Predicting the Bandwagon – GW32 Preview

In this weekly FPL series, our resident data expert attempts to predict the newest bandwagon according to underlying statistics, formational changes and more. Here, we preview Gameweek 32 of Fantasy Premier League.

by @FPL_FIDGET

Just as FPL becomes more predictable in some ways, it grows less predictable in others: popular midfielders and forwards have been smashing it recently, but clean sheets have vanished once more.

In fact, the top 10 teams for xGC over their last two GWs have managed only 5 clean sheets in 20 as almost all of them have underperformed their xGC (United and Brighton have performed marginally better than their xGC). The worst offenders include Chelsea, who have conceded 6 from 2.02 xGC, Leicester who have conceded 5 from 2.27 xGC and Fulham who have conceded 4 from 1.97 xGC.

While I hope recent Bandwagon recommendations continue to deliver, I do anticipate that top teams will start delivering on clean sheets once more.

In this section, I evaluate two popular bandwagons currently being jumped on for GW32, and whether they can come good for their new owners…

GW32 – Most Popular Transfers In

Bandwagons continue to have great success, reinforcing the recommendations given in these articles over the last few weeks.

Popular current bandwagons Son, Kane, Salah and Trent (230,000 transfers in between them) were explored and recommended last week and these recommendations were vindicated by their combined 3 goals.

Other popular bandwagons, Nacho, Jota and Laca (230,000 transfers in between them) were discussed in the GW30 preview, and have since rewarded managers with 6 goals in their combined 6 games.

Lingard in GW29 and Mount in GW28 have also been explored. Arguably, Mount has proven me wrong with his 2 attacking returns in 4 GWs since, and the major alternative Lingard has shown why he was my preferred pick with 6 returns in 3 GWs.

In fact, only Dallas of the top 10 hasn’t been explored in bandwagon articles recently -although he was recommended in GW25. We turn to him in the next section and also revisit Mount because of his continuing improvement.

Mount continues to impress, with many commenting on how well he has done to increase his goal threat, without sacrificing his creativity or his all-round play. Mount’s excellent season has made him among the first names on the team sheet for Chelsea, which is pretty impressive considering both his age, and the relative wealth (or at least expense) of midfielders in Chelsea’s squad.

While it’s still to be seen if Mount can be one of those special cases who consistently exceed their xG, it is increasingly clear that his performances are underpinned by talent, rather than merely form inspired by the upcoming Euros. What I mean is that, we are expecting Mount’s underlying numbers to continue, or even to continue to improve.

Selected Non-Premium Midfielders Last 6 Matches – Goal Threat Per 90

Non-Premium midfielders are becoming increasingly interesting and it’s worth comparing many of them at once to avoid over-looking high-performers. While Lingard continues to stand out for Goal Threat, the options for the second non-premium slot are rich.

All of the highest scoring non-premiums are significantly over-performing their xG, but none more than Jota (by 134%). Mount’s xG 90 of 0.26 over the last 6 matches is actually the second lowest of these 10 players. Likewise, he has received the second lowest Big Chances Per 90 at 0.21 and has the third fewest Penalty Area Touches per 90 at 1.43.

For those after goal threat, it would be seem that Jota and JWP with 0.32 xG 90 and 0.45 xG 90 respectively may be better shouts than Mount. Both of them have also been receiving significantly higher numbers of Big Chances, although JWP’s penalty surely skews those numbers.

Pepe – like Lingard – is exceeding Mount across the board, and his recent performances may be enough to extend his place in the team, which is always insecure under Arteta (as is the case for most of his squad).

Selected Non-Premium Midfielders Last 6 Matches – Creativity Per 90

Mount performs much better on chance creation, leading the way for xA 90 (0.24), Big Chances created per 90 (0.64) and second best for Chances created per 90 (2.35).

Given how important Big Chances are for leading to assists, it could be argued that Mount is unfortunate not to have more assists. If Big Chances were converted 50% of the time, Mount would have 0.32 Assists per 90 (as opposed to only 0.21). This point is best underscored by exploring how other midfielders’ Big Chances Created have converted into assists.

Greenwood and Raphinha boast assists at half the rate of their Big Chances created, Pereira has assists at a higher rate than half his Big Chances created, and Harrison, Lingard and Pepe have recorded assists as frequently as Big Chances created. Only Saka has been less fortunate than Mount when it comes to his teammates’ finishing.

