Hello and welcome back to FPL 2018/19!
This is the first of our articles for the new season (although we did do a little work over the summer on the continuation of our Prospecting the Prospects series). Much like our future trends series from 2017/18, these articles reflect on what we learnt from the season before and how that might impact FPL managers’ strategies for the new campaign.
Talisman Theory: background
On our podcast towards the end of the season, we began to advance the idea of bang theory.
No, not the one with Sheldon Cooper and co.
This was the notion that, week-to-week, you should have as many players as possible in your lineup that you believe will hit you 7+ points. This was conceived as a way to help populate Wildcard squads and aid decision making when faced with the choice between:
- 1 premium tier player + 1 budget
- 2 “upper-mid” tier players
It’s hugely simplistic, but the idea is packing as many of the players you think could “bang” into your team as possible, meaning you’re more likely to score well week to week.
Taking that one step further, though, we realised towards the end of last season that we could distil our options even more. This is because there tended to be one individual across many sides who was involved in most goals the team scored: think Marko Arnautovic at West Ham, or Wilfried Zaha at Crystal Palace.
This led us to the notion of Talisman Theory, which is the idea that you should always favour buying in this key individual – the Talisman – over any others in a team as they are the most likely source of FPL points.
This concept sounds perfectly intuitive, and could help us in the build up to Gameweek 1 in terms of team composition and player collection.
But this led us to wonder: is Talisman Theory actually backed up by FPL data?
We mined the data from 2017/18 to find out…
We worked with the data taken from last season in a few steps to explore Talisman Theory.
First, we looked at the total points scored by every team, with some surprising results:
Last season, 31 438 points were scored in FPL. Man City scored the most, and Stoke scored the least.
As you can see, Liverpool would actually have been runners up if we look at FPL points over the course of last season, pushing Man Utd and Spurs down to 3rd and 4th.
Swansea outperformed their relegation season in terms of their FPL performance, coming 12th in overall points scored.
Perhaps unsurprisingly given the stodgy performances under Big Sam last season, Everton underperformed in an FPL sense, coming 12th in the table – four places below their final Premier League finish of 8th. Bournemouth also underperformed, coming 16th in the table of points scored but 12th in the league.
Having collated this data, we then moved to step 2 of our analysis, which is a simple look at which player scored the most points from each team:
At the top level, none of this is particularly surprising.
I can’t overstate enough what a great season Mo Salah had: he scored 1% of all FPL points last season with his haul. Simply incredible.
Below him in 2nd-5th, it’s perhaps as we’d expect: Riyad Mahrez (now of Manchester City), Xherdan Shaqiri (now of Liverpool), Pascal Gross (our new cover hero) and Harry Kane were, unsurprisingly, the Talismen for their sides.
What gets interesting is what happens next.
Of the outfield players, players like Arnautovic and Raheem Sterling appearing is not a surprise.
But, from 6th-20th in terms of proportion of points scored per team, seven of those 14 players are goalkeepers. Take Nick Pope, for example – his 153 points was 36 points more than the highest scoring Burnley player (Johann Gudmundson on 117). Jonas Lossl at Huddersfield scored 22 points more than his nearest competitor, Aaron Mooy.
Cesar Azpilicueta is the sole defender featuring in the list – it seems that defenders need to score goals or assists (as Azpi did, in particular a record seven assists for Alvara Morata) to feature ahead of their keeper.
So what’s fuelling this?
One answer may be appearance points. For some teams – especially those where there perhaps wasn’t a standout player like Bournemouth – we might see a bit of data skew due to those more functional points being scored.
This tends to be a slight bone of contention, as for some appearance points are a useful way of understanding how good a player is: someone playing regularly should always be favoured over a riskier pick. However, for the purposes of this analysis, understanding what players added beyond appearance – the “added value” – is what’s important. For example, if a player plays 38/38 games for 90 mins (a Jack Cork, for example), they score 76 points for the season as a base line.
So let’s take those appearance points out and see what we’re left with:
This shifts things round a bit, but we still see those seven keepers there.
However, we begin to enhance the picture of Talismen in teams, and perhaps clarify the issue a bit.
Shaqiri’s role at Stoke becomes even more amplified, with him scoring almost a quarter of their non-appearance points. There are also some movers: Arnautovic, for example, moves from 9th into the top 5 if we remove those appearance points.
What I also think is interesting is the teams down the bottom: Bournemouth and Man United really didn’t have a notable Talisman, with their key men earning under 10% of the points, and, additionally, they were goalkeepers in Asmir Begovic and David De Gea. In both cases, this is very interesting: last year Josh King weighed in with 12% of their overall points for the Cherries. Similarly, Romelu Lukaku bagged 11.9% of Everton’s points in 2016/17, so you’d have reasonably expected him to become the Talisman for United as well. Plainly, this did not happen.
Additionally, in the middle orders, we still see goalkeepers reign supreme; this on one hand underlines the value of goalkeepers in many teams.
But can a defensive player be a true Talisman?
Probably yes – think Charlie Daniels in 2015/16 to some extent.
