Statistics in Football, Relevant or Not?
Statistics in football is something that has gained a lot of traction in the past few years, and it has always intrigued me. Initially it was something I didn't like, but as with all things, it is important to be open to learn and adapt to new things, so I gave it a try.
I have come to a definite stance on it, and I want to share. It's not the all and all, and I am still open to learning and changing my views, but this article is what I think about Statistics in Football today. I want to answer these questions; Is it relevant? Does it do more harm than good? Is it really used in the right way? Let's take a look..
I looked up a definition of statistics, and it says "Statistics is a branch of mathematics dealing with the collection, classification, analysis, interpretation of numerical facts, for drawing inferences on the basis of their quantifiable likelihood (probability) of data." ( Wiki/Oxford dictionary ).
The most important part of that definition for me is the "for drawing inference" part. Statistics on its own most times involves just numbers, but these numbers should tell a story, the right story. The way I look at it, if it does not tell the right story, or it is used in a way that distorts the true story, then of what relevance is it?
I've selected a few popular stats, and I'll go through their definitions according to Squawka, then i'll explain my issue with it.
Chances Created
The definition of a Chance Created by Squawka is " A pass that leads to a shot on goal ". I don't know how many people have looked up the definition used to come up with this, but when I did, I was surprised. So many "chances" are created that do not lead to a shot on goal.
A chance created is independent of the final outcome, and the final outcome should never take away from what the chance creator has accomplished. Imagine a Central Midfielder ( or anyone ), gives a through pass that has clearly split an opposing defense and put the attacker clear on goal with a keeper; The attacker skies the shot or plays it wide. Do you know this would not be counted as a Chance Created for the Midfielder?
If this happens 3 times in a game and the team lost and I look up the stats for that Midfielder, I would see Chance Created = 0. Does this tell the true story? Does this not take away from what that midfielder actually did? Has the actual relevant occurrence not been distorted?
Let's look at the next stat shall we?
Key Passes
The definition of this is "A Pass that leads to a shot on goal that is not converted". If I had a problem with the definition of "Chances Created", then I obviously have a problem with this, simply because a pass can be "key" without leading to a shot on goal.
Let's actually assume we want to have a stat like this, right off the top of my head I can imagine David Silva splitting a defense with a behind the lines pass to Kyle Walker the fullback, who then squares it up for Jesus to tap in. That pass by David Silva is my key pass. ( My definition would be: A pass during the ball circulation phase of the attacking organization that signals the point where the opposition defensive set up has been bypassed ).
I wrote about the different phases of an attacking organization here
But this snippet may help,
"Movement and Circulation of the Ball
This phase is how you move the ball around in midfield or the final 3rd of your opponent. Think of Xavi and Iniesta moving the ball around with passes, or De Bruyne and Silva, or Pogba and Mata, or Pjanic and Dybala or Khedira, or any midfield and forward line. The aim is to disorient and bypass the opponent's last line of defense, and progress the ball up towards the opponents goal to lead to the next phase of offense. "
Important to know that Silva's "key pass" should be independent of what Kyle Walker ends up doing, and what Jesus ends up doing. It should tell the story of Silva's ability to do what he did. If I was a scout watching David Silva, I would focus on his ability and what he has done, as opposed to what his teammate did.
Assists
Another very popular stat. A pass that leads directly to a goal. It is going to dovetail nicely with what I've said earlier because my first problem with this is how so much emphasis on an outcome determined 100% by player B ( the finisher ) is used to judge player A ( the passer ). If player A sends a perfect pass to player B and player B misses, why does this go on to affect player A in terms of less assists?
If you put 2 playmakers behind 2 different forwards, and both playmakers have similar creative abilities, do you think they would end up with the same number of Assists stat? Would the Assist stat tell the true story? Why should someone's ability be judged on what another has done? Shouldn't the smart thing to do be focusing on the creative ability of the playmakers? So why not focus on the fact that the playmaker has created a chance ( my own definition ), or a key pass ( my own definition: A pass during the ball circulation phase of the attacking organization that signals the point where the opposition defensive set up has been bypassed ), which is what actually points to his ability? I think this is the right thing to do.
This should lead me to what I would define as a Chance Created; A Chance in its raw form is basically an opportunity, an opening, a possibility of something happening ( a goal ). Whatever happens next is the outcome, which takes away nothing from the fact that an opportunity existed ( do you see the difference between my definition and Squawka's? ).
It does not have to lead to a shot on goal, but it recognizes the existence of the opportunity in itself. So a Chance Created should ideally be a pass or move that leads to an opportunity/the possibility of a goal occurring. When you watch a chance created, you will know.
