TCW's data and analysis writer - Alan Morrison - looks at where Celtic are putting themselves most at risk in transitions and the 'toxic combination' they must address ahead of next season's Champions League.

A feature of Celtic’s recent loss to Hibernian was the number of times they gave the ball away in their own half. StatsBomb had the home side with 30 pressure regains – where they applied pressure to the ball carrier and ended up regaining possession. The Hoops managed just 19.

Despite the disparity in points over the season, Hibernian were the team with a target to play for – a European qualification place. Their intensity reflected this, much more so than from a Celtic side who had already achieved their objectives.

Nevertheless, the ability to win the ball back high up into the opponent’s half and the distress this caused to the defence was clear to see


As a data point, the impact of that can be discerned using the packing concepts which I have previously introduced.

In this case, two data points – pack turnovers and pack recoveries.

Pack turnovers are where a player loses the ball and in doing so puts their own team-mates on the wrong side of the ball. That is, they are behind the ball and therefore, out of the game.

Pack recoveries are the other side of that transaction. This counts the number of opponents that are now out of the game because you caused the turnover and/or recovered the ball.

Two examples of this can be seen in recent fixtures:

In the first minute at Easter Road, Alexandro Bernabei played a blind pass inside towards Tomoki Iwata. The Hibernian attacker anticipated this and nipped in to win the ball.

Bernabei had committed a pack turnover. By giving the ball away he took all but three of his team-mates out of the game. Only the two centre-backs and Bernabei himself were between the goal and the ball. From this incident Hibernian had two shots at goal, the first a good chance for Elie Youan.

We can quantify the pack turnover impact by taking one defender (Anthony Ralston), all five midfielders/wingers and the forward out of the game. In data terms, that is a negative score of 14.

The Hibs forward who intercepts the pass gets a positive pack recovery score of 14.

From the recent derby and again inside the first minute:

The ball is played out left to Reo Hatate, who cuts inside but is then challenged by John Lundstram and the ball is lost, which is picked up by Todd Cantwell in an advanced position further up the field. There is a degree of luck as to where the ball ends up following the tackle, but it again highlights the dangers of being dispossessed anywhere from the midfield area and back towards your own goal.

READ MORE: Celtic boss Ange Postecoglou opens up on Tottenham links

In this instance, following Hatate’s pack turnover, there are only four Celtic players between Cantwell and the Celtic goal. This is a negative pack turnover score of 12.

Lundstram is awarded a pack recovery score of 12 for forcing the turnover.

Toxic Combinations

A concept I like to use to frame selection risks is something called toxic combinations. That is, are there players who if played together introduce increased risks.

This concept crystalised in my mind during the awful 2020-21 season when Celtic struggled on and off the field. On it, a combination of a centre-back used to a deep, well-protected back line (Shane Duffy), was protected by an ageing Scott Brown and flanked by two full-backs in Diego Laxalt and Jeremie Frimpong, who were often caught up the field in transition. Neither full-back nor Duffy himself were reliable passers of the ball. It was turnover and transition mayhem.

Pack Turnovers

This season is the first I have collected data for these packing concepts.

What can it tell us about the current squad?

Firstly, here are the pack turnovers per 90 minutes – simply the number of turnovers by each player per game:

It was slightly coincidental Bernabei and Hatate were chosen for the examples above in that both moments occurred early in those matches so I could find them easily, but also because those two players are the most prevalent turnoverers (new word!) in the squad.

It is something of a speciality for Bernabei, with 3.55 per game, the only player over three. Now remember not all these turnovers occur in the defensive or midfield thirds but we’ll come to the overall score which reflects that.

What I would like to draw attention to is that there is a prevalence of turnover potential on the left of Celtic’s alignment.

Bernabei is a standout, and Greg Taylor is fifth on the list. Of the centre-backs, normal left-sided centre-back Carl Starfelt provides the most turnovers from that position. Of the midfielders, Hatate is normally used as the left-sided of the number eights.

Back to toxic combinations.

If you have a left side of Starfelt plus Taylor or Bernabei in defence, with Hatate ahead as the eight, then you have a concentration of players with a high propensity to turn the ball over.

It is one of the main reasons virtually all opponents target Celtic’s left side.

As an aside, this is another metric that highlights how solid Cameron Carter-Vickers is. And Alistair Johnston is also the least profligate of the full-backs and a huge improvement on Josip Juranovic by this metric. Teams rarely attack Celtic’s right side.

If we consider the overall packing score from turnovers:

Because the score takes into account the volume of players taken out of the game, and their relative position (e.g. if they are defenders you lose three points as opposed to one for a forward) what is noticeable from this view is that Starfelt, in particular, ranks much higher.

That is, when Starfelt gives the ball away, it generally means many Celtic players are now out of the game.

READ MORE: Celtic must accommodate Callum McGregor - James Dailey

If anything, the impact of the turnover, as expressed by how many of your own team-mates you are taking out of the game, reiterates the nature of Celtic’s left side and the potential for toxic combinations.


We can now quantify the impact of giving the ball way. If you station several players together on the field that have the propensity to lose possession relatively easily, then this provides the opposition with an obvious weak area to exploit.

Clearly, the better the opposition, the more likely this weakness will be exploited. However, we see week in and week out SPFL-level teams attacking Celtic’s left.

It is an area needing to be strengthened ahead of another daunting Champions League campaign.