THE World Cup 2022 has come alive – without theItalyas we know, and not without controversy – and they have already reserved the first surprises: the defeat of the super favourite Argentina against the much less noble Saudi Arabiafor example, or that of the fearsome Mannschaft against the Japan. In short, an eventful world championship that is giving us quite a few surprises and overturning many predictions. And if we refer to the predictions it is impossible not to think about the forecasting algorithmsat the data science: it is these days, for example, the news that a team of scientists from Alan Turing Institute has developed a forecasting model which, based on 100,000 simulations of World Cup matches, has predicted that the national team will be Brazil to lift the world cup on December 18 next to Lusail Stadium. But science and technology are not only concerned with these aspects, which perhaps interest bettors more than fans: today, in fact, science and technology are profoundly transforming football, supporting coaches and athletes ever more closely to develop better tactics and improve own performance. Numbers and algorithms in hand.
Here’s how it will end
Let’s start with the perhaps more “trivial” and certainly more dated aspect, that of the predictions for the 2022 World Cup. As he says David Adams in a piece recently posted on the site of naturesfor decades statisticians Those involved in football have focused almost exclusively on the number of goals scored and conceded by a team and on finding the best model to predict them. Variants of these methods are still used today to predict the result of matches: one of them, for example, assumes that the number of goals scored and conceded is distributed around a certain average value, and was developed by a team of epidemiologists from the ‘University of Oxford. It did us well, given that he predicted (correctly) that the Italian national team would beat the English national team in the final of the 2020 European Football Championships (more precisely, he predicted the greater probability of this with respect to goals scored and conceded); but not only that: he had also nailed six of the eight teams that reached the quarter-finals. “Basically we want to get to give each team an offensive and defensive ‘score’ – he explained, always a natures, Matthew Pennone of the developers of the model – calculated starting from the total number of goals that each team has scored and the strength of their opponents: by inserting these parameters we obtain a set of equations to be solved to calculate the two scores, and it becomes relatively easy to make predictions for each match”.
The template we mentioned at the beginning, referring to this year’s World Cup, works in more or less the same way: its creators have given each team an offensive and defensive score, eliminated the “home farmer” (a parameter that increases the probability of victory for the team playing at home and which of course in this case applies to all teams except Qatar) and fed to the algorithm the results of many pre-world international friendlies. With this tool they have therefore simulated about one hundred thousand matches thus arriving at the conclusion that it will be the green and gold national team that will have the better of the others. Another group of researchers from the University of Innsbruck, Austria, with a slightly different model. The insurance company Lloydhowever, used yet another model (which had correctly predicted the victories of Germany and France in the 2014 and 2018 world championships) arriving at the prediction that this time it will be theEngland to win (beating Brazil in the final). Still different conclusions, finally, for the Penn group, whose model instead crowned the Belgium. If the algorithms don’t agree either…
From the match analysis to scouting
But, as we said, predictions are just a small slice of the pie. And not even the most succulent, at least for those who practice football as a professional. “Science and technology – He tells us Vanni Di Febofootball data analyst for the Italian soccer federation (Figc, i.e. for our national football teams) – they enter football in at least four major areas: the so-called match analysis, the scouting, injury prevention and rehabilitation and finally all aspects of a more nature corporate”. In match analysis, Di Febo explains, data from cameras and GPS worn by players are used to obtain variables of interest and build game patterns: which player is stronger at hitting the header, for example, who concludes more passes, who is stronger in tackling, which area is played more or less, which side is more likely to score (or concede) a goal. “The cameras give us details on the position of the players and the ball thirty times a second – he explains Matthew Giacalonewho dealt with match analysis for theInter – It is an impressive amount of data, which we process and from which we derive indicators which we then share with the team’s technical staff. Our information is stitched together with video clippings from matches and training sessions and synthesized into a video that the manager can then review and show to the team.”. Not only: “With these data – continues Giacalone – it is possible to build a offensive danger index which is essentially the linear combination of various variables – for example the chances had, the number of crosses, possession of the ball, etc. – and which measures in an ‘objective’ way, after the match, which was the best team”. Even if winning the game is another story: “Football is a very different sport from others, for example from basketball or baseball – says Di Febo again – as it is a continuous game low score: a single episode is enough, the ball that goes ten centimeters further or further, to condition an entire game. It is precisely because of this high ‘volatility’ of the game that making predictions is so difficult. Basketball, for example, is different: chance can have a certain weight in scoring a shot, but given that many more points are scored overall, it is legitimate to expect – and indeed it is – a ‘large numbers’ effect which causes that almost always the strongest team wins”.
THE bigdata they are also very useful in the process of scouting: “Thanks to these tools – Giacalone says – we are able to download and analyze the parameters of hundreds of players and understand which ones are the best to buy, both in economic terms and in terms of technical characteristics. We can refine the search up to 3, 4 players and then tell the coach which one is the best.”. Di Febo takes care of this: “Right now we aim to monitor all potential international players – explains – and we’re scouting players from the under 15s onwards, examining all the data we get from the match lists: appearances, goals scored, bookings. In this way we try to identify the most interesting names and go and see them for yourself”.
Then there is the question of injuriesand even in this case technology plays a fundamental role: the data collected by football players, in fact, says a lot about their state of health. “If for example we know that a fit player has an acceleration of 35 m/s2 – says Di Febo again – and instead we see that in training he can’t get to that acceleration, we can imagine that there could be something wrong. We are also able to monitor any excessive and too close efforts and alert the technical staff that the athlete could get hurt”. And finally the whole part corporatewhich is equally relevant in professional football: “We can make forecasts and assessments, for example, on the bonuses received by players Giacalone explains. but also on which are the teams from which you buy (or sell) the most, what is the percentage of foreign players and so on”.