Talisman Theory – March 2019 update

Before you get stuck in: if you’ve not read them before, I recommend reading our summer articles that introduced the Talisman Theory concept and then tracked how Talismen emerged last season to get up to speed.

Note: Joe from FFS has his own way of uncovering Talismen by using stats from their member section. If you are an FFS member, you may access this here

Huge thanks to Mitchell Stirling (@MitchellSt on twitter) for running the data for us.

The talisman in FPL – who more often than not is the key man at a club outside of the top 6 – represents true value. They are the focal point of any team’s returns in the game of FPL, which is often discrete from the football stats. This means, for example, whilst one player may sparkle to the eye it’s actually another with end product who is our FPL hero.

Our talisman data is generated through focusing on non-appearance points as a metric, which are the “individual” points generated by players rather than the more baseline ones. This therefore includes goals, assists, saves, pen saves and bonus. We exclude negative actions in the same way – pen misses, yellows, reds etc – so we can focus on positive contributions only. We have taken the decision to omit clean sheets – this is because this is more of a “team” award of points than an individual one. Defenders performing particularly well are compensated with bonus, but this idea of finding Talismen is more of an attacking than defensive concept. This also means points deducted for defences conceding goals are not considered.

With the double gameweeks to come, it’s worth looking at this in detail.

Data correct up until Gameweek 30.

Findings in team data

Let’s check in on the team-level data first and see which Premier League team has scored the most FPL “non appearance” points.

Back in the November update, City lead the way, followed by Chelsea then Arsenal. City are still top of that metric, but things have changed elsewhere:

Man City continue to lead the way with 687 non appearance points scored, in FPL; astonishingly, this is more than 5 times the output of bottom club Huddersfield. They’re 88 up on Liverpool (609),who nonetheless have travelled from 4th to 2nd from the November update – they’ll no doubt be happier to look at the league position, too.

It’s a 60 point gap between the Reds and Arsenal (549), with the Gunners’ North London rivals Spurs (528) very close behind them. It’s then Man United (478) and Chelsea (473) in the sub 500 club, but we can see a distinct top 6 – and indeed a distinct top 4 – vibe emerging when we look at the data this way as they’re all in a similar range.

The momentum is currently with United since OGS took over: they were 8th back in November.

Between the top 6 and the rest a yawning gap is seen – 103 points separates United from Watford (370), who spearhead the “best of the rest” followed closely by 6 clubs in close range of each other. These are the usual suspects for mid table mediocrity like West Ham, Palace and Bournemouth, although they’ve been joined this year by the Wolves juggernaut – they’re performing very well in 9th for a newly promoted (do they even count?) team. Leicester (309) may well see an advance on their current below par standing of 13th for non appearance points scored after managerial change.

Burnley, Southampton, Brighton and Newcastle are shown to be no great shakes, with only the middle duo of real interest from this point onwards due to double game week scheduling. Beyond Cardiff, their fellow relegation candidates Fulham and Huddersfield are surely doomed.

Findings in the player data

There are two ways of looking at this:

1/ Looking at the total non appearance points scored by one player

2/ Looking at the proportion of non appearance points scored by one player in relation to his team’s total non appearance points.

Let’s do 1 first, looking at the top 20:

It’s a wonder Raheem Sterling has somehow contrived to remain under 20% owned for large swathes of the season – he’s been the most effective FPL player this campaign, scoring 135 non appearance points. Mo Salah is hot on his heels though, just 7 behind – probably attributable to the Christmas pen spree. Hazard is in third, 6 behind Salah and 14 behind Raz, perhaps showing his quality endures on a season level despite being a frustrating asset to own.

Wilson may have some scratching heads as he’s level with Firmino, but it should be reminded that this is on a season level: remembering his opening blitz of 5 goal and 7 assists in the first 10 games he became highly prized, with over 30% ownership at one point.

Pogba, Rashford and Martial have scored fewer points than you’d expect but this is due to their slow start: only Martial had a decent beginning to the season, with Pog/Rash’s spurt over Xmas propelling them into contention.

You’d expect Rashford to overtake the likes of Richarlison and Felipe Anderson, who are just hanging on in the top 20.

This means Sterling and Salah have both out performed bottom club Huddersfield’s output on their own. Combined, they outscore each of the bottom 6 clubs according to our talisman metric.

So, in some ways, these guys are talismen in a sense that if FPL points are scored, then these men are likeliest to get them so far this season.

We’re now at the point of the season where, perhaps, the talismen are more easily identifiable.

The true beauty of our talisman model helps us to identify those individual players who are likely to provide value in an FPL sense – for example Pascal Gross (5.5m last year) who scored a huge proportion of his teams non-appearance points last campaign.

Let’s see how proportions of points scored looks:

Mitchell note: numbers may look slightly flatter: most of that is taking out the negative points for conceding goals which I had to else you end up with a player scoring 25 points out of -12 which means it makes little sense.

Forced into looking at it this way by Huddersfield essentially.

HUD have scored 122 individual points this season. Of which Jonas L has 26. If I hadn’t added the lost goals conceded back on, he’d have scored 26 out of 19. HUD have lost 103 points due to conceding 2+ goals!

Tom note: we may recalibrate older data to bring it back in line with this more streamlined model in future.

Perhaps unsurprisingly given his consistent quality, Jimenez emerges as the top talisman. Though it may look close between him and Mitrovic, the context is key: the Mexican has comfortably outscored the Serb by 96 points to 55.

Mitrovic is a good case in point for talisman theory: we are not saying buy Mitrovic right now – but if Fulham (or a team of that type) do show signs of getting in form, or have favourable fixtures, it may be worth considering them. In that same vein, Jimenez in the Wolves team is Talisman King.

Worryingly for Cardiff and Huddersfield, their goalkeepers Etheridge and Lossl are their top non appearance points scorers by dint of saves and penalty saves. Not a great situation as it shows a lack of points elsewhere in the whipping boys.

Hazard is again favoured by the model as a top 6 player whom operates as the attacking fulcrum for his side – his 25.6% score sits him in third. Though that % seems low, in terms of raw points it’s pretty huge – his 121 attacking points up until GW30 were comfortably more than double second placed Pedro (54).

On that note of ‘top 6’ talismen, it’s worth noting that only Salah also figures with Hazard representing that group in the top 10, with the remaining 8 scoring the highest proportion of non appearance points for the better sides in the bottom half. This is unsurprising, as points are likelier to be spread through the team for these clubs despite a nominal “star player”. Only 12 points separate Lacazette from Aubameyang, for example, with the Gabonese making it into the top 20.

The three teams represented by the bottom three are of interest to the double gameweeks, as Milivojevic, Ings (still) and Deulofeu/Deeney scoring lowest. Palace, Southampton and Watford all benefits from these fixtures but it’s worth noting that points have been more spread around the team for these clubs. Also worth noting how well Murray (11th, 20% of Brighton) is doing, though there would be doubt over how often he can play.

Conclusion

So there we have it. Jimenez reigns supreme currently as the Talisman King, though Raheem Sterling, Mo Salah and Eden Hazard are duking it out for top non-appearance points scorer overall.

It’ll be interesting to see if Jimenez and Sterling can preserve their leads over the rest of the season but it looks more than possible.

We’ll update this at the end of the season.

In terms of the upcoming doubles, it may be worth just highlighting the talismen (and also the #2s) for all teams so you can see who you might want to fit in as first port of call if making a team abiding by this metric:

We’ll discuss this more on the pod in the near future, but for now hopefully this is useful in giving further statistical evidence behind why Talismen are important in FPL. It’ll be interesting to see how these guys do over the final few action-packed Gameweeks.