When a side barely drops a point in the season, it is difficult to pick fault in their performances.

Two years into Ange Postecoglou’s reign, we now have a significant body of data to assess the extent to which the Australian has improved this Celtic side.

Results and league table measure outcomes, but it is in performance data that the more predictive data lies.

In the week Celtic are crowned champions again, how does this league campaign measure up to last?

I’ll start in this piece with attacking trends.


StatsBomb provides a range of team-level attacking data for us to compare the last two campaigns.

Let's start with what has not improved. Overall expected goals are down, which may surprise some given the side are on a remarkable 105 league goals after 34 matches, compared to 92 in 38 last.

Yet the drop-off is there, slight as it is, and may reflect Celtic have outperformed their expected goals once again. There is an element of having by far the best attack in the league, they outperform the average SPFL strike rates. Also, after scoring more than three goals 13 times, there is a tendency for teams to collapse slightly when xG gets above 2.5 and actual goals get beyond three.

As I have discussed before, add the concept of “we don’t stop” to being able to bring on five fresh forwards of high calibre late in the game and Celtic can pad out a result quite comfortably.

The total number of shots is down slightly as well as the percentage of crosses successful into the box.

Counter-attacking shot volume is also down, again, because Celtic are attacking increasingly deep defences, the opportunity to hit teams on the break is reduced.

My take on these is we are seeing a reduction due to the opposition increasingly sitting deeper and being more compact. Pack passing rates are down which tends to happen against solid deep defences. 

Set piece xG is down, but this tends to represent a low overall percentage of the chances Celtic create and reflects also the tendency to play corners and free kicks short and try and play through sides rather than throw long balls into the box given the lack of height in the side.

Shot quality is up – the average xG per effort at goal. This is encouraging and we see this increased attacking cohesion leading to the number of low hard crosses into the six-yard box area central to the goal. I am less worried about overall shot volume decreasing if the shot quality is on the increase. It infers a better quality of shot decision-making.

The major movers upwards are high press shots and clear shots. For the former, this again hints at both a consequence for the defending teams of a deep-set defence – that Celtic can aggressively press high and that the quality of the press, something that takes time, communication and teamwork, is improving.

Regarding clear shots, see the point above looking at the higher average xG per shot. Celtic are working the ball into increasingly better areas – simply central and close to goal – to better effect.

Finally, there is a slight decrease in the goalkeeper's average pass length metric. Being a geek, this gives me a warm glow. It means that, as the opposition are generally starting deeper, Celtic can play out shorter and therefore are more likely to keep possession when Joe Hart has the ball.


Firstly, teams are sitting deeper and more compact, therefore volume and indeed overall xG are trending downwards – opponents are fixated on stopping the Celtic juggernaut.

However, the team's attacking patterns and pressing responses are improving with time and practice. Despite packed defences, Celtic are shooting from better areas and working better quality chances. They can also take more risk when pressing the opponents given how deep they now sit.

Defensive trends will be covered in a follow-up piece next week.