In his debut blog for us, our new resident “statto” Ewen (@ewen_) does some magic with last year’s Fantasy Premier League (FPL) numbers to look at the ceilings and captain hauls. He’ll be in the “free role” for us this year, looking into issues as and when they arise / take his fancy.
A familiar pattern emerges at the precipice of a new FPL campaign. As soon as die-hards finally get their mitts on the game, a cycle through possible team compositions (and crippling self-doubt) looms. One permutation that can get a lot of traction is the prospect of a team firmly weighted in favour of an all-action, hard-as-nails defence for the ages (think top 4 candidate, penalty taking, perma-swashbuckling wingbacks). Pepper the frontline with hipsters favourites and a mightily interesting set-up is born.
Why might this not be the best strategy? Let’s search for an incompassioned defence of the star-studded attack. One that doesn’t just laud exotic new forwards for being mysterious, but crunches the numbers to see if you might regret skimping on attackers when it matters most. Or not.
The Importance of Being Captain
Something that’s often under-sold during such debates is the role of the captaincy, which has an undeniable impact to a players’ scoring potential (hint: it’s doubled).
The captaincy decision can be influenced by all manner of things, right down to a player’s birthday/arrival of baby (those factors are slightly out of scope for today). Going back to basics, it matters where a player spends their time on the pitch i.e. their position. Here’s why.
The chart shows how position affected players’ theoretical captain hauls (gameweek points doubled), controlling for appearances of less than 60 minutes (which unfairly impacts on defenders and goalkeepers) and games against the top seven (which tend to be avoided, captaincy-wise). ‘Premium’ players, from here on in, are defined as in the upper two quintiles for price i.e. they are priced in the top 40% of players in their position (this is an attempt at a fair and consistent representation across positions – sometimes anecdotal definitions of ‘premium’ players end up comparing roughly the top 10% of defenders against the top 25% of midfielders, which seems harsh).
Notice how all positions bunch up a lot around the 4-6 point mark, with this being their highest ‘peak’ – this tells us that in most gameweeks, against ‘non-premium’ opposition, even premium players are scoring just 2-3 points regardless of position (something to try and remember, next time your man blanks). This makes sense, intuitively – in any game, usually only a few players on each team do anything of note FPL-wise. Defenders and goalkeepers have a similar second hump around the 12-point mark (this being the clean sheet + captain standard, of course).
The other distinctive feature of this chart is the long tail to the right for forwards and midfielders – this tells us that big points hauls are more likely for attackers, and suggests the ‘ceiling’ of these players is higher. This may seem like obvious stuff, but it’s vital to demonstrate and understand such phenomena empirically, rather than to take FPL tropes as a given.
Ceiling, What Ceiling?
Speaking of FPL tropes, everyone knows that any captaincy decision worth it’s salt needs vindication via double digit performances. How many times did this happen last season, by position? This might help explain what’s driving the attacker’s long tail, above.
It doesn’t make sense to look at a pure count, because there’s so many ruddy midfielders in the game compared to, say, goalies. Instead we’ve considered the rate at which premium players demonstrate such returns – a glance at the above indicates that, going from front to back on the pitch, double digit returns got progressively less likely last season. To be precise, premium forwards got into double figures more than four times as often as premium defenders did.
Just who was behind this? At a Gameweek level, rather than season, we can identify the players who were breaking this ceiling most often.
A couple of things are very noticeable – one is how much Kane and Lukaku were responsible for dragging
the premium forwards up (notable party poopers included Rooney and Slimani, managing just one each). Still, as mentioned earlier, this kind of perspective is biased in a sense that midfielders will likely dominate simply because of the sheer number of them. Also, just counting big displays will overlook those who were accumulating impressive responses at a rate of knots, but over a shorter time period.
Hats off to Cesc, who got into double figures more often than not when he was given at least an hour on the field. An honorable mention must also go to Jesus, who was excluded from this analysis based on his lack of minutes meaning a shaky sample – his rate of double digits returns was over 50%, which is startling. Besides this, we’re still looking at a big repping from the midfield. Any way you cut it, it looks like 2016/17 was won in the midfield.
There are many strong arguments and possible season scenarios to support the hypothesis of a squad prioritising value for money and a sexy defensive unit. Nevertheless, it’s important to remember that the captaincy can play havoc with points-per-million metrics.
So, if you do take this path, beware – when a premium forward’s unique ability to outgun players in other positions in singular occurences coincides with well-timed captaincies, that mini-league foe could lay waste to your carefully crafted, value-optimised squad.