Thursday, 8 August 2019

2019-20 Fantasy Premier League: A preview

2018-19 FPL review

A frantic final week of the Premier League transfer window (purchase window, rather - there is still time for outgoing surprises) saw Harry Maguire (Leicester to Man Utd), Giovani Lo Celso (Betis to Spurs), David Luiz (Chelsea to Arsenal) and Ryan Sessegnon (Fulham to Spurs) secure major moves to the Premier League top 6, while Belgians Eden Hazard and Romelu Lukaku were notable big-money exits from the league during the summer window. West Ham have spent significantly in bolstering their ranks after Arnautovic's move to China, while Villa have made full use of their victory in the richest game in football by spending over £100m in a bid to stay in the top division.

The league also welcomes VAR this season, which should stir up memories of the 2018 FIFA World Cup, not only for its effective application, but also the fantasy implications of having increased penalties awarded.

Fixture difficulty review

Let us first review which teams are set for a soft start to the season that could provide value for picks from their sides. Last season, FPL assigned difficulty ratings from 2 to 5 for a player's fixtures based on opposition and venue. While the FPL Gods decided for 2018-19 that Brighton, Liverpool and Crystal Palace would be a notch easier to face at home than away, the powers that be have allotted a single difficulty rating to each team this season - Brighton and the promoted sides get a 2, last season's record-setting title racers City and Liverpool are rated 5, the rest of the Big 6 are rated 4, and the remaining teams have been slapped with a 3.

To start with, below is a list of teams sorted by the average difficulty of their first 6 matches this season.

Let's pause for a moment. Just how reliable is this difficulty rating anyway? A quick study of the PP90 scored against a team's difficulty rating lends credibility to the notion of taking this number a little seriously. When you break it down further by home/away, the point is even clearer:

  1. While there isn't a huge gap (relatively) between the 2s and 3s as compared to the other brackets, it is nevertheless more beneficial to go up against a 2.
  2. You get more points home than away, no matter who you face.
  3. Fixtures against the big 6 are real toughies, particularly against the top 2.

Player selections

Having reviewed some of the major numbers from last season, let us see how we can tie those to the difficulty of the fixtures the players are set to face.

Looking at players who provided a PP90 (points per 90 mins) of 5 or more with an opening-six fixture difficulty of 3 and under, let's see if we can pick off some high-potential picks.

The above chart, which plots last season's PP90 (vertical axis) against the opening-six difficulty (horizontal axis), provides a quick view into picks that could form the core of your FPL team. The color (low: green, high: red) indicates FPL ownership, and size indicates price.

The obvious

Everton and Bournemouth have been handed rather kind openings to the season. We've already seen Bournemouth assets deliver value with a kind opening schedule last season, which will make Wilson (8.0), Sigurdsson (8.0) and Fraser (7.5) enticing mid-price value picks. Digne (6.0) looks like a must-have in every backline at this point.

The next tier in fixture difficulty brings with it a leap in PP90, as we move into the Liverpool and Man City assets. It is quite subjective which mix to go with, but Mane's near-certain unavailability for GW1 puts Salah (12.5) and Sterling (12.0) at the top of premium picks to have. It wouldn't be too outlandish to bank on de Bruyne (9.5) improving his PP90 of 5.5 from an injury-blighted 2018-19. PP90 and set-piece action could also be used as ammunition to select Alexander-Arnold (7.0) over his similar-priced partner Robertson (7.0).

Premium options with difficult openings that can nonetheless not be ignore are Aubameyang (11.0) and fit-again Harry Kane (11.0), whose fixture-proof marauding should keep them on FPL managers' minds

A wait-and-watch approach is best taken with United and Chelsea assets, who begin the season facing each other, although it does look enticing both from PP90 and pre-season to bank on Barkley (6.0) to deliver the goods early on, with the Englishman having played his way into Lampard's plans.

The fringe

From the fixture difficulty list, it seems like cheap picks from seasoned Premiership sides Brighton, Palace and Burnley can be used as a decent fringe to enable a sturdy core of Liverpool, City, Bournemouth and Everton players. Leandro Trossard (6.0) could be a decent 4th midfielder, lining up for a more progressive-minded Brighton team, while their trusted FPL stalwart Matt Ryan (4.5) already features in nearly 14% of team selections. Elsewhere, Palace's Martin Kelly (4.0) is another popular enabler, finding himself in a staggering 24.5% of FPL teams. Burnley's best value option seems to be their keeper Pope (4.5), while Chris Wood (6.5) isn't the worst mid-priced striker option to start the season.

The Blogger's Team

Hopefully, this post has given you some ideas to start the season. The Friday night deadline is fast approach - Good luck!

