What is Celtic's solution to their winger problem?

We all tend to attribute that which our internal biases lead us towards. The reality is usually multi-faceted. Here is where the underlying performance data is compelling. Did Celtic struggle to draw with Kilmarnock because of the manager, the centre-backs, the wingers, and the strikers?

All will play some part. We discussed this on the Huddle Breakdown this week and Enda Coll queried the role of the midfield in movement and getting on the ball to beat the press. We discussed the pros and cons of Adam Idah pushing Kyogo Furuhashi back to the 10 position (and more materially limiting Matt O’Riley’s forward movements).

There are many other issues not least finishing the game with arguably eight reserves and only three genuine first-team starters on the pitch. Without getting bogged down in a general whinge fest, I want to return to a common topic this season – the quality of the wing play.

Wingers vs Kilmarnock

The speed of passing and physical movement of the centre-backs is a genuine area for concern, but despite that, the wide players were provided with the ball on regular occasions.

Despite being on the field only 22 minutes, Nicolas Kuhn received 12 pack passes, the equal-most in the team. Luis Palma received ten in his chaotically poor first 45 minutes. Yang Hyun-jun received nine in his 45 minutes. Only Daizen Maeda (one pack pass received) was either not given the service or could not find his way into space to receive the ball in wide position from a forward pass that took out some opponents.

READ MORE: The Celtic second-half slumps that may cost them the title

Therefore, the wingers received ample passes that had, to some degree, disrupted the Kilmarnock shape. From that service, there was a sum total of four shots generated, three by Palma and one by Kuhn. The cumulative xG was 0.35. The wide players created two chances from this possession one in each half for O’Riley and Paulo Bernardo with shots inside the box.

A combined expected scoring contribution of 0.77 from the four wide players. Maeda ended up with a negative packing score (-2) due to turnovers. Kuhn’s packing score of 83 was the highest in the team (in 22 minutes) further illustrating this wasn’t a failure to feed the wide forwards, it was what happened next (or didn’t happen next) that was the issue.

Historical perspective

I wanted to look back over the back catalogue of data to see how the current crop is faring versus wide forwards from yore. Equivalent data is available back to 2018-19.

It seems good news that Palma is the most prolific provider of shots and assisting passes back to the 2018-19 season. We all suffer from recency bias, however, and there are three wide forwards with slightly less prolific output all from the 2022-23 season in Jota, Liel Abada and Sead Haksabanovic.

It is a delicate ecosystem, however. Having high-volume wingers on one side allowed Maeda on the other to be more of a threat in behind as well as utilising his off-the-ball effectiveness – in other words, there was a balance to the attack. Both Maeda and now Yang are the lowest volume providers of shots and assists of this sample, unfortunately.

More important than volume is quality. We can measure this by the average xG per shot taken and the average xA of the chances provided.

Despite the very low volume of Maeda’s attacking output, when he does get into position to either pass or shoot, they tend to be high-quality efforts. There simply isn’t enough of it.

Whilst Palma provides high volume, his quality is generally low. In comparison, Jota was providing chances and shots equivalent to James Forrest levels, who was a very consistent operator in this regard. Yang is a toxic mix of low volume and low quality.

We can assess the extent players ‘make things happen’ in the final third as opposed to simply giving the ball away. The final third effectiveness percentage tells us the number of times the player ‘makes something happen’ i.e. a shot, a chance, a completed pass, a corner one versus losing the ball in the final third. And then we can add in the volume of final third ball losses per 90 minutes.

Essentially, this shows wastefulness or not.

When I discuss James Forrest, I talk about him a propo the modern winger as being a “conservative” wide-man. That is, he waited until the shot or pass was on before committing and was happy to recycle and keep the ball if not. The modern winger, like Palma and Jota, are encouraged to be riskier in their play and try more outlandish plays. Forrest’s peak seasons between 18-19 and 20-21 were characterised by maximising use of the ball in the final third. That may be too “safe” for modern managers.

Back to 2022-23, manager Ange Postecoglou and his wide-men are in the middle of the chart and therefore a balance between risk-taking and delivery. The current crop is outliers in the extent to which they give the ball away in the final third without delivering a chance, crossing or otherwise maintaining possession. That may be ok if the quality is high, but as we have seen that is not the case outside of the sparsely productive Maeda.

Summary

It comes down to balance. Having high-risk takers is ok if you also have some steady ball retainers and a consistent stream of quality delivery from elsewhere.

The problem Celtic has with wide play now is that the quality is not there in high enough volumes, and the level of wastefulness in the final third far exceeds the players’ ability to produce meaningful output.Either you provide a more rigid support structure around the wing play as per Postecoglou’s three-versus-two overloads (a full-back, number eight and winger ganging up on the full-back and wide midfield opponents) or you find a more effective use of resources.

READ MORE: Why Kenny Dalglish should have lifted the European Cup for Celtic

Palma has utility and given his strengths are passing and shooting whilst his weaknesses are pace and take-ons, I’d prefer him as a ten behind the strikers. Similarly, Maeda is highly effective when close to goal and his best performances are as a striker.

Hence some kind of 4-4-2 diamond or 4-3-2-1 Christmas tree would be preferable. A drastic change to shape is needed as the wing woes show no sign of abating.