Think Different: An FPL Tale Part II

Part two of this series will discuss a key metric I found a few years back, but until now I really haven’t used it during the season. In fact, I didn’t use it last year when building my squad ahead of the first game week. Like many metrics, you can take away from these numbers what you want. It’s not the end all, be all. Like many metrics used in FPL, there are some shortcomings, but I do like what it does provide when it comes to players who got it done all season long. In fact, it might hold a bit more value when looking at a 3-5 game run a potential replacement player. Just depends on how plugged into the FPL game you want to be. However, the more time you dedicate to the game doesn’t mean you will finish any better.

KEY METRIC – VAPM

A few years ago on Reddit  I came across an interesting metric called Value Added Per Million. It takes points scored per match (PPM) and subtracts two points from the PPM, for appearance points. As the original poster states, “This also diminishes the value of substitutes who score 1 point for a cameo.” Then again many statistics used in the FPL game have shortcomings, just find the ones you rely on and use them. I am not here to argue validity of metrics individuals managers use, as I don’t waste time with xG or xA. Does that make the metric wrong? Nope.

The key metric using VAPM comes in final number achieved when you take into account minutes played (games), total points and price. Based on calculations, players that achieve at or above 0.35 VAPM are those we should be targeting when it comes to our starting XI. This number is the result of doing the arithmetic of how many total points you hope to score on the season. The winner of the 2019/2020 FPL game scored 2557 points, but for our example we will use a more “realistic” figure.

Delving further into the numbers, one can come across some problems with the metric that make you rethink the situation. Take Joe Gomez as an example, the Liverpool defender, priced at £5.3 scored 93 points in 2004 minutes. He played in 28 matches, but started 22 of them. Originally, I was using the minutes played, divided by 90 to give me games played. Using this method, and his 2004 minutes equates to 22.3 games played, well short of his 26 matches he played in. This method gives him a 0.41 VAPM, but that isn’t realistic.

Instead I’ve calculated VAPM using matches played, but the minutes no longer figure in this final number. Only when I compare VAPM to PP90, do I take minutes played into account. Using 28 games Gomez has a more realistic VAPM of 0.25, which comes in well below the 0.41 if using the previous method based off minutes played.

Ahead of the season, 2300 points would have made you competitive with captain selections scoring 500 of those points, which is realistic, you would need to score 1800 points to achieve your goal. There are 38 game weeks, which is rounded up to 40 to simulate “double game weeks and the extra points from chip use.” Again, I stress no metric is perfect, take from this what you want or nothing at all. The 1800 points are divided by 40 to give you 45 (points). Subtract 22 (starting XI playing 90 minutes) from 45 for a final score of 23 points, which is what is required of your players each week.

Partial list of all players listed by VAPM.

Given that  £18  million (I had £17 million to start last season) is a typical amount for four bench players, your starting XI needs to average approximately 0.35 points per million invested. Based on my calculations there were 11 goalkeepers, 9 defenders, 14 defenders and 7 forwards who finished at or above 0.35 VAPM. That’s quite a small pool of players to be looking at when making decisions. Its nothing more than a guideline to help with player selections.

DEFENDERS

Sadly, I can only claim Virgil van Dijk in my GW1 starting XI. After activating my Wild Card ahead of GW3, I added John Lundstram  but dropped van Dijk.  Throughout the season, I owned everyone in this group listed below, except John Egan and Ricardo Pereira. As mentioned in part one of this series, “Trent Alexander-Arnold and Andrew Robertson, had a combined score of 391 points on the season.” My defenders as a group scored only 457 points on the season. Pretty amazing to consider those two defenders could have made up a bulk of your point. More surprising was the defense play of Sheffield United, as John Egan and George Baldock,  ended up above 0.35 VAPM!

There were and are players you could own over the course of 38 weeks who could score points in spurts, but not fulfill that 0.35 VAPM. Serge Aurier (0.34) and Kyle Walker (0.34), Enda Stevens (0.33), Lewis Dunk (0.33), James Tarkowski (0.32) and Çaglar Söyüncü (0.31) all came up short. Looking at the results and defenders I played over the season, my failure started ahead of the first game week. I had the right representation, two players from Liverpool and Man City, but only consistent starter, van Dijk (Joe, Gomez, Oleksandr Zinchenko and John Stones were my other starters).

These players were compiled from the top 25 defenders based off total points for the 2019-2020 Premier League season. Many of these players are worthwhile to own. This metric allows you to better plan player transfers when you consider a £4.5 defender needs just 11 points to score a VAPM of 0.37 or one clean sheet over that period. Any other points that players scores during that three game stretch is a bonus and only decreases the total points needed to achieve your season goal.

This chart provides points needed to meet or exceed the 0.35 baseline put forth in the first article. The higher priced defenders need to pull their weight and score more points over a 3-game period. Using Alexander-Arnold to start the season, he scored 11 points in 270 minutes, but at £7.0 his VAPM was only 0.24.

Using FBref.com, I’ve compiled all the VAPM statistics for each player in the league for the upcoming season (excluding the promoted teams & relegated teams). Depending on how FPL prices defenders, there is some good logic behind starting Robertson and TAA, even without having to consult any key metrics. The challenge will become to round out your starting defensive unit. While this list is an endpoint, it’s also a starting point. While we haven’t seen all the transfers and player movement shake out, this metric and list could provide you a starting point when building your team.

Next up in the series, we will look at the midfielders and see what comparisons and conclusions we can draw when we look at them, based on their value and performance last season. Do we see prices continuing to rise? Do some defenders unjustly get moved to be midfielders?

Leave a Reply

Your email address will not be published. Required fields are marked *