In Pursuit of Science Over Art

There are a number of reasons that Fantasy Football managers transfer players in and out of their squads, among them injuries, recent form, future fixtures, general quality and team performances. These, though, are mostly intangible, based on luck, opinion and clairvoyance. There is, however, another factor that managers should also take into account. Something more scientific. Something, crucially, more predictable. Value.

Let’s start at the beginning. When we pick our teams for Gameweek 1 we get to spend up to 100.0 on our squad of 15. Each player has an assigned value, and that value rises and falls throughout the season. Take Jamie Vardy for example. You would have had to spend 6.0 to select him in Gameweek 1. To bring him in now you have to spend 7.3. Vardy has accrued 1.3 so far this season. For those that selected him in that first Gameweek, the value of their squad has increased 1.3. That might not sound important, but think what you could have done differently if you’d had 111.3 to spend for that first Gameweek ….

Player value increases are alternately added to the player value you see in your team or to your bank. The money in the bank can be spent on upgrading any player in your squad – the value added to your player can, of course, only be realised when you sell him. To see this in practice let’s look again at Jamie Vardy. His value on the open market is 7.3. His transfer value in my squad is 7.2 – not much difference, but then I only brought him in a couple of weeks ago! Dimtri Payet has an open market value of 8.3, but a transfer value in my squad of 7.9. The difference, 0.4, would have gone into my bank.

Player values are recalculated every day, sometime in the middle of the night UK time and always by 0.1. Players may therefore see multiple rises and falls within a single Gameweek. Famously, Anthony Martial’s value rose 0.3 in Gameweek 7 – his value rose on three separate occasions in a seven-day period! It is interesting to note though that his price has now fallen back to 8.0 where it started!

Details on week by week player value changes can be found on their individual information pages, and a summary of price changes for the season or the latest Gameweek are options when filtering player statistics (as you might by form, points etc). You can also filter by total transfers in and out, across all fantasy managers, again for the Gameweek or the season. Vardy, Mahrez and Payet have made the biggest overall gains. Among the biggest fallers are Hazard, Fabregas and Walcott.

The exact formula determining these changes is a closely guarded secret by the games developers, primarily related to the amount of transfers in and out of the player made by fantasy teams (lower value players being more sensitive to price changes than higher value ones). There are, however, websites where it is predicted with a high degree of accuracy. I use fplstatistics.co.uk. This has some important implications, not least because it allows managers to time their transfers to take advantage of likely upcoming price changes, thus benefiting from the 0.1 price rise (or avoiding the 0.1 price loss). A single 0.1 gain/loss isn’t going to make much difference, but as with Dave Brailsford’s British Cycling Team, repeating the process leads to the accumulation of marginal benefits for an overall much larger gain. Buying and selling Martial at the right times would have made your team 0.4.

This is all well and good, but in practice is there a correlation between fantasy team value and fantasy team performance? Well it turns out that there is. The average team value for the top ten teams in our league is 103.0. For the bottom ten teams it’s 99.5. There is an element of self-fulfilling prophecy about this: better performing teams inevitably contain better performing players who are inevitably more likely to rise in value. Still, there is a reasonable correlation between team value and points accrued: for fans of regression analysis the R² value is 0.4024. Not statistically significant but not to be dismissed either.

At the time of going to press the highest value team in FFUK was Tim McEwan’s “£ cantbuyyoustones” at 105.6, his team lying in 39^th position. A table –topper? No. But in contention? Certainly. Imagine how handy that extra 5.6 is going to be with Kun coming back from injury soon!

In short – will a focus on team value alone win you the League? No. But it is clearly an important factor to consider when making transfers and, crucially, when you time those transfers. And it adds a little science to the otherwise dark arts of Fantasy Football Management.