A common narrative coming out of the Champions League campaign has been that Celtic performed reasonably well, all things considered, and that things could have been quite different had they just taken more chances.

The StatsBomb data for the group stage confirms this narrative to a degree, but it also appears that Ange Postecoglou’s tempered comments after Wednesday’s game were realistic. 

The Hoops did create a significant amount of chances, given the club’s relative amount of financial resources. Out of the 32 teams involved in the group stage, the Parkhead side's average non-penalty xG of 1.40 ranked 11th, xG from open play ranked eighth, and xG per shot 13th.

In addition, the Bhoys were ninth in clear shots created, which are defined as shots taken with only the opposing keeper between the shot taker and the goal. All these metrics confirm that the Scottish champions did indeed create a lot of chances, and also a lot of quality ones. 

However, the remainder of this column will be dedicated to attempting to examine where performance levels fell short and contributed to Celtic finishing 29th in non-penalty goal difference and 23rd in non-penalty xG difference. 

Celtic Way:

The table immediately above breaks down the Hoops' 2022 Champions League campaign into various performance metrics, while comparing them with the 2021 Europa League group stage. The “Core On-ball Value” metric is something I calculate, which combines three of the five OBV components: pass, dribble & carry, and defensive action. Shots and related goalkeeper actions are relatively low-volume events in games, obviously, and as a result, their volatility from game to game, and even over six games in a group, can be significant. 

The core OBV components are relatively high volume and offer a gauge for the non-scoring aspects of gameplay. Obviously, goal scoring and prevention are paramount and will explore those variables shortly. The Parkhead side's core OBV ranked 19th out of the thirty-two teams, with core OBV difference ranked twenty-fourth. Perhaps as supporters may already know via the eye test, Postecoglou's men's attack-first style of play aligned with the OBV differences amongst the three components. Pass OBV difference ranked 14th, dribble & carry 24th and defensive action 30th.

This suggests that the Bhoys' ability to progress the ball and create chances relative to limiting the same for opponents was right near a “last 16” level within the group stage. Unfortunately, that appeared to be the only phase of play that reached that threshold. Core OBV difference suggested the Scottish champions were just above the level for finishing third in a group and qualifying for the Europa League. 

Despite creating chances at a healthy rate and level of quality, Celtic ranked 25th in npxG conceded, as well as 31st in xG per shot conceded - only Viktoria Plzen conceded higher. The attack-first style of play also left the Hoops vulnerable to opponents dribbling and carrying the ball in space, as the Bhoys ranked 32nd in both dribble & carry OBV as well as defensive action OBV. The balance between chance creation and the vulnerabilities those efforts created appeared to be deficient.

While not shown, the issues in both group stages from last year and in the Champions League were heavily skewed versus the top two opponents. In the previous campaign, the Parkhead side dominated Ferencvaros in most core-related metrics, and again this season versus Shakhtar Donetsk, while suffering the opposite versus the teams from the big five leagues. 

Finishing was obviously poor, and by shot OBV Celtic ranked 31st, while the opposition average was sixth highest. However, it is at the goalkeeper position where a worrying trend has emerged. As shown in the table above, the Hoops' keeper's play, including shot-stopping, ranked poorly in both group stages in the last two seasons.

Given Joe Hart’s struggles in seasons immediately prior to arriving at the club, perhaps this should not have been a surprise. Here are Hart’s shot-stopping-related metrics in all league games to date and compared to those from this season’s Champions League games:

Celtic Way:

We can see from the radar that the Englishman's shot-stopping has been right around what StatsBomb’s model suggests, or about average, playing in the SPFL Premiership. Those metrics dropped well into the negative during the Champions League games.

Even in the SPFL, the 35-year-old's deficiencies in claiming crosses and sweeping duties have dropped his goalkeeper OBV into negative territory, while at the Champions League level, the net performance result was severely negative. That followed last season’s Europa League group, in which Hart played five games:

Celtic Way:

I have written in the past about the former Manchester City man's poor fit relative to the requirements of Postecoglou’s style of play at the keeper position, and that could be true even if Hart were at least stopping shots at an average rate at the Champions League level.

Finishing is a highly variable aspect of football, and there is little doubt that, over time, Celtic’s will balance out. As shown in last season’s Europa League, when the Hoops ranked 6th in shot OBV. However, if the Scottish champions are to progress in European competition next season playing the same way, then increased athleticism in midfield and a significant upgrade at the keeper position should be top priorities.

Better finishing may make some of the score lines closer, but goal prevention needs to be addressed if serious progress is to be made.