Chelsea: How to use football statistics when analyzing players

There are always comparisons in the Chelsea fanbase among Chelsea players. Some don’t like it when people use statistics. They think the purpose of it is to hoodwink them, this is because many – either those who hate it, or even sometimes those who use it – don’t know how to use statistics. Football statistics are raw data. They don’t give you information. They don’t insinuate anything. They just give you raw facts.

Of course there’s a reason some facts are the way they are, and that’s where the nuances needed to interpret and analyze statistics come in. First of all, statistics are objective data, with specific, understandable definitions of each “stat”, so as to avoid ambiguity or vagueness in interpretation. In data collection, these things are called events, so long balls, passes, key passes, recoveries and other data collected are referred to as events.

An event almost always has a universally accepted definition, much like the rules of football. A goal is a goal in a game in any part of the world, same goes for fouls, yellow cards, corners, throw-ins and so on. No ambiguity. This ensures that comparisons between players are fair and evaluation of all players are done in a uniform way, making it easy to differentiate players who have performed well and those who have performed poorly. The single most important thing you need when analyzing football data is context. Context lets you know when a certain statistic is an anomaly. Some football statistics look like an anomaly and some even look impressive, but when you apply needed context, you realize that they are in fact not as impressive as they originally seemed. Here are some key guides in using football statistics effectively and correctly.

What does the football statistic say?

Football statistics are objective. They are data, sometimes raw data. On their own, they don’t tell you much. You must, however, know what the data is saying and what it’s not saying. Pass completion for example, is a ratio of the total number of pass he has completed to the total number of passes a player has attempted. Professional footballers are generally very good passers of the ball. However, pass completion requires important context before you can say a certain player is better at it than others.

Raw pass completion numbers take into account short passes and long passes together. Long passes however, are attempted far less than short passes, so while they make up part of the pass completion ratio, they hardly ever affect it decisively. An example is that if a player attempts 90 short passes and 10 long passes, he could end up with a pass completion of 90-percent, even if he doesn’t complete any of the long balls, so long as he completes all the short passes.

The statistic in this instance says the player completed 90-percent of his attempted passes, which is true. It, however, does not say the player is a good passer, it just says the player attempted a number of passes and completed 90-percent of it. It is left to whoever is researching or looking at the statistic to dig deeper into that statistic to see if there’s more to that number than meets the eye. On average, good passers of the ball average about 55 to 60-percent long pass accuracy over several seasons.

Midfielders and centerbacks of possession-oriented teams, often average between 90 to 95-percent passing accuracy. Therefore pass completion alone does not tell you much, it only tells you what percentage of the passes a player attempted where completed. Likewise the same with cross completion, aerial duels won, and every other statistics that is judged based on completion or success rate. An example of this is Andreas Christensen in the Manchester United Premier League fixture under Thomas Tuchel and Bernardo Silva in the Champions League fixture against Paris Saint Germain that was played a few days ago. Christensen and Silva both completed 100-percent of their passes in those games, however Christensen attempted and completed all 5 of long balls, while Silva attempted no long ball.