If goal threat and creativity are considered together, the closest players to Mount’s underlying numbers are Raphinha, Harrison and Pereira, all of whom are over £1.5m cheaper. Raphinha seems a steady hold, Harrison a slightly weaker alternative, and Pereira a potential shrewd differential. Pereira has had many explosive returns this season separated by blanks, but his explosiveness has become more frequent, and the importance of West Brom’s run-in lifts his appeal.

Upcoming Fixtures

Chelsea’s fixtures are hard to gauge because all of their remaining opponents have posted excellent defensive numbers at one point in the season. Admittedly, many of these have since fallen away in the likes of Villa and West Ham, and others such as Brighton, Fulham and Arsenal have often failed to convert their defensive prowess into stopping goals.

That said, I don’t think this is a favourable run-in. A combination of teams fighting for Europe and safety playing against a Chelsea team who have often preferred passing to shooting.

VERDICT- SWERVE

Before looking into the underlying numbers once more, I was expecting to be able to recommend Mount this time. I’m growing to really like him as a player; he is playing better than ever; he has become a much more consistent goal threat; and Chelsea have shown early signs of improving as an attacking side.

However, the reality is that Mount’s numbers rely on over-performance when others are cemented in better underlying numbers. And those players with similar underlying numbers are much cheaper.

What a player Dallas has been for FPL! He has been a delight that has gone largely under the radar, popping up in excellent positions and converting with incredible consistency.

The big question is whether Dallas can continue this form and end the season as the best FPL defender in the game.

Selected Defenders Goal Threat Per 90 – Last 6 Matches

Dallas has been scoring (0.32 per 90) far beyond his xG (0.08 per 90) and regardless of his lack of Big Chances (0 per 90). This would hint that a regression is due, but Dallas has been outscoring his xG all season; he’s shown an incredible propensity for long shots.

In watching Dallas, this facet to his game is readily apparent, and it is backed up by the statistics. He has 4 goals from 22 shots outside the box this season with the shots averaging 0.04 xG per shot (according to Understat), meaning he has scored more than quadruple his xG.

Dallas’ finishing aside, his xG is still good and is only beaten by known goal threats Saiss and Stones and the devastation of a more attacking Coady.

Leeds’ Defensive Numbers Per 90 – GWs1-25 vs GWs26-31

GC: 1.72 vs 1

xGC: 1.64 vs 1.40

Clean Sheets: 0.32 vs 0.17

Big Chances Conceded: 2.5 vs 1.5

Shots on Target Conceded: 5.4 vs 5

Shots in Box Conceded: 9.1 vs 8.5

Although Leeds have played City, Chelsea and West Ham, their defence have continued the improvement that led me to recommend Dallas last time I looked at him in a bandwagon article.

Leeds have conceded fewer goals, shots on target and shots in the box, but most importantly of all, they have conceded only 1.5 Big Chances per 90 in their last 6 matches compared to 2.5 per 90 before that. All this considered, Leeds have been unlucky to keep so few clean sheets recently.

Upcoming Fixtures

After 2 difficult games, Leeds have an excellent run-in. It’s worth noting that Leeds’ performances seem to be irrespective of opponents, which is best underscored by Dallas’ double against City.

VERDICT – HOLD

I wouldn’t be selling Dallas – or indeed any Leeds asset – as I consider Raphinha and Harrison decent options for the run-in.

However, I’d hold off bringing in Dallas for two GWs. I very much doubt that Leeds keep a clean sheet over the next two GWs, and Dallas may even receive negative points for goals conceded. While Dallas will still be a threat, I think there are better uses of a free transfer at the moment.

From GW34, I think Dallas is an excellent option, and may even be the defender to target. I’d recommend coming back to him then and re-assessing whether your defence can offer as much as Dallas for the run-in.

In this section, I look ahead to anticipate who will be the two biggest bandwagons in GW33.

Pulisic’s time at Chelsea has been undermined by odd team selections and by injuries. Under Lampard, Pulisic came in and out of the team, despite his constant goal threat and directness. Under Tuchel, it’s yet to be seen what part Pulisic can play. However, his performances since returning to the side have contributed to Chelsea finally being able to score more than 1 goal in a game under Tuchel.

