Value 2018-2019
This article aims to revisit some of the key points discussed in previous articles last season (Season review, Mid-season review), but with fresh eyes and new starting prices making the foundation for analysis. This study contains numbers from 223 currently active players from the game, where all of them have taken part in at least 1 previous season. As this study builds on numbers from previous seasons, I’ve not included promoted sides and players signed from different leagues. The aim is to evaluate player prices to see “where the value is” amongst positions and “price brackets”.
Adding a second formula
In the comments-section on FFS there was some good discussions and Hot Topics regarding value. I tried to soak up some of the knowledge from the regulars on the site participating in the discussions. With those discussions in mind I’ve decided to include an alternate way of looking at value. Although Points Per Match Per Million (PPM) is a good metric and one I’ve personally championed several times, it can be viewed as discriminating against the pricier players of the game. To avoid sounding like a broken record, whilst also broadening my own horizons and hopefully adding some further insight, I’ve included VAPM (Value added per million) in this study.
This Reddit-article offer some good insight on VAPM.
To swiftly summarize what this article and its comments suggest, basically, because the game is constructed in such a way that all players are afforded 2pts due to merely playing at least 60mins, those points can largely be ignored when assessing value. Therefore, there’s reasons to argue that VAPM’s formula (total points–appearance points)/matches/price is a more accurate measure of value rather than points/matches/price. This was well pointed out by Roosta when he compared Sergio Aguero and Ayoze Perez, both in terms of VAPM and PPM.
In 17/18 Ayoze Perez (124pts) had a PPM of 0,63 and a VAPM of 0,26.
In 17/18 Aguero (169pts) had a PPM of 0,59 and a VAPM of 0,41.
Because appearance points alone counts for such a high percentage of the total points accumulated by the cheaper players, an unproductive period won’t make the same kind of “dent” in their value compared to expensive players. Excluding appearance points might therefore give a more nuanced evaluation of player value than PPM.
Reviewing the results
The table below shows the players from my sample that top the respective categories. Based on new player prices, VAPM 17/18 and PPM 17/18 shows the players who are the most valuable based on their points last season (players with less than 19 matches have not been included). VAPM history and PPM History shows the same, but uses data collected from all of their seasons in the league (players with less than 2 seasons have been excluded).
The fact that 9/20 players in PPM/VAPM history are the same, and 16 in PPM/VAPM 17/18, probably show that the two formulas to some extent appreciates the same players. But at the same time it clearly shows that PPM gives a higher rating to the lesser-priced players. Earlier, especially in comments to my articles, I’ve discussed why there’s still a place for those pricey players who have next to zero chance of defending their price with regards to PPM. I’ve usually pointed to captaincy and “spending your budget” as the main motivations for selecting them. However, for some of them, maybe VAPM alone, can more than justify it. The table below might suggest just this.
Applying the knowledge
The columns have been colour-coded to highlight how last season (17/18) or previous seasons (history) suggests where value is with regards to PPM and VAPM this season.
Takeaways:
– Both PPM and VAPM suggest that goalies will offer great value at all price points this season
– While PPM suggests that cheap defenders are a good value bracket, VAPM does not agree. Expensive defenders do however offer good value using both formulas.
– It’s when looking at mids and attackers that the formulas differ the most. While both seem to agree that cheap mids and attackers will be a mediocre source of points, both midpriced mids, premium mids and premium attackers look far better when using VAPM rather than PPM. VAPM history actually goes as far as suggesting that premium mids are the most valuable players within the game.
- As those opposed to using PPM have suggested, my studies do suggest that PPM does discriminate towards the pricier players of the game. But as of yet I’m not quite willing to let go of this metric. I’ll get to why, but first lets look at the example in the mentioned Reddit-article.
Reddit-user HappyGrinch suggested:
- 2,300 would be a superb final score, probably landing you around the top 10k.
- If we make a reasonable assumption of 300 captaincy points, that means you need to achieve 2,000 points without captaincy.
- There are 38 game weeks, but to approximate for double game weeks and the extra points from chip use, let’s round the denominator to 40.
- 2,000 divided by 40 equals 50 points per week. 22 points are scored by all of your players playing 60 minutes. Therefore, your team needs to score 28 additional points.
- The average squad has about £18M on the bench, so your starting XI needs to return around 0.35 points per £1M invested.
His thoughts on this matter certainly intrigues me. His point on 2300pts being some sort of a hallmark has been the case for most of the recent seasons. The 300 captain points sounds about right. And the 40 gameweeks and the average bench thinking sounds fairly accurate.
However, his thinking can also be used with regards to PPM. For this scenario let’s remove captaincy points and go with the 40 gameweeks. 50 points per week divided by 82 = 0,61, meaning your squad would need to average 0,61PPM to achieve 2300 points.
The thing is: I ran both PPM and VAPM through the crop of players in my studies after last season and decided to use the 0,35VAPM and 0,61PPM-criterias to see how many players who fit the mould. As this comment shows, after trimming the pool of players, by removing goalies and the players who played less than 25 games, the pool finally consisted of:
VAPM: 12def, 14mids, 1striker.
PPM: 21defs, 19mids, 2strikers.