However, I’d argue that defenders have a way of eliciting points (clean sheets), which are more contingent on the team performing rather than the individual player. Yes, yes, of course the team contribute to attacking points, I get that, but equally those points – in an FPL sense – if more attributable to an individual are what might distinguish a Talisman from a defensive group effort.
With this in mind, we took a final hatchet to our data, this time expunging appearance points, clean sheet points (and therefore defenders/goalkeepers entirely) to see what that tells us:
This final look is perhaps the most derivative, but also the most interesting of all.
It shows the pure attacking points scored by any team, and picks out the one player who contributed most to that total.
This highlights the value certain individuals contribute to certain teams, and our eye immediately flicks to those who are still FPL viable in the top 5 (following moves to big clubs for Mahrez and Shaqiri), who are Gross, Arnautovic,and Mooy.
What interests me about these figures is that they rank above Salah, Kane, Sterling and others.
Although Salah had a fantastic season at Liverpool, and of course is an auto-include in any team of the year retrospective (and, indeed, a great many Gameweek 1 squads for the season to come) the fact remains that he was in a team of players who contributed strongly in terms of attacking returns: Bobby Firmino and, to a lesser extent, Sadio Mane and Phillipe Coutinho in H1 of the season diluted the concentration of points his way. If anything, that highlights what a ridiculous FPL season Salah had.
The same is true of Sterling and also Eden Hazard’s seasons – though they were the men with the highest attacking outputs for their sides, the points were much more spread, making them a Talisman by a short rather than long way.
In the middle reaches, we see a more mixed picture if we try to understand the idea of the attacking Talisman: the now-stateside Wayne Rooney emerging as the top man for Everton, for example, highlights the fact that for some teams there may not be a Talisman in an attacking sense.
However, what that also shows us is that for some teams the reliance on one man is acute.
Pascal Gross in particular emerges as a big hero of ours through this analysis. If any attacking points were scored by Brighton, you had a 2 in 5 chance of him getting points from it. At 5.5m (or 7.0m as is now), that’s great value for money. A Talisman indeed.
- We lack emergence data. This is very hard to get at – if anyone can help, let us know – so we have to look at the season on a total level rather than from a cumulative/progressive standpoint.
- Context is key. Pascal Gross scored 40% of the points in a team that scored 34 goals. We know that, you know that. This data is more about understanding if a Talisman exists, and what we as FPL managers can learn from it.
…so what have we learnt?
If one emerges, the Talisman is the man to own. Pinpointing these individuals in certain teams should probably be our priority in terms of our team composition when it comes to spots on our roster, because they are the likeliest to return you points week in, week out.
The top Talismen are more likely to play for mid-tier clubs, fitting your 3rd – 4th midfield spot. They form the important differential alongside the big hitters that you and many others may own. Getting on these guys quickly, and owning them as a constant source of points, will help you succeed in FPL. For example, although Gross’ Brighton only scored 34 goals, he would’ve elicited you points above appearance on 13 occasions (including one random 3 bonus point performance v Man Utd) – a near enough 1 in 3 delivery-to-games ratio, which is exactly what you want from a player in the 5.5-7.0 range.
The Talisman may not always be who you expect. Staying with Brighton, prior to kick off of last season we had big hopes for Anthony Knockaert. He quickly faded into irrelevance for the Seagulls as Gross took over the mantle of key man. Another example – though more through circumstance of Zaha’s injury – is Luka Milivojevic coming to the fore for Crystal Palace through set piece mastery. Both are instances of where the less likely lad has come through – and I still chide myself for not seeing the value in either example quicker last season.
Talisman theory isn’t as relevant to the bigger clubs. Though Mo Salah was undeniably the Talisman at Liverpool, points were shared far more given the breadth of quality in the frontline that could contribute and cannibalise points from our Egyptian King. Similarly, the proportionally lower points percentages for the Talisman (with defensive points removed) for clubs like Man United, Chelsea and Arsenal should be noted. The spread of points at the bigger clubs shows how we need to take a more case-by-case approach to their key men: where some (like Salah) are perhaps an auto-include due to ownership and/or performance, in other instances (e.g. which City man/men you buy) it’s who you judge will score you the most points in the current context.
Not all clubs have a Talisman. Season on season, this changes – where in 2016/17 Romelu Lukaku scored 11.9% of Everton’s points overall and was the Talisman, last season his replacement as top scorer, Three Lions hero Jordan Pickford, is not. Ditto the attacking points Talisman Rooney. Similarly, where FPL royal Gylfi Sigurdsson attained 14.4% of Swansea’s points in 2016/17 and therefore was the obvious Talisman, last season’s save-driven top scorer Lukasz Fabianski, or their diffident attacking points leader Jordan Ayew probably were not real Talismen.
So we come to the end, and come to a picture which is, to some extent, a mixed bag.
This all opens up all sorts of interesting questions surrounding value in the context of player price, which Nick will tackle in his upcoming articles on price’s relationship with cost.
But the key takeaway, I think, is that Talismen should be high on our agenda when deciding who to buy in at any point.
This seems is particularly true when looking to maximise value for the price you pay in the key position of the 3rd and 4th midfielder, meaning Talisman Theory should be top of mind when you’re selecting whether to go with a key man or gamble on a differential: in the case of something like Gross versus a gamble, always plump for Pascal.