TCW's data and analysis writer James Dailey explains why Celtic must continue to accommodate Callum McGregor as part of their midfield transfer policy.

Celtic’s midfield has experienced periods of significant transition since Ange Postecoglou’s arrival in the summer of 2021.

The initial iteration for early season friendlies included players such as Ismaila Soro and Liam Shaw, with Soro persisting as a starter into the beginning of the league and European campaigns. In fact, he even started the first Europa League group stage game away to Real Betis, with Callum McGregor out injured, alongside David Turnbull and Tom Rogic.

Upon McGregor’s return from injury, a period of relative stability endured with he, Turnbull and Rogic logging the vast majority of minutes until Turnbull was injured in the League Cup final against Hibernian. Nir Bitton and James McCarthy contributed to cover the limited number of minutes prior to the Covid-induced early arrival of the winter break. 

The remainder of the season brought with it the arrival of two subsequent mainstays in Reo Hatate and Matt O’Riley, with McGregor largely slotting into the deepest midfielder role. While Aaron Mooy has contributed significant quality minutes domestically during the current campaign, the other three have persisted in providing relative stability. 

As we all look towards next season’s campaign, there is obvious uncertainty about potential player departures and arrivals in the upcoming summer transfer window. While it may be plausible that relatively young players such as Hatate or O’Riley could attract and perhaps be tempted by transfer interest, Celtic’s captain is unlikely to be going anywhere. What kind of role may McGregor be poised to play next season and who will join him in the middle of the park.

One of the challenges Celtic have faced is moving between dominating games domestically versus the vastly different level of opponents at the Champions League level - particularly pot one and two sides. To place the disparity into a little bit of context, Celtic have averaged 71 per cent possession and 718 passes in league games this season up through last weekend’s draw versus St.Mirren, which compared to 44 per cent and 490 in this season’s Champions League group stage.

READ MORE: Celtic's poor form is over as the Scottish Cup final beckons

Celtic played much of last season’s Europa League group stage campaign with two attacking midfielders in Turnbull and Rogic, while this season Postecoglou appeared to tweak his tactics for the Champions League. Here was a pass map for the starting 11 players from last season’s first group stage game away versus Real Betis:

While movement and player positioning are dynamic in Postecoglou’s style of play, we can see from this map the average locations pretty clearly conformed to an attacking 4-3-3. Here is a comparable map from this season’s first group stage game versus Real Madrid:

The alignment was more in line with a 4-2-3-1, with O’Riley playing the role of a supporting striker, as Celtic’s rest defense transitioned to a more traditional 4-4-2 shape. This tactical tweak persisted for the remaining group stage games until the trip away to Real Madrid:

With McGregor out injured, O’Riley was the deepest midfielder, with Hatate and Mooy more advanced in what was closer to a 4-3-3. By half-time in that game, Celtic had conceded two penalties, 2.61 in total xG, with Vinicius Junior having not scored on two high quality chances. Obviously, the game would end 5-1, with Real Madrid having accumulated a total of 3.78 in xG. To place that amount of xG into perspective, the Rangers side that ended up conceding a record number of goals in this season’s Champions League group stage posted a worst single-game xG conceded of 3.26 versus Napoli. 

Single-game xG is often widely divergent from results, as things like finishing and keeper shot stopping introduce significant variance relative to the quality of chances xG is intended to measure. Conceding north of 2.00 in xG at the Champions League level can manifest in lopsided results, as the quality of attacking players in pot one and two sides is very high. For example, Celtic conceded far less xG in lopsided results under Brendan Rodgers, with attacking forwards such as Edison Cavani, Kylian Mbappe, Luis Suarez, and Lionel Messi having done severe damage.  

If Celtic are to make progress in next season’s Champions League, the ability to do a better job at limiting opposition chance creation may be paramount. Can Celtic afford to play an attacking 4-3-3 system with McGregor as the sole deep midfielder? Is Hatate sufficient enough defensively to be an adequate partner for McGregor if two midfielders are played deeper?

While the sample size is obviously very small, despite Celtic averaging only 44 per cent of possession in the Champions League, neither player offered high levels of defensive or pressing output. McGregor played an important role in linking up play in possession, but can the club afford to pair him with another midfielder who is not more defensively productive? 

READ MORE: Why Celtic's end-of-season dip in form is not unusual

McGregor will be entering the upcoming campaign aged 30, which is just at the front end of normal midfielder physical decline curves. While it may just be a fraction of a step-in pace or quickness, McGregor is unlikely to be a level-appropriate attacking midfielder next season. 

Along with upgrading the keeper position, investing in a young athletic and defensively imposing midfielder to play alongside the captain should be a top priority as Celtic look to make advances in Europe next season.