywall and they want to work out what people like reading, whereas here I always stress the importance of writing about what you want to write about (that's where the interesting, unique articles tend to come from). Sharing feedback, in comments, I think are a great way of showing appreciation for articles like these but I guess making that easier would be good?

  2. FPL Xylophone
    • Fantasy Football Scout Member
    • 8 Years
    2 years, 9 months ago

    Interesting read. Thanks!

  3. Marmalade Forest
    • 6 Years
    2 years, 9 months ago

    Superb article

  4. RedLightning
    • Fantasy Football Scout Member
    • Has Moderation Rights
    • 13 Years
    2 years, 9 months ago

    Another way if looking at it is "Is Salah the best value player in FPL?"
    In FPL's Stats tab, it calculates Value (season) as the total points scored last season divided by the player's new price.
    If you are expecting to captain Salah 4 times out of 7, then his effective value becomes Value times 11/7 and increases from 231/12.5=18.48 to 18.48 x 11/7 = 29.04,
    and if you expect to captain Bruno 3 times out of 7 then his effective value becomes Value times 10/7 and increase from 244/12=20.33 to 20.33 x 10/7=29.05.
    The only non-captains with greater Effective Values according to this would be Martinez, Dallas and Meslier.

    1. Andy_Social
      • 11 Years
      2 years, 9 months ago

      Would it make sense to risk perma-captaining Salah - or Kane if he starts well (assuming no injuries) and having no other premiums (i.e. 10m+)?

      1. FPL Disorder
        • Fantasy Football Scout Member
        • 12 Years
        2 years, 9 months ago

        This is definetely a viable strategy but I would be open to selecting mid priced assets that are captainable in the difficult fixtures for your perma-captain.

      2. Scrumper
        • Fantasy Football Scout Member
        • 4 Years
        2 years, 9 months ago

        This is my plan. Going with Salah to start, will switch to Kane if he moves to city. Happy to have punty VCs as back up

    2. FPL Disorder
      • Fantasy Football Scout Member
      • 12 Years
      2 years, 9 months ago

      True, this is one way to take this further. But, then you are dependant on good point projections. Using last year points is not very accurate. Especially for players like Dallas who have changed positions in FPL. I would also argue that for GKs you need to add 4m to the player price since you have to select two GKs but can only play one.

      1. RedLightning
        • Fantasy Football Scout Member
        • Has Moderation Rights
        • 13 Years
        2 years, 9 months ago

        I agree that last season's points totals may not be the best estimates of this season's points.
        But whatever method you use to estimate how many points players are likely to score over the next few games, adjusting their effective values according to the proportion of the next few games that you expect to captain them could still be useful.
        The initial 4.0m overhead on a non-playing bench goalkeeper applies to other positions too - there is also an initial overhead of 4.0m on every goalkeeper and defender, and 4.5m on every midfielder and forward.
        I prefer to compare sets of players at the same combined prices in the same sets of positions and see how these affect the squad as a whole rather than trying to compare values of players at different prices in different positions.

        1. FPL Disorder
          • Fantasy Football Scout Member
          • 12 Years
          2 years, 9 months ago

          I totally agree that the goal is to maximize value from out budget but then it becomes too complicated for a single article in my opionion.
          I dont think you can put a 4m or 4.5m overhead on other positions than the goalie. Thats because the “overhead” is shared between the 10 outfield players and the bench itself isn’t a pure overhead since it can actually bring you points when your starting XI doesn’t play and even rotate with some players.
          My point with a 4m overhead on GKs was purely based on If you would have Martinez or Meslier you in all likelihood would have a non-playing 4m GK on the bench as a pure overhead.
          You could have two 4.5m and rotate then it doesn’t make sense to calculate values for each GK instead you would need to calculate the value for the 9m combo.
          But to reiterate I agree with the value approuch it is just really hard to simplify for a short article.

          1. RedLightning
            • Fantasy Football Scout Member
            • Has Moderation Rights
            • 13 Years
            2 years, 9 months ago

            I took the main point of your article as demonstrating how premium players who are frequently captained can provide as good or even better value than cheap or mid-priced players who aren't.
            Taking the value approach to the entire squad is a far wider subject and also involves considering the compromises that arise from dealing with budget constraints. If you had an unlimited budget then value would be irrelevant, and if you got all the 'best value' players but left money in the bank then that money in the bank would be wasted.
            For goalkeepers, I would look at them as a pair rather than trying to value them separately. You should consider whether you are likely to score more points from 4.5m+4.5m or 5.0m+4.0m. And you should also consider whether the squad would benefit more by spending 9.0m on goalkeepers and 91.0m on the rest or by spending 8.5m on goalkeepers and 91.5m on the rest. There is no point in giving an individual value to your non-playing 4.0m goalkeeper or in comparing this value with any other players.
            The initial 4.0m and 4.5m overheads have the same effect on all players. These are the minimum prices you have to pay whether you like it or not, and they have an artificial effect on all players when calculating values, so it is better just to look at how many points you can expect to get by spending the 36.0m that remains after overheads in different ways - and that is not trivial.