Pulisic Per 90* – 18/19 vs 19/20 vs 20/21

Goals: 0.39 vs 0.47 vs 0.30

xG: 0.51 vs 0.46 vs 0.39

Shots: 2.65 vs 3.26 vs 2.3

Assists: 0.39 vs 0.21 vs 0.07

xA: 0.29 vs 0.22 vs 0.15

Key Passes: 2.06 vs 1.55 vs 1.41

*stats taken from Understat

In his last year at Dortmund, Pulisic hit some truly impressive numbers (0.51 xG 90 and 0.29 xA 90), which he was largely able to sustain in his first season at Chelsea (0.46 xG 90 and 0.22 xA 90). However, Pulisic has been less impressive this season (0.39 xG 90 and 0.07 xA 90).

I find it useful to look at historic numbers to see if they hint that a player is underperforming, which we have seen from Pulisic under Lampard. Encouragingly, there are signs of a revival under Tuchel.

Pulisic Per 90 – GWs1-19 Under Lampard vs GWs20-31 under Tuchel

Goals: 0.1 vs 0.75

xG: 0.29 vs 0.46

Shots: 1.97 vs 2.74

Shots in Box: 1.38 vs 2

Shots On Target: 0.49 vs 1.5

Big Chances: 0.3 vs 0.75

Penalty Area Touches: 6 vs 5.99

Assists: 0 vs 0.25

xA: 0.09 vs 0.14

Big Chances Created: 0.30 vs 0.25

Chances Created: 1.08 vs 2

The difference between Pulisic under Tuchel and Lampard is marked. Pulisic has registered a much higher goal threat with 0.46 xG 90 compared to 0.29, over a shot more on target per 90 (1.5) and 0.75 Big Chances per 90. Because of Pulisic’s high number of Big Chances, his xG overperformance seems more sustainable than many other top midfield prospects.

Crucially, Pulisic’s xG per 90 under Tuchel is in-keeping with Pulisic’s numbers last season (both 0.46) and in his last season at Dortmund (0.51). Likewise his shots per 90 (2.74) are more similar to previous seasons (3.26 last season and 2.65 the season before).

What this tells me is that Pulisic, and his incredible outlet (that saw him valued at £8.5m this season), are back. Notably, Pulisic has since dropped to £8.2m. In a season when money is more disposable than normal, Pulisic may offer borderline-premium numbers for a very affordable price.

Pulisic’s Minutes

Given how poor Pulisic’s numbers were, it’s incredible that Lampard started Pulisic in his final 7 games at the club. Unlike Lampard, whose team selections often failed to reflect form, Tuchel restricted Pulisic’s minutes at first, forcing Pulisic to fight his way into the team.

Crucially, Pulisic is seeing more minutes recently with an average of 15 minutes in his first 9 games under Tuchel, increased to 67 in his last 3 games under Tuchel.

GWs16-19 vs GWs20-31 Touch Maps

The graphic above shows that in these minutes under Tuchel, he has played on the right, compared to predominately on the left for Lampard. Actually, for Lampard, Pulisic played across the front, but this doesn’t seem to suit Pulisic as we’ll see below.

Pulisic’s Left Foot Key to Goal Threat

Pulisic playing on the right means he can cut in on his left foot, which is a major strength of his as a two-footed player who excels at left-footed shots.

Amazingly, in his career, Pulisic has scored 12 League goals with his left foot compared to 11 with his right despite taking almost half as many shots with his left foot (59 shots with his left foot compared to 107 shots with his right). This is no doubt due to the quality of the chances Pulisic has been able to fashion when cutting in, with his xG per shot on his left foot at 0.2 compared to 0.12 on his right foot (according to Understat).

Mid-price Midfielders Goal Threat Per 90 – Last 6 Matches

Over the last 6 matches, Pulisic has far exceeded all non-premium (and most premium) options for goal threat. Pulisic’s resurgence has been short-lived, so sample sizes should be taken into account. However, boasting an xG 90 of 0.55 is highly impressive and an incredible 1.02 Big Chances Per 90 underscores Pulisic’s threat.

It is much easier to over-perform xG when players are receiving Big Chances rather than an array of low xG shots, which is also encouraging at a time when over-performance is being fretted over. Even a regression to xG would see Pulisic outscore over-performers Jota and Lingard.

Upcoming Fixtures

This image has an empty alt attribute; its file name is image-37.png

As discussed above for Mount, Chelsea’s fixtures are not highly attractive from an attacking point of view.

It is worth noting that Pulisic’s return to the side has seen Chelsea score 6 goals in 3 games after 11 goals in the 9 games that Pulisic hasn’t started. Pulisic has been heavily involved in these goals (having scored 3 himself), and it seems he has done enough to warrant a run in the side, and with him in the side, Chelsea have found scoring easy.

