In my last piece I looked at how Celtic’s attacking play has changed, whether improved or otherwise, between Ange Postecoglou's two campaigns.

Thoughts now turn to defending. Last season the Celtic team conceded 22 goals in 38 matches (0.58 per game). This term it is 25 in 34 (per game).

Has Celtic’s defence declined in quality, or are there other factors at play?


Using StatsBomb data we can compare several data points across this season and last.

The headline is that the xG conceded has dropped by a small margin despite conceding more goals. While Celtic may have benefitted from fortune in scoring many more goals than expected, similarly, they may just be slightly unlucky to concede more than expected.

Certainly, as regards the eye test, nothing is worrying about in Celtic’s domestic defending compared to last season. In that campaign, it took Carl Starfelt a good while to settle, there were injuries, and there is no doubt his data has improved this season. 

Meanwhile, Cameron Carter-Vickers is as consistent as ever, as I wrote last week.

Celtic Way:

Joe Hart, over the season, has conceded less goals than expected, so again, there are no negative flags there. He has conceded 25 goals but post-shot xG suggests an average of 28.71 goals should have accrued given the quality of the opponents’ efforts.

Shot quality from the opponents is reassuringly low at 0.08 xG per effort. Many opposition chances come from corners or free kicks which tend to be low-quality chances given the number of defenders around.

The biggest positive is that Celtic are conceding 57 per cent less chances from counter-attack situations this season. The positive for Celtic is that the rest of the defence – the placing of players not engaged in attacks, is stronger and more rehearsed. Secondly, as opponents are sitting so deep, they simply have less opportunity to break on this Celtic side.

READ MORE: The four unlikely Celtic players who should start against Rangers

A further knock-on effect is that Celtic’s average defensive line is now 48.86 metres from goal as opposed to 46.09 last season. Simply, Celtic can push their defence a bit higher up the pitch as a starting position – closer to the opponent’s goal in fact and play a higher line.

The combination of pressing effectiveness increasing and restricting the opponent’s space to play by playing higher up the pitch sees their passing completion drop from 68 per cent to 65 per cent. Teams may be going more direct against Celtic, but this will ultimately give them more ball control and thus, in concert with the five-sub dividend, tire the opponents out quicker.

Indeed, passes completed per opposition defensive action is up slightly as Celtic are pressed less effectively.

Simply the number of shots conceded is down which makes sense in conjunction with the decrease in average xG conceded.


Despite conceding more goals this season in the league, Celtic’s defensive data support the evidence that they are increasingly more effective in the press and forcing the opposition further back and restricting their space.

Individual defensive performances have generally improved or remained consistent from a very high benchmark (Carter-Vickers). Celtic are an increasingly efficient defensive unit. 

The concession of more goals does not appear to be a cause for alarm and could just be an anomaly that is not uncommon in such a low-scoring sport where luck plays a significant part. 

And the part played by the VAR decisions on penalties over the season may also be impacting this.