          2. RedLightning
            • Fantasy Football Scout Member
            • Has Moderation Rights
            • 13 Years
            2 years, 9 months ago

            When players are captained, their effective value is enhanced.
            Similarly, when players in your FPL squad are benched (either through your choice or through their being rotated, injured or suspended in real life), their effective value is reduced.
            For example, if you expect one of your players to be in your first eleven, or auto-subbed into it, 4 out of every 7 matches, then you could multiply his basic value by 4/7 to get his effective value.

            It isn't really quite as simple as that though - you may only captain players for their more favourable matches, so their expected scores in these matches should be higher than their average scores in other matches. This means that if you captain a player 4 times out of 7, then doubling his expected score for the matches in which you captain him will probably gave a different result from simply multiplying his average score by 4/7.

      2. Tarby
        • 4 Years
        2 years, 9 months ago

        It's a good strategy I think FPL Disorder.

        I try to benchmark value. For example with defenders the very top is only 7.5m (lets predict 180 points) and cheap options can be 4.5m (100 points maybe), the extra 3m is likely to get something like 80 more points perhaps.
        180/ 7.5 = 24.00
        100/ 4.5 = 22.22

        But the extra cost is actually getting better value still: 80/3 = 26.67

        What else could be done with that 3m though?!?!

        For defenders you could have 5 x 4.5m players and rotate 3 of them, perhaps could hope to get 400 points from doing that if you do well. 400/ (4.5 * 5) = 17.8
        Or you could go big with Trent, Robertson, Digne/ Stones/ Maguire perhaps and 2 4.0m fodder but hope for 500 points.
        500/ 28 (total cost of those 5) = 17.9

        So would be marginally better value, but you'd likely to 2 bench players who rarely play so a weaker squad and an injury to one would be problematic so have to think of those things as well.

        With keepers this method is obviously much easier!

        Combine the cost and if rotating estimate your points total and compare with more expensive with a 4m option. With rotating it can also be helpful if there were games postponed or for when DGW come into play too!

    3. circusmonkey
      • Fantasy Football Scout Member
      • 13 Years
      2 years, 9 months ago

      I am going one premium this year (Salah to start with) and my next most expensive is Havertz although that will change if they sign a striker. I can get 4 premium defenders, 3 min price mids, 2 mid price strikers, cheap goalie, so this article confirmed my view, which let's face it is why we like articles.

      1. circusmonkey
        • Fantasy Football Scout Member
        • 13 Years
        2 years, 9 months ago

        I mean mid price mids!

      2. FPL Disorder
        • Fantasy Football Scout Member
        • 12 Years
        2 years, 9 months ago

        I like this approuch in theory although I usually end up with at least 2 premiums.

      3. Andy_Social
        • 11 Years
        2 years, 9 months ago

        Similar here, with Sancho as 2nd-most expensive. The challenge will be sustaining one premium when Kane goes to City!

        1. FPL Disorder
          • Fantasy Football Scout Member
          • 12 Years
          2 years, 9 months ago

          Kane will be popular come gw 2-3. I think early WCs will be extremely popular this year

          1. Andy_Social
            • 11 Years
            2 years, 9 months ago

            For me, there's really 2 windows: after GW3, TDD, 1st int break, or later after GW7, 2nd int break.

  5. LangerznMash
    • 7 Years
    2 years, 9 months ago

    What I took out of this article is there is no point owning high priced players (£8.5m - £10.5m) unless you plan on captaining them.
    I only have one player in this price range, Mahrez, however he is only a short team choice and I might consider captaining him GW2.

    1. Andy_Social
      • 11 Years
      2 years, 9 months ago

      Pretty much. It depends where you draw the line. I said 10M which rules out Son but rules in Sancho. Maybe 8.5 should be the maximum - ruling out Mahrez but Foden or Havertz would be fine. In my current draft after Salah, my most expensive assets are 9.5, then 9.0 then 4x7.5s. My defence and midfield are pretty powerhouse, but the attack is 2x7.5. It looks good.

      But the real dilemma will come later when Kane moves - or doesn't and carries on where he left off at Spurs. Will I lose Salah to have Kane, or will I keep Salah and pass on Kane?

  6. Fpl Old Guard
    • 12 Years
    2 years, 9 months ago

    Top 3K any year league code “lacz6l”. Aiming to create a worlds best league.