VERDICT – JUMP

This has some disclaimers attached. Pulisic is high-risk because his ownership is low, and his place in the side far from guaranteed.

As much as I advise caution with this recommendation, if you need a risky rank-climber, Pulisic could be your man. If he plays in GW32, Pulisic will top my shopping list.

As a United fan, I’m thrilled to see his season take-off in recent weeks. You may think I’m biased to include him as a potential bandwagon, but I know that United assets are often-overvalued in people’s perception because of the media and also the sheer number of us United fans who are milling around.

Predicting Greenwood to become a bandwagon is different from recommending him, but let’s explore him as an asset first.

Greenwood Per 90 – 19/20 vs 20/21*

Goals: 0.69 vs 0.19

xG: 0.23 vs 0.29

xG performance: +200% vs -34%

Shots: 2.69 vs 3.01

Assists: 0.07 vs 0.13

xA: 0.1 vs 0.13

Key Passes: 0.97 vs 0.77

*Stats from Understat

Greenwood has gone from a major over-performer of xG to an under-performer. It would seem that Greenwood has undergone a terrible period of form, but that would be an unjust evaluation.

Last season, it seemed Greenwood operated solely to get the ball in front of his feet to take a shot at goal – and my God could he shoot. This season, Greenwood has gradually added to his game. His runs, his decision-making and his creativity have improved, which is reflected in his underlying numbers – 0.3 xGI per 90 last season vs 0.42 xGI per 90 this season.

Greenwood Per 90 – GWs1-28 vs GWs29-31

Goals: 0.07 vs 1.5

xG: 0.18 vs 0.56

xG performance: -61% vs +168%

Big Chances: 0.27 vs 1.5

Shots in Box: 1.66 vs 4.51

Shots on Target: 0.66 vs 2.25

Penalty Touches: 4.39 vs 6.76

Assists: 0.07 vs 0.75

xA: 0.1 vs 0.12

Big Chances Created: 0.2 vs 0.75

Chance Creation: 0.66 vs 1.5

While I don’t expect Greenwood to maintain such incredible over-performance of underlying numbers, the underlying numbers themselves are encouraging.

An xG 90 of 0.56 is a premium score, above any we have seen in the article this week. Greenwood’s Big Chances per 90 of 1.5 is absolutely massive, and does hint that should Greenwood continue to receive so many Big Chances, you can expect him to slightly over-perform his xG.

Rather than go through each individual stat and point out the improvement, it is worth simply saying that Greenwood’s current numbers are hugely exciting. The sample size (3 games) is miniscule, but Greenwood is a talented United player, who you’d expect managers to readily get behind. Another excellent game, and I think managers will flock to him.

Top Performing Midfielders Goal Threat Per 90 – Last 3 GWs

Many will feel encouraged by Jota’s and Salah’s numbers, and we’ve already seen how well Pulisic has done. However, Greenwood is in the next band and him being among the very top-performing midfielders in recent gameweeks demonstrates his potential.

Again, the caveat here is that the sample size is incredibly small. However, Greenwood clearly exceeds the apparently high-performing Mount. I do not expect Greenwood to score so frequently, but players in form do generally convert Big Chances at a high rate as can be seen by almost every other player in the table.

Upcoming Fixtures

It is hard to evaluate fixtures for United who often perform better against tougher opposition. Burnley and Wolves could be difficult opponents to break down, however, the others are likely to stretch themselves to score and so leave plenty of space for United’s counter attack.

Overall, I’d expect Greenwood to benefit from playing these more confident sides.

VERDICT – HOLD

It’s important to balance acting quickly with taking a measured approach – two contradictory concepts. To rationalise this, I’d say that at the moment there is an absolute wealth of excellent midfield options. While Greenwood could be the latest thing, he is far from the only one.

Another gameweek will give us more data, and hopefully a sample size weighty enough to act upon. After DGW32, we’ll be in a better position to plan out our midfield options for the remainder of the season. With the hard-to-predict Burnley game out of the way, Greenwood will also be well-placed to deliver in the run-in.

Thank you for reading!

I hope you’ve enjoyed reading the latest article in this new Weekly Series on Bandwagons. Please retweet, follow and join the discussion.

If I get calls wrong, let me know about it!

All statistics are taken from FFScout (which I have a paid subscription to) unless indicated otherwise.

Total Assists includes Assists and Fantasy Assists

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Start a Blog at WordPress.com.

Up ↑

%d bloggers like this: