A reoccurring subject that generates much emotional response is the topic of corners/set pieces (let’s call them set-plays) both attacking and defending.

Set plays are one aspect of the game where every supporter of every club has a similar thought: we don’t score enough from them, and we concede too many goals from them.

The reality is that if we just take corners, they are not a very efficient way of scoring, given around 1.27% of them result in goals.

That being said, it is an opportunity to put into place rehearsed attacking moves and in the interest of marginal gains, should be maximised as opportunities.

Celtic have had a small team, physically, for the two years under Ange Postecoglou. With two centre-backs barely six feet tall, and the tallest outfield player (Matt O’Riley) not renowned for his heading, it is no surprise that set-play opportunities seemed to result in short passing options and the retaining of possession. Very little piling of the ball in the mixer.

That is my perception, but does the data support that?

READ MORE: Celtic's recruitment mistake that they must not repeat

Defending Set-plays

We’ll start at the back.

When I record shots at goal I differentiate between corners and set pieces which we’ve lumped together as “set-plays”. Then, I indicate whether the shot resulted from the direct set-play pass, or from a secondary action. The latter is more of a judgement. An example would be a Celtic corner is played into the box, a Celtic player flicks it on, and the ball is then intercepted by a defender but falls to another Celtic player who gets a shot away. This shot is deemed a secondary action from a set-play.

I have such data going back to the middle of Brendan Rodgers’ first spell.

Pleasingly, the trend of actual shots resulting from set-plays and goals conceded has reduced over the six seasons recorded.

It fell over Rodgers’ reign, which is encouraging, but was lowest in Lennon’s 19-20 season - before spiking up in the disastrous COVID-19 season - and then falling again under Postecoglou.

This feels a little surprising, given the sheer size profiles of the respective teams. Lennon had the likes of Christopher Jullien, Kristoffer Ajer and Shane Duffy in his sides, who were some of the tallest players at Celtic in the last few years. Rodgers struggled to have a settled backline pairing and - as mentioned above - Postecoglou had small centre-backs, relatively.

This view records actual outcomes, and of course, I’d recommend looking at expected goals as being a more accurate reflection on the quality of chances created from set-plays.

Here is the xG conceded by season from set-plays:

The trend for expected goals follows actual quite closely.

xG conceded fell gradually through the latter two years under Rodgers before spiking upwards under Lennon. Under the tutelage of the Australian, Celtic went back to Rodgers’ year three levels of threat against.

Attacking Set Plays

Is the perception that Celtic does not score enough from set-plays justified?

Let’s look at the same profile in the attacking sense:

The attacking trends are a little more subtle. Rodgers had more success in year two.

Celtic’s best recent season for generating shots directly from set-play passes was under Lennon, which may be another manager-style factor.

Under Postecoglou, the trends of shots generated went back up to the same level as Rodgers in 2017-18.

Irrespective of the number of shots generated from set-play scenarios, the number of goals scored has been quite flat throughout.

Let’s consider xG again, in terms of that generated by Celtic:

The trend for the quality of set-play chances created is a falling one with a notable spike under Postecoglou in his first season.

He accounted for both the second-highest and lowest values in his two years.

This may be reflective of the impact of relatively small samples.

Summary

A slightly dull but important conclusion is that the trend in set-play defending and attacking is far less extreme than the emotions of supporters who form views on effectiveness.

In reality, there are subtle trends over time that appear to reflect levels of organisation and structure rather than personnel. That is - despite having a small team – Postecoglou - being a highly positional-based manager with high levels of systematic team structure and organisation - was able to reduce the threat from set-plays despite having the smallest physical team in memory.

READ MORE: The Celtic problem position the club must address soon

Rodgers is a similar type of manager and shows similar trends in the set-play attacking and defending numbers.

Lennon is more of a relational coach with less emphasis on positional structure and rigidity. It also suggests when confidence ebbed and team selection became erratic, one victim was the set-play organisation which lurched from being highly successful in 2019-20 to becoming the worst of recent years in 2020-21.

What does this all mean? I think that under a highly structured coach like Rodgers we should expect a higher degree of organisation and that will drive up better numbers irrespective of the size of the team.