Just four points separate Liverpool in first and Everton in seventh after 17 matchdays of the 2020-21 Premier League seasons. The title race is seemingly well and truly on.
The Premier League has probably been the most competitive league around Europe in recent memory with five title winners across the past decade, even if Manchester City and Liverpool have dominated since 2017-18.
Liverpool may have the best attack (37 goals) and the highest Expected Goals (xG) total (31.7), but they did not win any of their three games over the Christmas break and failed to score in their last two. It was the first time they didn’t find the net in back-to-back league games since April-May 2018 (2).
Under Pep Guardiola, we have got used to seeing a very prolific Manchester City, who averaged a terrifying 2.5 goals and 2.2 xG per game between 2016-17 to 2019-20. However, this season their offensive numbers have dramatically dropped off to 1.6 goals per game (ninth-best) and 1.7 xG per game (third-best).
Conversely, they do have the best xG against total of 13.2. To steal an adage from American football, “Offense wins games, defence wins championships.”
Offensively, Everton (0.12 xG/shot) and Tottenham (0.11 xG/shot) have the highest xG per shot in the top-flight this season (penalties excluded), meaning they have been consistent in finding themselves in good shooting positions. Spurs will continue to rely on Harry Kane and Son Heung-Min as their attacking outlets – the forward partnership have already combined for 13 league goals in 2020-21, no other duo has ever combined for more in a single Premier League campaign.
The South Korean has also scored 7.4 goals more than his xG suggests he would be expected to score (12 goals, 4.58 xG), which is the best differential in the league. Tottenham are third in terms of their team xG differential (+6.35), only Manchester United (+6.67) and Southampton (+9.2), have overperformed to a greater degree. Will those sides keep that pace? Only time will tell.
The Stats Perform AI team has provided a simulation to generate predicted league tables for all of Europe’s top five leagues. The statistical model estimates the probability of each match outcome – either a win, draw or loss – based on each team’s attacking and defensive quality. Those ratings are allocated based on four years’ worth of comprehensive historic data points and results, with more weighting given to recent matches to account for changes in form and personnel over time.
The model predicts that Manchester City (68.7%) have far by the best chance of wrestling the title back from Liverpool (19.6%). Manchester United (80.6% chance) and Spurs (72.3%) are favourites to join those two sides in the top four. At the bottom, unsurprisingly, the model doesn’t like Sheffield United’s chances, giving them just a 4.7% chance of staying up. West Brom aren’t far off from that either (5%).
While the Premier League has been the most (it’s all relative) competitive league, the Bundesliga, alongside the Serie A, is quite the opposite. Bayern Munich have won each of the last eight championships, with Dortmund coming the closest to breaking that dominance in 2018-19 where they finished two points behind.
This season there is hope among non-Bayern fans that the dominance would change, this time thanks to RB Leipzig or Bayer Leverkusen. Those hopes took a hit when Robert Lewandowski popped up with a last-minute winner against Bayer two matchdays ago, allowing the reigning champions to jump back into top spot.
Unsurprisingly, the reigning champions have by far the best attack. They’ve netted 44 goals already this season, the second-best tally ever in the Bundesliga after 14 games and only behind themselves in 1976-77 (48). They also have the best goals-xG differential in Europe this season at +14.4 (44 goals, 29.6 xG), which is made all the more impressive given they’ve played just 14 matches.
Having the best goal scorer on the Continent in Robert Lewandowski (19 goals) and the two players with the most assists in Germany – Kingsley Coman (7) and Thomas Müller (8) – certainly helps.
Defensively, FC ‘Hollywood’ are not that imposing, with ‘only’ the fifth best xG against (17.0). They’ve already conceded 21 goals, which is against the norm as Bayern have not conceded more at this stage since 1991-92 (23), while they’ve also only registered two clean sheets thus far which is lower than in any season over the last two decades. Hansi Flick’s side are certainly all-in for attack.
RB Leipzig might be the main rival for Bayern as they sit just two points behind them. They are sharp offensively with the second-best xG total in the Bundesliga, while also defensively solid, shown by having the best xG against within the top five European Leagues and the fewest goals conceded in the German top-flight.
Their aggressive pressing is one of the hallmarks of their play. PPDA – which looks at opposition passes per defensive action in the attacking two-thirds of the pitch – can be used as an effective proxy for pressure. Leipzig rank third in the league for this, behind only Hertha Berlin and Bayern.
Sadly, for the neutrals, the predicted finish is still very lop-sided. Bayern have a whopping 93.8% chance of winning their ninth consecutive Bundesliga title. The model gives RB Leipzig just a 4.6% chance of overthrowing the Bavarians, and if we’re looking for a ray of hope then at least that number is not zero. There’s always a chance!
While Bayern Munich seemed to struggle a bit defensively it’s quite the opposite for Atlético Madrid, as after 15 games into their LaLiga season they’ve conceded only six goals. Just one team has conceded fewer at this stage in the history of the Spanish top-flight – Deportivo La Coruna’s tally of five in 1993-94.
It’s hard not to think of Jan Oblak when looking these numbers. As outlined in this dedicated piece on the Slovenian stopper, he’s been a titan in goal for Atleti since his first game with the Colchoneros in March 2015. From this debut, he has nearly registered as many clean sheets (115) as conceded goals (127) in LaLiga – a phenomenal record. In 2020-21, he’s saved 87% of the shots that he’s faced in the league, which is the best rate among regular goalkeepers from the top five European leagues.
