What’s the best way to identify talent? The statistics vs visuals debate

What’s the best way to identify talent? The statistics vs visuals debate

WATCH ENOUGH HIGH-LEVEL MATCHES and one begins to appreciate the complexity of true player development – or at the very least, acknowledge it. The problem, however, is that many mistake appreciating the complexity of player development with understanding it. What makes player development fascinating is the fact it’s the ultimate inexact science. People find intrigue in studying the practices and pathways that produce prospects in the hope that one of them is a pure phenom.

Player development is the art, science and alchemy of hedging bets on performance and potential. It’s no secret that talent identification is extremely difficult to assess, which either leads some down the rabbit hole of over-complicating the game for future generations and others down the path of disregarding valuable player traits and tendencies altogether.

We live in the age of nuance. The ability to see the forest through the trees is as important as the ability to see the playmaker in the marauding horde of youngsters romping up and down the pitch on a Sunday morning. It’s the ability to split hair and discern whether a talent showcasing skills at speed yet unopposed is a real player or merely an apprentice in performance art. This murkiness exists because we also find ourselves in the digital age where the game is argued, parsed apart, learned and consumed online.

Never before has the divide between sports analytics and data science been so at odds with what the trained eye and experienced scout can figure out within a few minutes of watching a player in meaningful competition. The really good scouts have learned to assess a player by how they carry themselves in warm-up or away from the pitch. The art of understanding the person and the player is not assessed in spreadsheets and datasets.

However, trends, objective behaviours in training and match play cannot be ignored. The expansion of applied data science has permeated the world outside the chalked white lines, so it makes its integration into football inevitable. And we aren’t talking about armchair analytics for the scores of pseudo coaches on Twitter. This is the task of collecting data across vertical markets and applying the metrics at such a level, rate and degree that coaches can consume the information and apply it.

In essence, what’s measurable is marketable, and as such, the influx of analytics in football has a way of dividing the traditionalists – who watch football and understand it in their bones – from the agents of change with doctorates in economics, data science and analytics-based disciplines. The irony is that the game is ultimately about outcomes. The score at the end of 90 minutes is a measurable and objective figure frozen in finality. The game itself has a finite set of statistics and analytical moments that, when combined, account for the actionable periods of play and further study by all.

Read  |  Where do you draw the line between progression and success?

The old guard of scouts rely on the intangibles only they can recognise, rationalise and ultimately document. Part of this is true brilliance, but it can cross over into the realm of safeguarding their own role within a club or federation to remain relevant in the fast-moving age of football threatening to leave those who don’t evolve and adapt choking on its exhaust fumes. To ignore this class of football knowledge savants is surely folly. To rely solely on this sect of assessing player development in an age of unlimited access, data and predictive technologies is insanity.

So what does data science offer football if it’s a discipline well above and beyond the armchair analysts’ level of understanding? In short, a good principal data scientist helps people make decisions through large-scale analysis, and by creating technology ranging from performance predictors run on complex algorithms that calculate goal scoring based on shots on target, for example, to ultimately get an xG figure – expected goals.

An unintended consequence of this influx of data is what is known as data exhaust. Exhaustive data pools form of useless datasets that confuse most and empower the stubborn to make a case that the extraneous statistics matter when in reality they may not. Perhaps this is what places football at the crux of such a clash of methodologies and ideologies.

The game has forever been run through the grinder of method over madness and ruled by ideologues demanding adherence and subscription to their method. Where the fan sees the possession statistic as a measure of quality along with distance covered, passes completed and shots on target, the football mind sees what type of possession is at play and in what section of the pitch it occurs and when.

The point is that most would rather have higher possession in the attacking third than more overall possession (however it’s likely both are achieved if a team has the ball more often in the attacking third as they are dominating play). Some scouts and analysts don’t care for possession because it’s a misleading statistic and instead look at transition play and rate of transition-to-goals or goal-scoring opportunities. These examples indicate the multi-levels of the game present in a single instance or match.

Again, these professionals may never kick a ball nor do they need to because they specialise in the distillation of insight from large volumes of data and thrive in the creation of intelligent systems to offer customised solutions. Top clubs and federations pay top dollar to keep these assets on their staff and their work is largely proprietary and unpublished because it creates a competitive advantage in a savage global competition pool ad marketplace.

