Introducing the FGPE 100…
A way to rank value in Fantasy Premier League every week and a quote from Moneyball...
“Your goal shouldn’t be to buy players. Your goal should be to buy wins. And in order to buy wins, you need to buy runs.”
That is one of the key quotes in the book-turned-film Moneyball, the true story about how the Oakland Athletics revolutionised baseball through their use of data. I was reminded of it recently (and the great scene in the film with Brad Pitt and Jonah Hill, video above) by the Ankler newsletter, which covers the entertainment industry. It used the quote to illustrate how Netflix's most valuable programmes were not blockbusters but regular US series like Criminal Minds. You can read the full piece here
In Fantasy Premier League this quote is even more relevant, albeit with a minor alteration. Your goal shouldn’t be to buy players. Your goal should be to buy points.
We have written a lot in Fantasy Gameweek already about the importance of value. We have compared the example of buying shares in a company to buying a player for your FPL team. You shouldn’t buy shares in a company just because you like it, you should like the price too. For instance, you may think Tesla is going to become the biggest car maker in the world, but if its shares are already priced on the assumption that it is going to be bigger than all other car companies combined, then you need to rethink whether to buy them.
It’s the same with FPL players. You may think Erling Haaland is going to be the best striker in the Premier League this season. But if his £11.5 million price means that he needs to score 30+ goals to offer a greater return than cheaper players then you may need to rethink whether to buy him. Everyone starts with £100 million, the winner is the one who turns that into the most points.
There are numerous quirks in the FPL rules and scoring system that we need to take into account when considering value too. This includes that forwards score only four points for a goal, midfielders get five and defenders get six. Midfielders also get an extra point if their teams keep a clean sheet, forwards do not. Furthermore, last season’s top bonus point scorers were led by midfielders and defenders, especially attacking full-backs - the system seems to favour their contributions.
This thinking is behind our ZeroForward strategy, our look at mispriced players and our analysis of whether Mo Salah should be in your team.
Now we want to take this look at value further. We want to offer you a way to assess the value of players in a similar way to how you can assess the value of company share prices through a price-earnings ratio (in simple terms, how many multiples of a company’s annual profits its share price is) or dividend yield (how the regular payout to shareholders measures against the share price). We are going to call this index the FGPE 100: the Fantasy Gameweek Player Evaluator 100. We think it will help you make decisions about your team for the next Gameweek by offering you a better understanding of what each player offers.
If you want to skip ahead to the index now, please go ahead. It is below. The players are colour-coded by position with the most valuable first. Goalkeepers are grey, defenders are blue, midfielders are yellow and forwards are red.
First though, a quick explainer of what the FGPE 100 shows and why it matters.
How do you track how valuable a player is in FPL? In simple terms it is how many points a player delivers per £1 million. The FGPE 100 tracks this but also more. That is because points-per-£1 million does not take into account other factors. How do you differentiate between a player who costs £10 million and delivers 10 points and a player who costs £5 million and delivers 5 points? How do you measure opportunity cost - the risk that your investment elsewhere (another position) could be more profitable? For example, how do you take account for the fact the 10th best defender might offer better points-per-million than the best forward but you don’t care because you have your five defenders already and you HAVE to play at least one forward? How do you also take into account your potential captain for the week, which is simply the player you think will score the most points?
We have designed the FGPE 100 to track all of this. To get the FGPE score for each player we have measured their points per million for the Gameweek but then put that through a multiplier of how their score compared with others in their position in that Gameweek. So, for example, the top scorer in a position has their points-per-million for that week multiplied by 1 and someone who scores half that has a multiplier of 0.5. This puts their performance in perspective and proportion. The overall FGPE score for each player over the season will be the average of their weekly score.
We think the FGPE system, including the multiplier, is a much more powerful way of looking at value than simply points-per-million. Take Haaland, for example. On a pure points-per-million-basis he ranked just 37th for Gameweek 1. This was behind 19 defenders and any goalkeeper that kept a clean sheet. But is that really an accurate reflection of what's going on in FPL given that we can’t select 19 defenders and HAVE to pick at least one forward? No. Haaland was the joint-top scoring forward in Gameweek 1, only four other forwards scored even half as many points as he did. Most teams who had Haaland in Gameweek 1 did ok. The global number 1 is top after triple-captaining him (!). Under the FGPE index, which reflects that Haaland was the joint-top scoring forward, his ranking goes up to 14th.
Anyway, with that explainer out of the way, let’s get on to the FGPE 100 for Gameweek 1. A few things to note for the first set of rankings:
-Only players who scored more than 2 points are included for this week.
-This ranking will get more and more useful as the weeks go on. This is only one Gameweek of data, so treat it with caution.
-We have not built future projections and predictions into this yet but will look at ways to refine the model as the season goes on. Please get in touch with any suggestions.
-We will track player price changes. A player’s weekly FGPE score will be based on their price that Gameweek. However, we will refer back to the original price where relevant - for example, if many teams own the player at their original price.
-The FGPE score is just that, a score. Use it to compare players or groups of players. Don’t use it to say that Fabian Schar should score 3.3 times his value every week. He won’t. And that’s not how this index works.
-Harry Kane does not make the list yet as he only scored 2 points. For those interested, however, his FGPE score would have been 0.03 for Gameweek 1. For those who had him in your team, his return in the first match was fairly ruinous…
-It is called the FGPE 100 because we will eventually include the top 100 players in it every week.
With that, here you go. The index is below as an image and also attached as a pdf if you want it in a different format. If you want the raw data and spreadsheet, please get in touch. Beneath the chart there are some quick thoughts on some of the players who stand out…
-Schar leads after his goal, clean sheet and three bonus points for Newcastle. That won’t happen every week, but this shows the value of attacking defenders, especially those priced at around £4.5 million.
-As Tom mentioned in his “What we learned” piece, Kulusevski’s Gameweek 1 performance was eye-opening and not out-of-line with how he finished last season. He scores 1.41 compared to just 0.14 for Son on the FGPE 100 index. He could be a really valuable player in FPL this season in his own right, but also as a hedge against Kane and Son if you don’t have them.
-Mendy and Ederson being so high up the list shows that the top goalkeepers could be attractive options this season. They could offer the cheapest, most reliable route into the best defences given they may never be rotated.
-Zinchenko and Kelly didn’t even score and look what happened. An assist + clean sheet + three bonus points = big scores and real value. Zinchenko being on set-pieces and part of a solid Arsenal defence could make this sort of performance a regular occurrence.
-Mitrovic - or any of the £6 million to £7 million forwards - offer real value if they are decent. For context, Mitrovic would still have been in the top 50 in this list, and the sixth-placed forward, if he had just scored one goal and not two.
-I think there are two ways to look at Haaland’s ranking. Firstly, he does offer value if he keeps this up. However, the other side of this is that even after scoring twice and getting three bonus points he only ranks 14th. This suggests there is a way to build a team without him. The FGPE system will reward consistency, however, so I expect him to climb this list. Pascal Gross isn’t going to score twice every week. Haaland might…
Thanks for reading. Please get in touch with any suggestions or thoughts about this data and how you think we could improve it. The FGPE 100 will be back next week with more analysis on what it shows and less on how it does it…
If you enjoy Fantasy Gameweek and find it useful, please spread the word and share it…