Using our expected goals on target (xGOT) model, we can predict how many goals a goalkeeper would be expected to concede, given the quality of the chance (xG) and the end goalmouth location of the shot. As a result, we can directly evaluate the contribution of the goalkeeper and see how many goals they prevented for their teams.
Using this model, Oblak has prevented 5.3 goals in La Liga already this campaign, with only Edgar Badia preventing more in the competition (+6.1).
Of course, Atlético under Diego Simeone have always been strong defensively, but it seems like there has been a positive shift in their offensive play thanks to the rise of Portuguese wonderkid Joao Felix and the arrival of Luis Suarez. Atleti have scored 24 goals from open play in the Spanish top-flight this season, only Barcelona have more (29). Suarez has scored a third of these goals (8/24) and has scored a goal withevery 1.4 shots on target, his best rate within a season since arriving in Spain at Barcelona in July 2014.
With two games in hand, Atlético could eventually see themselves eight points clear at the top of the league ahead of city rivals Real Madrid, who have managed to stay close to them without convincing performances. Karim Benzema has been superb with 13 goals involvements in 15 appearances so far (eight goals, five assists), the second-most in LaLiga.
2020-21 was the first time since 2006 that neither Real Madrid nor Barcelona were top of the league at Christmas. With 15 of the last 16 campaigns seeing one of those giants crowned as champions, Diego Simeone’s men will be hoping to replicate their famous title win in 2014 to upset the apple cart come the end of this campaign.
Can they do it? The model seems to suggest so with Atlético (60.2%) predicted to end the Real Madrid–Barcelona dominance in Spain.
2020 was a renaissance year for AC Milan. Although they still sit top of the Serie A table, their 3-1 defeat to Juventus in midweek ended a 27-game unbeaten run in the league and was their first league loss since March last year. It had been their longest unbeaten run in the top-flight since between 1991-1993, where they went 58 in a row without defeat.
With 37 points from 16 Serie A games, the Rossoneri already have 16 points more than last season at this stage and have started better only at this stage in the 21st century with 39 points in their title-winning 2003-04 campaign.
Despite the end of their unbeaten run, Milan’s current record scoring streak of 36 consecutive matches is still ongoing. They also set a Serie A record of scoring at least twice in 15 consecutive games in 2020.
They have been a dangerous team in transition, with only Roma have scored more goals than Milan’s six from counter-attacks. Theo Hernandez is the perfect exponent of this – creating the most chances following a carry of five meters or more (12, level with Mkhitaryan). He also has the second-furthest total carry progression (3418 meters) and the third-most goal involvements from by defender (7) thanks to four goals and three assists, behind only Achraf Hakimi and Robin Gosens.
But as always, beware of the old rival. Internazionale have been only other side capable of following their neighbour’s pace so far. The Nerazzurri have scored the most goals in Serie A this season (41) and are taking open play shots in great positions – their average xG per shot is 0.14, the second-best in the league.
Conte’s pressing style seems to be well integrated now and his side have racked up 148 high turnovers so far this season, behind only the pressing-masters Atalanta (167). Will it be a one city race in Italy? It seems like it, for now at least and it could end the nine-season reign of Juventus as Serie A champions.
It’s set to be a wild and unpredictable three-horse race in Italy. Our model gives Internazionale (33.8%) a slightly better chance than rivals Milan (23.1%) and Juventus (20.7%) winning the Serie A crown. One thing’s for sure: Juve have got a big fight on their hands to make it 10 titles in a row.
While some fans can attend games around Europe, that’s not yet the case in France, and the lack of supporters has had a big impact on the league. The current season in Ligue 1 has the lowest home win percentage and the highest away win percentage in a single season in the history of the competition. Lyon currently top the table, three points clear of Paris Saint-Germain and Lille. Injuries and inconsistent performances have been the bane of PSG’s start to the campaign – they’ve not been able to select the same starting XI in any of their 18 league games.
While the momentum might not be in favour of PSG right now, their numbers do not seem that worrying. The French giants have the joint-most clean sheets (10), most goals scored (40), most chances per game (4.9), the most high turnovers (168), and they’ve conceded the fewest goals (11).
Lyon have also been ferocious going forwards, scoring 37 goals and accumulating the highest xG figure in the league (43.5).
What’s worrying for their rivals is that Lyon have the worst goals-xG differential in the French top-flight (-6.5), meaning they could and should get better as the season goes on.
These two Olympiques – and even a third one in Marseille – are Paris’ strongest opponents for the title. It will be a very interesting challenge for PSG’s new manager, Mauricio Pochettino.
The model expects Pochettino’s side to clinch another Ligue 1 crown however, giving PSG a 79.3% chance of winning the league. Lyon have a 17.9% chance of winning their first title since 2007-08, while Marseille and Lille fans, unfortunately, need to go back to dreaming.
Our Expected Points Model Explained:
- The model estimates the probability of each match outcome (win, draw or loss) given each team’s attacking and defensive quality.
- The team’s attacking and defensive qualities are based on four years of historic results, with more weighting given to their most recent results.
- The model will take into account the quality of the opposition that a team scores or concedes against and reward them accordingly. For example, scoring vs Man City is worth more than scoring vs Newcastle.
- We can simulate the upcoming matches using goal predictions from the Poisson distribution with the two teams’ attacking and defending qualities as inputs.
- Finally, we simulate the outcome of the season 10,000 times in order to estimate the likelihood of each team finishing in each league position
Enjoy this? Subscribe to The Analyst to receive five stories each Friday from Stats Perform. It’s free.