Recent Posts



The New Metrics

Statistical analysis has, of course, always been with us but it really took off with the birth of the Premier League as one of Sky Sports’ shiny new value-added extras to be presented to viewers along with ever more intrusive camera positions, touch screen video tables and that annoying ‘swoosh’ sound effect that accompanies slow motion replays. By the way, standby for 360 degree virtual viewing - you may not have asked for it but it is coming anyway.

Generally I do find these stats add to my enjoyment of the match. Knowing rather than guessing the ‘Minutes Between Goals’ or the ‘Completed Passes’ of a player does make a difference when one is about to scream at them from the sofa for missing a goal. This season, the BBC have adopted the 'Expected Goals' metric. Yet even with all I have just written I am in two minds about it.

The funny thing about EG is that it is not a stat at all, it is a prediction based on the statistical analysis of other goals. So if out of one hundred players in a certain position with the ball at their feet, ninety-nine scored a goal in the past, EG tells us that tomorrow a player in similar circumstances is very likely to score also. Two things:

1. I don’t need a computer to tell me that.

2. Show me a model that could predict Ronaldinho’s logic-defying nonsense.

To be honest I can take or leave EG but seeing the mainstreaming of this metric did have me pondering, if not this, what sort of analysis would be good to see. ‘Shots, Blocks, Yellow Cards’ and ‘Dismissals’ are all useful but I would also quite like to see something like ‘Minutes Before Conceding a Goal After Scoring.’ Surely it would be a measure of a team’s all-round alertness and discipline, two qualities which however talented your players are essential if you want to win something.

This next one works for a team or player but was difficult to pin down. I tried very hard to put the ‘wet Tuesday night in Stoke’ question into a stat but the best I could come up with was ‘Win Ratio Long/Short Sleeved Jerseys.’ It is imperfect, I know, so if you can come up with a better metric to gauge the show-pony player who can only deliver the goods when the sun is out and the grass cut just so, do let me know.

A player stat that I really do think would be helpful for fans and managers is one that could measure, how to put it, not the fouls an individual player may suffer but the ‘Percentage of Perceived Fouls Suffered Not Given’ or alternatively ‘Time Spent On Pitch While Not Standing/Running/Heading’ (i.e. on the damn ground when they’re supposed to be on their feet). I suppose this is one stat that would have to be reviewed post game to validate the reason for the studied players reaction which means it will likely never happen.

At this point things can start to get silly very quickly with one imagining fans and sociologists requesting stats on ‘Number of Haircuts in One Season’ or ‘Amount of Jewellery Worn On Substitutes Bench.’ I think we can all recognise this is the thin edge of the modern-football-is-rubbish wedge. Modern football has its quirks but it is not rubbish. I will say this, though; I think we could all learn from a new stat that would measure the amount of time spent in pointless conversation with a referee or any other official after a decision has already been made. I’d probably call it ‘Time Wasted Arguing Toss’ (TWAT).

Midfield General (ret.)