If you rely on the stats provided mainstream, you will not get the full story of the opportunities created by a player, unless they change their definitions, and try to tell a better story.
My second problem with "Assists" is that it just gives you a number. I may give a very short sideways pass to a player beside me in the middle of the field and he goes on to dribble everyone and score. It goes on my Assist numbers, but the point of the whole Assists stat is about creative ability. Would this tell the full story? Would this not distort the truth? So why depend on this fully?
The "Chances Created" stat as defined by Squawka is then a combination of the above 2,
Chances Created = Assists + Key Passes.
If there are issues with the individual components of what makes up the stat, then there are issues with the stat itself ( the result ).
Let me pick out a few defensive stats to touch on;
Number of Tackles/ Number of Interceptions/ Number of Blocks/ Number of Anything
More volume does not automatically equate to better ability. People use these numbers to compare players, forgetting they are a function of how the team plays, or the specific role of a player within his team.
A DM playing in a super dominating team will obviously have less blocks or whatever than a DM playing in a team almost always under pressure. So why put up numbers side by side for both players and use these volume numbers to somehow infer one is better at that parameter than the other?
A defender playing in a team always under pressure will have more tackles than one in a dominating team. If you put up defensive numbers for a season to compare 2 players, will the fact that A had higher numbers mean he was better than B? These numbers give no bearing on how the different teams played, or how they were set up. Why does this now translate to ability comparison?
Clean Sheets
This is another stat used to compare goalkeepers that should never be used. Clean sheets are a terrible way to compare goalkeepers. They are more a function of so many other things than solely the keeper. A less able keeper in a better defense can have more clean sheets than a better keeper in a worse team/defense.
Clean sheets are a team award, or if you want to go deeper, the whole defensive crew award. If you want to compare keepers, look at their agility, reflexes, shot stopping, positioning, aerial ability, command of the area, discipline, consistency ( and this comes from actually watching the keepers for a good period of time, over many games ).
These may or may not be parts of what contribute to clean sheets, but they are just a component of what makes up a clean sheet ( players do too ). So why not isolate the Keeper ability as opposed to using a parameter that involves a team effort to directly compare goalkeepers? Do you see the folly in using clean sheets?
The Golden Glove award and it's contemporaries in other countries is nice and all, but it's not a true keeper comparison parameter. David De Gea after recently winning his first Golden Glove ( Imagine using the fact that he had not won a Golden Glove prior to this to somehow talk down on his ability; this is an example of a big problem with this stat ) said It's a team award, and personally I believe Clean Sheet should be a team award.
Summary
Clearly, the major problem with statistics in football is 2 fold;
1- Their definitions and how they are recorded
2- The distorted stories they tell
My Solution
1- Re-define the definitions of these stats ( and how they are recorded ) to better reflect on the abilities of the players.
2- Restrict stats to the relevant player position or roles. I've seen people include "chances created" in comparing a DM and a CAM. How does this align? What is the basis of comparison?
3- Major emphasis on the Eye Test, with stats as a minor supplement. This is the most important part of it all; Never ever discard the eye test. It is the only thing that tells the full story. You want to know the full story, watch a player. Watch over a period of games. Watch and see how he can intercept, or how he can create, or how he can score. Watch and see his ability. Imagine using "Assists" stats to judge Andres Iniesta Lujan, you would never ever get how good a player he was.
4- Stop using stats for skewed agendas. Everyone is guilty of this, even mainstream media. They pull up statistics of different players, put them up side by side and compare just to push an agenda. It defeats the whole purpose, and mis-educates the public. It doesn't tell the whole story ( like I've pointed out ), it isn't well defined ( like I've pointed out), and it can be presented in a skewed way to push an agenda. The public gulps it all, and that's it.
There are many more stats I can delve into ( take-ons, pass accuracy, Expected Goals, etc ) and pick out what's wrong, but for brevity sake I'll stop here.
If stats are not telling the true story, or tell half a story, or are used to push false stories, then they do not have a place in our beautiful game. But if these things can be corrected, then I'm all for statistics in football.
In the meantime, remember to always critically think when stats are presented to you. These are the questions to ask yourself;
1- Is it relevant to the role/position of the player?
2-Does it tell the story of the ability of the player independent of the outcome decided by another player?
3- Is this just a volume stat that doesn't really say A is better than B, and fails to take into account how both teams play?
I believe asking yourself these critical questions would help in bringing better context to stats as they are presented today. I hope one day, we get to a better place with statistics in football.
P.S: The World Cup is around the corner. I'm so excited and cannot wait for it to start. So many questions.. Who will win? Who will be the breakout star? Which team will be the surprise package? I hope to get a few World Cup articles out, keep an eye out for them.
You can always reach me with any comments or questions on Twitter: @otoiks