What’s even more intriguing is that members of this class of science may not have a passion for the crest of a club but their passion lies in applying the power of the theory and experimental design from the behavioural sciences into online adaptive systems.

Read  |  Why we need to teach young players the importance of winning

To illustrate just how far the game changes and just how indiscriminate football can be, take a figure like Arsène Wenger, who told FourFourTwoAt 12 you can detect if technically a player can make it or not. At 14 to 16 you can detect if physically he will be able to cope with the demands of professional sport. And from 16 to 18 you can start to see if a player understands how to connect with other players. At 20 the mental side of things kick in. How does he prepare? How does he cope with life’s temptations and the sacrifices a top player must make? This is a job where you must be ready. If you get a chance, you have to take it.”

On the surface, the only objective figure is age in this quote; however, Wenger is talking about aspects that, once again, don’t necessarily show up on a dataset or spreadsheet into account. Technical ability, maturity, the complexity of life in general. Perhaps these things have always been measurable – we just always haven’t had the tools, licence or wherewithal to actually see them, nor the ability to cull what’s applicable from what’s just extraneous information.

This is a man who revolutionised football in the Premier League era, and while he’s certainly lost his effect, as do all great influencers in time, the man changed the pedigree of ‘boring boring Arsenal’ to a free-flowing football club. His best sides played some of the most exciting football and his ability to recruit young players from the continent and forge them into a squad composed of a lithe blend of traditional taskers like Tony Adams, Lee Dixon, Martin Keown with the flair of some great French and Dutch players amongst others remains impressive.

Wenger’s early approach and opinions on diet, match preparation and recovery laid the foundation for others to improve on similar approaches. A few examples of Wenger’s impact early-on are as simple as banning booze during stretches of the season. Other simplistic methods include Wenger wearing a stopwatch around his neck and bringing an end to the archaic and old-fashioned training methods that turned into lengthy crucible training sessions – instead opting to calculate the time each player stretched and took part in short, sharp and intense sessions.

These sessions, naturally, became catalogued as did a player’s’ participation and performance in them so the staff knew when to press a player or when to back off to optimise them for match day.

Away from the argument of metrics lies that of methodology. Perhaps no clubs have mastered this better than Sporting CP and Porto – both of whom have revolutionised the way they produce young players capable of being selected and inserted into the top sides in Europe because of their technical, physical and psychological abilities to excel in one-on-one situations.

Read  |  Football’s next frontier: the rise of detailed analytics

A former professor at the University of Porto is Vítor Frade – a scholarly figure and an architect of total player development in ways others have adopted and reaped the rewards from elsewhere. This is a man who didn’t play at a high level, nor did he manage a team. He theorised an approach to the game called Tactical Periodization.

Because football is a fluid game that combines different game models and tactical implementations, the route to achieve winning football requires a process that deconstructs the game into consumable segments in specified areas and timeframes of synthesis. Such an approach is present is academic disciplines, thus reinforcing the use of academia in the realm of sport.

But to what extent? Where one competency, say fitness, is developed, the right type of fitness can be overlooked, under-trained or undervalued. The same is true with body composition. All the power, size and strength in the world means very little if the player cannot fit into the right system or lacks the technical refinement or the decision-making ability to be an asset instead of a hindrance on the field.

From the lore of discovering the diamonds in the rough, picking the scrawny player from a swarm of nomads romping around a cinder pitch, court or street for these represent the romanticism and mysticism of the game’s most infinitesimal degree of discovery. Capturing lightning in a bottle through happenstance means less and less when we consider the methods of the tried and tested scouting practices used by the top clubs.

The reality is that when it comes to player identification and player development, there is no singular answer. Furthermore, most players, by the time they’re old or good enough to be discovered, have been passed over and have had their career. The ceiling has been hit and they don’t even realise it – a ruthless facet of the game that will never change.

Whether one subscribes to data-heavy practices to assess talent or uses the eye test and experiential analysis, the game endures. Both ways to indicate player development remain an enigma tempting theorists from inside and outside the game to attempt to solve its quandaries and quagmires. The war between the schools of thought will persist, and football will be the ultimate victor as the best of both will rise above the noise.

Those lodged in one mindset over another are just as likely to change their stance as an elephant is to climb a tree if you ask it to – but in the end, the process is what matters most. The answers to player development and scouting remain elusive, especially since the variables continue to change as do the stakes.

By Jon Townsend @jon_townsend3

No Comments Yet

Comments are closed