Thursday, 30 May 2019

2018/19 Fantasy Premier League: The Numbers

To say this has been an incredible campaign would be an understatement. Between the best title race in memory, the worst top 4 race in memory and the goal-laden comebacks in European competitions, it has been a breathless gift of a season for Premier League fans that is already being sorely missed.

Things were just a little less exciting in the FPL race, with the winner nearly assured of his crown a week in advance. Adam Levy from New Zealand captured this season's FPL overall prize with a mammoth 2,659 points averaging just under 70 points per GW, setting the standard for next season's aspirants.

There's nothing like an FPL season review to relieve some of the post-season boredom, or at least kill some time ahead of the European finals. Let's jump in.

The players

Time well spent

The "points per 90 minutes" is a popular aggregate metric to gauge which picks deliver the most FPL points for the time they spend on the pitch over the course of the season. The upper end of the PP90 list is dominated by Manchester City, whose attackers form half of the top 10.

Fantasy Premier League 2018-19: Points per 90 minutes top 10

Ruben Loftus-Cheek's presence at #6 bodes well for his FPL prospects next season once he comes back from injury, as he seems to have established himself as first choice for the third midfield position. Things look as uncertain as ever for the likes of Mahrez, Jesus and Shaqiri, however, despite their efficient usage of minutes to deliver value. Liverpool's marauding full backs Trent Alexander-Arnold (6.77, #13) and Andy Robertson (5.96, #19) are the only defenders in the top 20.

Safe choices

The picture is slightly different from the PP90 if we want to look at players who blanked the least. Defining a "blank" as any score under 4 points in a gameweek, it is possible to come up with a "strike rate" - The percentage of matches in which a player scored 4 points or more.

Fantasy Premier League 2018-19: Top 10 players strike rate

One perspective on FPL captaincy is to sometimes consider the "safe option" - Someone assured of delivering points in a given gameweek. With that in mind, traditional wisdom is disproved somewhat with a defender leading the way - Liverpool's left-back Robertson posted a strike rate of an impressive 69%, ahead of the more-trusted Aguero (64%) and Lacazette (63%). Youri Tielemans is the only player outside the top 6 to make this chart, with an impressive 2019 with Leicester City that sees him lead illustrious names such as Kane, Salah and Hazard in this metric.

Bang for your buck

Introducing the aspect of price to PP90 provides insight into which players provided "value for money". If we use the prices from the start of the season, this is how the top 10 looks overall.

Fantasy Premier League 2018-19: Top 10 points per 90 minutes per pound

This, however, obscures the value of the PP90 metric with respect to price that led to the number in the chart. For example: Mendez-Laing of Cardiff blanked in 16 of his 20 appearances (80%) and features in this list by virtue of his 4 strikes propping up a mere 5.0 starting price. Additionally, players like Shaqiri and Lovren, while not quite as unstable, suffer from not having assured minutes.

If we look exclusively at players with a PP90 exceeding 4 and compare their PP90 with price, we can spot several players who over-delivered on their assigned price, and are worth keeping an eye out for next season.

Fantasy Premier League 2018-19: Points per 90 minutes vs Price
Some notable prospects:
  1. RLC (7.26 PP90 @ 5.5)
  2. Shaqiri (7.53 PP90 @7.5)
  3. Wilson (5.98 PP90 @6.0)
  4. Deulofeu (5.8 PP90 @5.5)
  5. Gray (5.53 PP90 @6.0)
The chart also clearly lays out under-delivering premium assets, such as Lukaku, Kane and KdB, who hopefully occupy more reasonable price brackets next season as a result - Particularly for the latter pair, who appeal and performances were diminished more by injuries than a lack of quality.

The Opponents

The idea here is to look at how stingy different teams have been overall, and home vs away, in allowing points to their opponents. The metric under focus is PP90 allowed per player faced.

Fantasy Premier League 2018-19: Points per 90 minutes allowed by team

Predictably, Manchester City and Liverpool (in that order) were the toughest teams to face, conceding a measly 2.3 and 2.4 points per 90 mins per player respectively. This is less than half the PP90 conceded by Huddersfield Town (4.92), the most lucrative opposition for players to pick up points against. A poor 2019 pushed Brighton (4.2) to the bottom of the "still in the league" list.

Most teams are stingier at home than away, as is intuitive, barring 3 notable exceptions: Crystal Palace, Leicester City and Manchester United, who allowed a higher PP90 at home. There are also 3 clubs who were markedly better at home than away: Bournemouth (gap of 1.46 between home and away), Chelsea (gap of 1.31) and Arsenal (gap of 1.19). Over the season, Spurs and Newcastle showed virtually no difference in points allowed based on venue.

I hope you guys enjoyed this post. You can access the full metrics for the teams and players used in this article here. If you have any suggestions or further data requests, just leave a comment and I'll see what I can do!