As always, the most important match of the year in club football, the Champions League final, provides us with many column inches from football pundits: will Ronaldo be crowned best player of the decade or will Buffon steal his thunder? Will Real Madrid cement its hegemony in European or will Juventus confirm the ascendancy of Italian calcio?
Read more: Ben Ainslie is gearing up for the race of his life at the America's Cup
As an alternative to opinionated punditry and biased judgements, WIRED is here to provide with a mathematically rigorous, fact-based analysis of the match, with the assistance of our friends at 21st Club, a football analytics consultancy that uses machine-learning algorithms to help Premier League clubs make intelligent decisions around areas such as squad evaluation and player transfers.
“We help football clubs think more strategically,” Omar Chaudhuri, head of football intelligence at 21st Club, explains. “Most clubs rely on very reactive decisions, on a game to game basis.” Blake Wooster, 21st Club's CEO, agrees: “You don’t have to outspend your rivals. You have to outthink them.” Here’s their smart take on the final in Cardiff.
If both teams field their expected lineups — Navas, Varane, Ramos, Danilo, Marcelo, Kroos, Modric, Casemiro, Ronaldo, Benzema and Isco for Real Madrid and Buffon, Chiellini, Alves, Sandro, Bonucci, Barzagli, Pjanic, Khedira, Dybala, Madzukic, Higuain for Juventus — Juve are 58 per cent more likely to win the tournament.
“This is mostly due to the negative impact that missing key players like Bale, Carvajal and Pepe,” Luis Usier, quantitative analyst at 21st Club, says. This is according to a statistical model that uses machine learning to isolate the contribution of every individual player to the team, assessing their recent performances, the opponents faced, among other factors, and computes how many goals each team is expected to score — a measure called “pre-match expected goals”. In this case, Madrid has 1.16 expected goals against Juventus’ 1.45.
However, should Bale and Carvajal be available and fully fit, Madrid would them become slight favourites, with 1.33 pre-match expected goals to Juve’s 1.26.
Still, the 21st analysts consider that the Madrileno’s attacking prowess is evenly matched by the Italian’s defensive coherence. “Juventus has managed to build a very good squad based on defense, which is usually more cost efficient,” says Chaudhuri. “Whereas Madrid has built a team that is as good overall, based on more expensive attacking talent.”
Should Madrid not be able to field the world’s best player, its chances of winning the final would decrease to a mere 31 per cent. That in itself spells out how important Ronaldo is for Los Blancos. Juventus, in comparison, is not as reliant on their charismatic goalkeeper: missing Buffon would only decrease their odds of winning to 51 per cent. Mathematically, they would remain favourites, but only just. “This means that Ronaldo has 50 per cent more impact on his team than Buffon does,” Usier says.
“The loss of Ronaldo would be devastating to Real Madrid.” Chaudhuri agrees: “This is consistent with a trend that we observe: the best attackers do have more impact than the best defenders. Attacking talent costs more, but justifiably more. The fees are in line with their contribution. The market is not as inefficient as people think.”
One of the foundations behind the two finalists’s success this season has been the smart rotation of their squad. Both teams have made, on average, more than four changes to the starting lineup each game (the average in the top five European leagues is 2.6). However, the Spanish team still boasts of a much more effective bench: all of their top substitutes — Vazquez, Morata, Asensio and James — have a much higher impact on their team than the Juventus’s substitutes.
Although Zidane is being hailed as the architect being Madrid’s success, the club’s performance haven't really improved from previous seasons.
Madrid has been performing at approximately the same level as they were under Carlo Ancelotti and even the unloved Rafa Benitez. In fact, Madrid’s goal difference this season has been worse than at any point since the 2012-13 season, when Jose Mourinho was their coach. “They also have benefited from their opposition having players sent off in key matches, most notably with Bayern, who got a red card in each leg of their Champions League tie and almost beat them nonetheless,” Usier says.
The impact of Allegri, on the other hand, can be quantified at around two extra points per season since taking over in 2014. “We look at the improvement in team strength over the course of a manager’s tenure. We know how our model rated Juventus when Allegri joined and we compare it with how it rate them now.”
Read more: Robo-journalists could soon be writing sports reports in the UK
The analysts at the 21st Club analysed how performance correlates with the age of the players, finding that attackers are at their peak in the range of 24 to 26 years, midfielders between 25 to 27, and defenders at the 26 to 28 range. This season, Juventus relied on many players who, while still excellent, are past the average performance peak: Buffon (39), Chiellini (32), Alves (34), just to name a few.
Indeed, this season Juventus played 71 per cent of the time with players past their peak. “So even though Serie A can sustain older players for longer, this really is the end-of-a-generation team for the club,” Usier says. On the other hand, Madrid has played 41 per cent of the time with players at their peak range.
“People completely overreact to one result, especially in knockout competitions, where one game dictates everything,” Usier says. “If Madrid win 4-0, people will say that were obviously the better team and everybody will be wiser after the fact. But that’s an overreaction to just one match.” Wooster agrees:
“Football is such a low scoring sport. The team who deserves to win a match, based on the quality of their performance, actually only wins the game 60 per cent of the time. That means there’s 40 per cent randomness involved in a game.”
This article was originally published by WIRED UK