    1. Phlajo
      • 5 Years
      2 years, 9 months ago

      A private league score is the average of the leagues top 5 managers, right?

      In that case a former top 3k means nothing, apart from trying to boost your ego (which clearly seems to be important to you, so GL)

      🙂

    2. RedLightning
      • Fantasy Football Scout Member
      • Has Moderation Rights
      • 13 Years
      2 years, 9 months ago

      We already have Top 10k Any Season, Top 1k Any Season and Top 100 Any Season mini-leagues, but when it comes to FPL's table of the best mini-leagues, which as PHLAJO says is based on the average scores of the top five managers in each league, none of these can compare with the FFS Open and Members Leagues, or others with very large numbers of teams.
      And the top teams in these leagues are often not the ones with the best career histories.

  7. Fpl Old Guard
    • 12 Years
    2 years, 9 months ago

    I didn’t know only the top 5 are used to work out the average, you are right then, the league with the most members will always be top of the charts purely based on luck

  8. Tarby
    • 4 Years
    2 years, 9 months ago

    I think like this but without having thought about actually assigning it value, I really like that.

    All about trying to find the opportunity cost of one player over another.

    I'm not sure simply applying a factor of 1 to all non-captains is ideal though, bench players could be rated in order.

    Keepers grouped together as a total cost

    If you have Welbeck on your bench at 6m vs Davis as 4.5m the EPP needs to be adjusted as they're on the bench. Welbeck will likely score a lot more points than Davis, but if either is mainly on the bench then the actual points might only be 10 - 20 or something. Spending 1.5m for lets say 15 points extra is a bad use of funds.

    I think I've suffered by having strong benches in the past, and could have 1 or 2 places with cheaper players as they incredibly rarely get into the starting team anyway! Why have Mitchell over Kelly and spend 0.5m if I expect them to be 3rd sub every GW!

    Thanks for the thoughts FPL Disorder

    1. FPL Theorist
      • 4 Years
      2 years, 9 months ago

      Magnus Carlsen is brilliant at this. I think he had a week last year where he got a CS or attacking return from both Struijk and Kilman in the same week. He actually spends a fair number of transfers getting the "right" 4.0 defenders for short periods of time, knowing that they're not nailed long-term.

      It's a super-elite strategy and I haven't really seen anyone else pull it off. Maybe a couple of other guys in the Great and the Good league at most.

    2. Blues for Salah
      • Fantasy Football Scout Member
      • 4 Years
      2 years, 9 months ago

      Very interesting idea. I'd add to that the idea of having a pair of 6.0 to 6.5 ish players who you think of as assets you rotate to try and take advantage of matchups. Even assuming that you pick correctly week after week, what's really the net impact of that rotation of your overall score, and would you actually get more productivity concentrating that value in one asset or the other?

      Having a couple reasonably productive 4.5 assets on the bench to cover for surprise absences from your nailed XI is good policy, but I wonder whether the marginal bench spending beyond that is justified by the numbers - especially given that one can swap out a member of the starting XI via transfers as need arises.

      1. Tarby
        • 4 Years
        2 years, 9 months ago

        Using transfers on the cheap players but getting them right might be a good strategy. The top players the opponents doesn't matter as much!

        As you've shown interest I'll share what I found yesterday! Based on last season (which might not be the best to base it on!) having 4 defenders in the 5 - 5.5 range and rotating would actually have been the best value possible! That's excluding using Dallas as obviously picking up from the start and combing with Targett without even rotating would have been great!

        That was based on avoiding the 'top 6' plus Leicester and also away to Everton where possible. And using the same 5 players with no transfers as well.

        I think TAA and Robertson will be back to 180+ this season though, which could change this. I'll have to do it for 2019/20 season as well and see what comes from that.

        At the moment I'm thinking three 5.0 - 5.5, with one at 4.5 and one fodder at 4.0 and look to transfer in what's likely to be the 4.0 bargain when they become clear and likely bring in another 4.5 when the top prospect there seems to have made themselves known!

        On keeper last season if you rotated Ramsdale and Guaita on easiest game, would have been 145 points (they got 123 & 124 individually). So the extra 0.5m would have been worth 20 points which would have been great value.
        Johnstone and Guaita (140 & 124) would have been 168 points, so more like 30 points for the 0.5m.

        So keeper rotation seems a very good idea, any 2 keepers who play 38 games and avoid the good teams and would create a top keeper with them

  9. Clouseau
    • Fantasy Football Scout Member
    • 8 Years
    2 years, 9 months ago

    I love articles like these, particularly where they provide a science-/maths-based approach which makes you look at the game in a different light. Well-written and well-explained, so thank you, Disorder.