What I suggested then, and as I will suggest again, is that I believe this shows that both ways of evaluating players are good. Where PPM might be discriminative towards pricier players, VAPM looks to be a bit discriminative towards cheaper players. While VAPM might be more accurate, PPM gives you a larger pool to make the selections needed to hit your target. As no season is similar, one needs to be flexible and able to adapt. Valuable players might become available in all positions and all brackets, and what happened last season doesn’t necessarily reflect what will happen this season. Regardless if you prefer PPM, VAPM, or a different way of looking at things, that highlighted part is the most important one.
Players that stand out
After stating that what’s happened in the past doesn’t accurately reflect what’s going to happen next, what’s better than using past statistics to highlight some players that might stand out? Here’s a list of some of the players who’s PPM and VAPM, 17/18 or history, suggest might be valuable picks this season.
Strikers:
– Aubameyang and Aguero: Both players had a VAPM of 0,43 and PPM of 0,61 based on their new prices. This suggests Auba should be terrific pick if he maintains the form from last season. Both last seasons numbers and VAPM history (0,36) suggests Aguero looks like a really solid pick as long as rotations and injury don’t hamper to much of his minutes.
– Wood: The 24 games he played last season is a small sample size. But his points production last season suggests that Wood, at 6,5, has a VAPM of 0,36 and a PPM of 0,67.
– Austin: Been a peripheral figure in FPL after his massive debut season with QPR. Priced at 6.0, Austin’s historical figures suggests a VAPM of 0,34 and a PPM of 0,67.
MIDFIELDERS:
– Salah: If he is to replicate his numbers, he actually holds a very healthy VAPM of 0,49 and PPM of 0,65 based on last seasons numbers. He probably won*t, but should offer some value, even beyond captaincy, as long as he reaches about 240pts.
– Alli and Sanchez: Although both had somewhat unimpressive seasons last year, their price drop has given them a very positive value this season. Their VAPM history suggest a value of 0,37 for both.
– KDB and Eriksen: As two of the most consistent players within the game, Eriksen has sported a points per match averages of 5-6pts in 4/5 seasons with Spurs. Similarly De Bruyne has done the same his 3 seasons with City. The two players looks good value as set and forget options in midfield due to thei consistency. Some might stay clear of them due to not being considered good captaincy options, but their historical VAPM of 0,34 suggests that they might not have to be.
– Pogba: With the addition of Fred, there’s speculations that Pogba might finally be less restrained in the attacking parts of the pitch. If this is the case, Pogba might become one of the most valuable players within the game this season. At 8.0m, he’s priced the same as last season, and although some might not be impressed with Pogba’s numbers last season, he actually sported a VAPM of 0,39 and a PPM of 0,64.
DEFENDERS:
-Moses, Alonso and Kolasinac: I’ve grouped these 3 together due to their attacking nature. None of them are currently safe picks, but if they should become such, it seems reasonable to expect a VAPM greater than 0,4 and a PPM greater than 0,75 at their current prices.
– Hector Bellerin: Most of the 5.5 arsenal assets looks good value with this seasons prices. I’ve decided to highlight Bellerin’s numbers, as they are the most impressive ones. If Arsenal performs along the lines of how they’ve done the last couple of seasons, Belerin should offer good value with his VAPM history suggesting 0,37 and PPM history 0,73. If they should better their results under Emery, Bellerin might become the standout option he was in 15/16.
– All Liverpool assets priced 5.0: As it stands, LFC looks likely to be starting each GW with at least one CB and one RB priced 5.0. Barring new prices, it will be Lovren, Matip, Gomez, TAA and Clyne battling for the two slots. If any of these players become “nailed”, they should offer great value, somewhere in the range of 0,35-0,4 VAPM/0,75-0,85PPM.
– Azpilicueta: To get consistency, one should look to one of the priciest players of the game. Although a change of manager brings some questions, Azpi looks like he’s worth the outlay to get into one of the best defensive sides of the last couple of seasons. 17/18 suggests a VAPM of 0,42 and a PPM of 0,73, and history suggests a VAPM of 0,33 and a PPM of 0,63. Having missed only 2 games in the last 3 seasons, Azpi currently looks like the best option amongst the priciest defenders.
GOALKEEPERS:
Honorable mentions to Karius, Heaton and Pope. With the likely addition of Allison, and uncertainty regarding Burnley number 1, I’ve not included them, even though their stats look brilliant.
Ederson and DDG: The two goalies compete with each other based on last seasons numbers depending on whether you use VAPM or PPM. The fact that Ederson’s season looks more repeatable than DDG’s probably suggests that he could offer more value than his Spanish counterpart as a set and forget option.
Ryan and Fabianski: Two of the standout budget options last season have been handed a budget price this season as well. If they should replicate some of that form, a total of 140pts should give them a VAPM of about 0,4 and a PPM greater than 0,8. Rotating two 4,5’s could arguably produce even greater value, but as Fabianski’s last 4 seasons have proven, the 4,5 goalies come with greater volatility with regards to steady returns.
5 years, 9 months ago
Great article, thanks for this. Really interesting stuff.
Definitely confirms some of my picks. Who are you looking to bring in? Wood looks to be my cheapest forward right now. I think I'm considering Austin again now too