Spread the love

Paul Gamble looks back at the Bees’ last two performances and evaluates whether the stunning long-range goals are a coincidence, or point to a change in attitude. And did Bryan Mbuemo’s worldie against Millwall take a deflection, or not?

And with an incredible comeback win against Millwall and an utterly dominant performance away at Swansea, our season seems to have finally taken off. In particular, the atmosphere at Griffin Park on Saturday is something we’ll all remember for a long time and will surely be one of the defining moments of our final season there. After it had all subsided and the bones had been picked out of the performance, aided by the odd pint of Frontier or two, the overriding emotion was one of relief. Finally, luck seemed to have turned and we had got the result our performance deserved. Even the debate over whether Mbuemo’s goal had taken a deflection or not seemed to have been influenced by the view that we needed a change in fortune. For what its worth, I’ve watched the highlights an indecent number of times and I still can’t tell.

On the other hand, Tuesday’s match appeared to have nothing to do with luck. We roasted the Swans with a combination of serious defensive strength, slick passing football and attacking flair culminating in two fantastic strikes and what must surely be my favourite own goal of all time. Unfortunately, I was only able to watch this one via iFollow but coming on the back of the Millwall win it was a good feeling. Scanning the Griffin Park Grapevine and the Beesotted twitter feed, it was evident how much pride we had taken in that performance and comments were becoming increasingly optimistic for the rest of the season. We had finally hit our stride.

The interesting thing is that the stats don’t tell the same story. According to the excellent experimental 361, we outperformed Millwall comprehensively (2.5 to 1.5) and deserved less against Swansea (1.6 to 1.1). There are some obvious caveats to this. The Millwall XG came from the penalty and a tap in whilst Swansea were behind us all the way until they started throwing everything forward in the last 10 minutes. Add this to the ongoing debate about stats and Billy’s excellent recent interview with Rasmus, it does raise questions about how useful these metrics are.

The popular narrative has been for the last two years that we have underperformed compared with XG estimations and that a table of justice would see us much higher. The problem with this argument is that if you argue that luck is the main reason for the underperformance in the league table, the only way we can improve is to be luckier. Of course, whilst luck is a factor, its not the only factor and there are all sorts of other reasons. For example, last year we committed an incredible number of defensive blunders which lead to goals against in tight matches. Whilst a statistical analysis based on XG might lead you to conclude we had been unlucky, that would not be the right conclusion. Similarly, we have hit the post or bar without one going in a large number of times this season already. A simple look at XG might lead to the conclusion that again we were unlucky but in truth some of those chances should have been buried.

Of the six goals we’ve scored recently, three have been from outside the box and would have a low XG rating. Its certainly unusual for Brentford as our playing style doesn’t tend to encourage that many shots from range, focusing instead on trying to create higher XG chances in the box. The problem is that this allows teams to defend inside the penalty areas leading to blocked shots and rushed chances that although positionally might have a high XG value, are actually more difficult.

Nowhere was this more evident that against Charlton. I have no knowledge of what goes on inside the club or what tactics the coaches are encouraging players to use but to me, this recent increase in shots from distance and goals scored looks like more than a coincidence. This would suggest that it is not a question of whether stats are useful or not, but more a question of what stats are being used and how they are being used. Certainly, as fans, we only have access to a fraction of the metrics the club uses.

Every Brentford fan can remember those defensive blunders last year, but I am sure that the club could tell you exactly how individually effective our defenders and keeper were. Over the last twelve months, the Directors of Football have almost completely rebuilt the central defensive elements of the side with Mepham, Konsa, Barbet, Woods and Bentley all moving on, some for very big money, yet this part of our team looks much stronger this term. It’s also no surprise that in that interview Rasmus was talking up the recent defensive improvement as it had clearly become the major focus for the club over the summer both in the transfer market and on the training ground. This work has to be grounded in statistical analysis.

A different website, fivethirtyeight.com, has published an expected end of season table based on their model of current and future performance. They have us down as finishing fourth this season. Betfair has us as fourth favourites for promotion at 9/2 and fifth favourites to win the league. Whether we achieve that will depend on a lot of different factors of which one will be luck. No one can control but a good understanding of the statistics that underly the performances can only help the manager and the players improve the things they can.

Paul Gamble