There are many telltale ways to differentiate between an FPL veteran and a casual just by looking at their team during one of 38 FPL game weeks over the course of the season. Maybe their team is riddled with injuries and suspensions, or they do not go with any of the 7 million plus pounds defenders because they are ‘too expensive’. Or, maybe they spend 11million on the goalkeeper position so that they can rotate the two to maximize points.
However, in recent weeks something about this felt out of place to me. Because while goalkeeper rotation is shunned as a rookie tactic in FPL, similar logic is applied to captaincy rotation, where teams usually keep 2 premiums in their team with the objective of rotating between them and maximizing points.
Of course, the logic isn’t exactly the same (mostly because goalkeepers usually benefit from difficult fixtures, so you don’t want to rotate them out), but I decided to look at what has happened in the season so far, to see if I can answer some important questions about our team structures and FPL premiums this season.
Method
I wanted to find out how many premiums that the optimal FPL team would have. To measure that, however, I must first decide what to measure. At first, I settled on looking at the points per game of the different premium arrangements, as FPL is a game to score the most points. However, I also wanted to take into account that FPL is played on a budget, and since we have an upper limit on price, premiums should also return in relation to their price (in other words, it does not matter that Salah scores 400 points a season if he costs 90million. Their price must also be reasonable). So, I added another variable to measure value: points per game per million. Using these, I will be able to determine the net return of a premium set up, as well as their return relative to their price.
I also decided to look at six premiums in particular: Heung-min Son, Harry Kane, Raheem Sterling, Erling Haaland, Mohamed Salah and Kevin de Bruyne, because I believe that these were the only premiums that were ever considered as FPL options so far this season.
Finally, I should mention that I performed this analysis before game week 6 had started, and as such it was not taken into account.
Analysis
So, I started by looking at the point returns of all premiums over the past five game weeks.
What was immediately notable was Haaland’s standout season so far. With 58 points in just 5 game weeks, while also being the second cheapest premium on my list, he is easily the best premium in the game so far, based on both raw returns and returns relative to price.
| Name | Price | Points (Rank) | PPV (Rank) |
| Haland | 11.5 | 58 (1st) | 1.01 (1st) |
| Kane | 11.5 | 31 (4th) | 0.54 (3rd) |
| Son | 12 | 15 (6th) | 0.25 (6th) |
| KDB | 12 | 28 (5th) | 0.47 (5th) |
| Sterling | 10 | 32 (3rd) | 0.64 (2nd) |
| Salah | 13 | 35 (2nd) | 0.54 (3rd) |
Besides Haaland however, the table makes for a poor reading for most of the other premiums. Good FPL players usually have a PPV of at least 0.7, but since premiums are usually kept for their captaincy potential (doubling their points, and therefore doubling their PPV), they get a little leeway on their PPV demands, and 0.6 is normally seen as fine. However, even at that, only Haaland and Sterling hit that benchmark. So premiums have not been incredible so far over the season. But what about when they are used in tandem with another premium?
The idea with 2 premium teams is that, you can avoid captaining player A during a difficult fixture by captaining player B. When done perfectly, points are gained by rotating the captaincy between the two players. So, assuming perfect captaincy rotation (the player with more points that game week has their points doubled), I calculated the points return of all possible 1 and 2 captain combinations.
- Note that I excluded Heung-min Son from this. Simply because he has been too bad an FPL option. If you started with him this season (as I did), you have my condolences.
As you can see, 2 premium combinations dominate. Haaland being the only single premium option that beats any 2 premium combination is testament to that. Logically it also makes sense, as I assumed earlier that players would make the correct captaincy decision every game week.
So, we know that 2 premiums have 1 premium teams beat on pure FPL points returns. However, we also want to know whether their points return is worth the extra spend. So, I calculated the PPV’s of all combinations assuming perfect captain choices.
Again, Haaland and combinations including him lead the charts. But a more important inference is that when it comes to PPV, 1 premium combinations are actually doing better than 2 premium combinations. What this implies is that adding an extra premium to our FPL teams is an attempt at maximizing points, even at the expense of our (FPL) millions.
After all this analysis, the next question came about naturally. If 1 premium combinations have a higher PPV, while 2 premium combinations return more points, which should be prioritized?
From this point, I’m leaving the realm of empirical evidence. This question has its roots deeply entrenched in the science of optimizing an FPL team, which is beyond my region of expertise, even when using events that have already occurred. This is because, in order to find out what captain evaluation metric is superior, I would need to find out the best FPL team so far, and determine if the combined cost of players exceeds our given budget of 100 million. If it does, it implies that some sort of efficiency is required when spending, and I would tilt towards looking for the best performers in PPV (1 premium teams). And conversely, if the optimal team was cheaper than 100million, it would imply that our aim should simply be to choose the top scoring players, in which case I would look for the premium combination that prioritizes points (2 premium teams).
That’s my reasoning, but I do not know the optimal FPL team. At least, not exactly. However, I can approximate by looking at the highest scoring players as of Gameweek 5. This team structure does not accommodate for transfers, which should push up the team’s total score, but you will soon find out why I did not deem that a problem. So, without further-ado, here is your (almost) optimal FPL team of the season so far:
The shaded-out names are generic bench options, and they can be any of the cheapest options in their position. What is important are the players starting, and more importantly their cost and points score. Assuming a perfect captain rotation between Salah and Haaland, this team scored 446 points as of GW5, which placed them at first place in the Fantasy Premier League! Yes, first place without using any chips! And comfortably so too, so comfortably that even midway through game week 6, 446 points would keep the team nestled in 118th place. This is why I did not bother to find a truly optimal FPL team so far; I decided that this team would be good enough to reach a conclusion using it.
So, what is the cost of this team? It may shock you that, including substitutes, this team’s cost as at when the game started rounded at 94 million. That is 6 million (or a Reece James/Martinelli) less than the starting budget of 100million!
What this implies is that, as I mentioned earlier, these first five game weeks have been a race to maximize points, and not player value, as the budget limit of 100 million is more than enough to obtain some almost-optimal teams. In other words, as of right now, though players like Salah aren’t worth their cost, they were still optimal choices in our FPL teams simply because they were some of the highest points scorers in the league (and as we saw, Salah was a member of the team I used earlier). This also supports the idea that 2 premium players in a team was optimal between these 5 game weeks (in the team I used earlier, we also noticed 2 premiums: Haaland and Salah).
However, I want to point out that throughout this analysis, I used data only from game weeks 1 to 5. What about future game weeks? This report has in no way been predictive, and should not be used for that purpose. However, I do have some thoughts on what to expect in the future. Expensive FPL players are expensive for a reason: they are expected to score the most FPL points. I would be very surprised if the optimal FPL team come end of season looked anything like it does now. In fact, I expect more expensive FPL options to score as their price potential suggests, and when that happens, we would need to pay much more attention to a player’s PPV. Maybe then, 1 premium teams would be more ‘meta’. However, for now I’m already planning my first wildcard, with Haaland and de Bruyne the first names on the list.
I write more articles like this as threads on Twitter @NumbersFPL. You should follow me there!
BallerzBantz 