Do Not Sell My Personal Information

Do these Chelsea & Man United stars prove stats aren’t always so reliable?

Statistics are more prevalent than ever in the beautiful game. Nearly twenty years on from OPTA first translating the Premier League into numerical terms during the 1996/97 season, utilising their data for analytical means is now a staple task for the pundits of English football.

Not that it’s a practice exclusive to the television studios; websites such as Squawka and Whoscored, laden with statistics ranging from the number of throw-ins during any given match to the Premier League’s most prolific goal-assist combos, are peaking in popularity, whilst most top flight clubs now view statistics as an essential part of the player recruitment process. Sky Sports even embarked on a new chapter this summer, using data from Sports Interactive’s highly popular Football Manager series to run the rule over rumoured targets of the Premier League’s top clubs for the very first time.

But how much can statistics really tell us about a sport whose greatest moments often defy all rationale? Are they an effective substitute for good, old-fashioned eye-witness interpretation? Do they always give us the full picture?

THE STATS DON’T LIE

There is certainly something reassuring about the objectivity of numbers on a screen. Statistics can give an insight into the strengths and weaknesses of a player you’ve never heard of, let alone seen in action, and on the most-part they’re relatively reliable.

Premier League fans, for example, won’t be surprised to find Arsenal’s Mesut Ozil leading the division’s creativity charts with 4.9 chances per match, closely followed by West Ham’s Dimitri Payet and Tottenham’s Christian Eriksen, or branching further afield, four-time Ballon d’Or winner Lionel Messi averaging the most successful dribbles per match, 6.5, of any player in Europe’s top five divisions this season.

They’re indicative of not only a player’s traits but also their form and ability when compared to like-minded counterparts. It certainly seems a more trustworthy practice than simply relying upon the subjective mumblings of Sky Sports’ Tony Gale when formulating your opinions on a player. The apathetic angst towards BT pundits Michael Owen, Robbie Savage and Owen Hargreaves, to name a few, shows that even those considered ‘experts’ hold views on players, teams and football in general that most would disagree with – but it’s much harder to have a heated debate with a set of statistics.

You could argue with Hargreaves until you’re blue in the face about whether Douglas Costa is in fact the best one-on-one dribbler in the world – as he claimed during Arsenal’s 2-0 win over Bayern Munich in October – with nothing to go on but your own instincts and eye-witness accounts. But you can’t dispute that four players, including the aforementioned Messi, are currently averaging more successful dribbles per match than the Brazilian. It’s written, quite literally, in black and white on Whoscored.com.

STATS CAN BE MISLEADING

That being said, statistics can often be misleading. Take last season, for example. Chelsea’s John Terry and Gary Cahill ranked in the bottom eight amongst the Premier League’s centre-backs for successful tackles per match, with 1.1 and 1 respectively, whilst QPR’s Richard Dunne came first with 2.8 and Liverpool’s Dejan Lovren came third with 2.4. The Chelsea pair also came dead last in terms of interceptions per match, averaging less than one per game.

Yet they both claimed places in the PFA Team of the Year and won the league title, whereas Dunne fell through the trap door with Rangers and ¬£20million man Lovren was branded one of the flops of the season. Even amid the Blues’ sudden decline, I’m sure most would select Terry and Cahill over Dunne and Lovren in a hypothetical starting Xi – so why isn’t that reflected in the statistics?

It’s not an anomaly; rather, a representation of how Terry and Cahill, Lovren and Dunne had very different experiences last season. QPR finished up with the worst defensive record in the Premier League by quite some distance whilst Liverpool had the second-worst in the top ten, so Dunne and Lovren made more tackles simply because there was a lot more defending to do. On the other hand, Terry and Cahill were part of a well-organised Chelsea side that dominated the vast majority of their matches. Thus, less defensive actions were required, but they still produced them when they had to.

It shows how much more than simply ability influences statistics. Tactics, philosophy and situation are also major factors – and that’s by no means limited to player statistics. Manchester United have created the fourth fewest chances of any Premier League side this term, 90, but they haven’t scored the fourth fewest goals – in fact, they’ve scored the ninth most – and in stark contrast to what they might suggest, they’re fourth in the table.

STATS CAN’T GIVE THE FULL PICTURE

No matter how sophisticated statistical analysis becomes, it will never be able to give the full picture. Manchester United’s Michael Carrick is a prime example. He averaged less created chances, passes, shots, interceptions and tackles per match than midfield partner Ander Herrera last season, whilst finishing the season with less goals and less assists, but his influence on the Red Devils’ results was far greater than the Spaniard’s. United won 15 of the 17 Premier League fixtures he started last term, compared to just seven in the 21 without him.

The England veteran’s influence is an incredibly subtle one, as a midfielder who doesn’t maraud from end-to-end but rather wait until play comes to him and make right decisions at the right time, and that has proved almost impossible to quantify adequately.

It rings true with many players, even those topping the statistics charts. Can any data truly depict the genius of Lionel Messi, for example? His sensational goal against Bayern Munich in last season’s Champions League semi-final, in which he bamboozled Jerome Boateng into collapsing on the floor before chipping the ball over Manuel Neuer, was one of the most magical goals I have ever witnessed. But statistically, that’s just one successful dribble, one accurate shot and one goal.

Its indiscriminate to Moussa Dembele jogging past Alan Hutton and poking the ball into Brad Guzan’s near post – in comparison, one of the most defensively lapsed goals I’ve witnessed for some time – during Tottenham’s 3-1 win over Aston Villa before the international break.

SO WHAT DO STATISTICS REALLY TELL US?

Well, statistics certainly have their uses, especially in a comparative sense, but it’s not always obvious what they’re actually telling us. Comparing Manchester United’s match stats from their 2-0 against West Brom compared to their scoreless draw with Crystal Palace the weekend prior might be demonstrating an improvement in performance, but it might also be highlighting the inferiority of the opposition.

Likewise, Terry and Cahill’s low stats from last season might counter-intuitively suggest that they’re actually doing their job properly by anticipating danger early and getting into the right positions before tackles, interceptions or headers are required. It’s not simply a case of the higher the stat the better the performance, or, in this case, vice versa. Statistics are black and white, but football has a lot of grey¬†areas.

Every statistic must be put into the context of real-life football and only in an honest manner. Earlier in this article, I argued stats are objective, but they can still be used to illustrate a subjective point. It could be audaciously claimed, for example, that Eden Hazard is actually performing better than last season because he’s a creative player averaging more created chances per match, albeit by the slight increase of 0.06, but there’s no question he’s been a shadow of the winger who claimed the 2015 PFA Player of the Year award this term.

Clearly statistics are an incredibly useful and powerful tool, not only for analysing performances but also when illustrating arguments or describing the characteristics of certain players. But it is by no means a complete, indisputable science and the biggest danger lies in misrepresentation and misinterpretation. It’s kind of like that Goldie Lookin’ Chain song, Guns don’t kill people, Rappers do – or should that be, stats don’t fool people, pundits do?

Article title: Do these Chelsea & Man United stars prove stats aren’t always so reliable?

Please leave feedback to help us improve the site: