Celtic must mix up their shot variety under Brendan Rodgers

“I say it’s very important (to have the weapon of Turnbull’s long shots). I was talking to our analysts the other day on it that when you are playing against teams that are low on the pitch and have a line of bodies, sometimes you have to shoot, you have to take shots outside the box because you can work the keeper, or it draws someone out and you can make a pass. But you got to do it and we have guys like Turnbull and Palma who have that quality shooting from outside the box and you see the benefits of it.”

READ MORE: The Celtic Numbers: Stellar Liam Scales shines in Dingwall

What we learned from this is that taking long shots is an element of Rodgers’ tactical play especially against low blocks. On Saturday, Ross County were down to ten players early and soon fell back into a low block, and boy, did the team deliver. Here is the shot map from the game:

18 of a remarkable 39 attempts at goal were from outside the box. Who can fault the manager and the tactical approach, when Celtic’s first and second goals were from rang from the aforementioned Turnbull and Palma?

Well, when you are averaging just 0.09 xG per shot, I would have concerns. Although Celtic racked up 3.53 xG according to StatsBomb, and won 3-0, what’s the issue?

The case against

We’ve read Rodgers’s case for long shots in certain scenarios. My concern is that when you are having the volume of shots Celtic achieved at Dingwall, the average per shot is a more accurate indicator of overall quality and of breaking down the opponent than the sheer volume of xG accumulated.

In other words, Celtic aggregated a high xG through volume rather than quality of effort. As we saw, it was only at 45 minutes the deadlock was broken and not until the 77th minute that the game was secured. It was ultimately comfortable but - if we are honest - a frustrating watch for long periods.

This is because long shots are inherently low-probability efforts. Most long shots have a between one and five per cent chance of scoring. That means - even at the higher margin - you’d need to manage to get 20 such efforts during a game to generate just 1.0 xG.

Why? Firstly, professional goalkeepers are highly skilled, and most will not easily be beaten by a shot from range where they can see the flight and trajectory and have time to position themselves.

Secondly, physics. You must hit a ball perfectly to get it into the small portion of the goal a goalkeeper will not have time to cover, given the distance and relative reaction time. A slight miscue when it leaves the foot becomes highly exacerbated over 25 to 30 yards.

Thirdly, the chance of the shot being blocked is quite high. Given you are likely to have two lines of defence between you and the goal – the midfield and defensive lines. The sheer number of bodies can generate a favourable deflection on occasion but is more likely to defend the effort.

Last season in matches involving Celtic, there were 442 efforts outside the box and 35 per cent of them were blocked. Inside the box, there were 843 efforts but only 19 per cent of them were blocked.

So, you are nearly twice as likely to not have your shot reach the goal from range than from inside the box. That means you won’t get the secondary potential benefit Rodgers talked about regarding the ‘keeper spilling a rebound. However, that is not to say Rodgers was incorrect in focusing on the requirements for that game under those specific circumstances.

Trends

What I would caution against is a return to the shooting strategies we saw the first time around under Rodgers.

There has been a steady decline in the percentage of Celtic shots taken outside the box, especially under Ange Postecoglou. The Australian wanted his side to work the ball into positions whereby the ball could be played into the centre of the box – generally, the place where the highest xG will be generated.

And he is in step with modern thinking as this article in The Athletic highlights. Volume of shots is less important than shot quality and this is borne out in Celtic’s own data.

There is a clear trend of taking fewer shots and this season is seeing the lowest average yet (this does not yet include the Ross County game which will raise the average significantly).

Improved analysis has shown the power of xG and the value therefore in taking shots as close to the goal as you can. A long shot has a further potential disadvantage in that you are likely to lose possession for a very low probability of scoring. The risk/reward ratio is against you.

Taking all that together, we can see, finally, that the average xG per shot has been on the increase as the volume of overall shots - specifically long shots - has fallen.

Under Postecoglou the average xG per shot exceeded 0.15 for the first time in these records, and significantly higher than Rodgers’ first season 0.11.

For context, last season Celtic took 911 shots. If they are averaging 0.04 xG higher at that volume, they could expect to score an additional 36 goals over a season! That is clearly significant.

Lessons from Stuart Armstrong

Trends in football are towards taking less long-range shots. Acceptance of xG concepts has largely informed coaches as to the value of working the ball into higher xG areas to take shots.

Of course, the same coaches are more appreciative of the need to defend such areas and marshal resources accordingly. Celtic have followed industry trends and under Postecoglou attacked low blocks by flooding the box in central areas; encouraging aggressive “pack” passing; moving the ball at speed; and getting strikers (especially Kyogo Furuhashi to expect low cross balls between the posts and close to goal.

Domestically at least, this was hugely successful with 147 goals in 53 matches. But Rodgers is also correct that, on occasions, when faced with a very low block, as evidenced on Saturday, expert long-range shooters introduce an unexpected element, and the attack should be varied. A specialist weapon to be used as a feature of variation and surprise rather than a central pillar of your attacking strategy.

READ MORE: Celtic's Oh Hyeon-gyu analysed following Dingwall showing

I’ll leave you with a reminder of a facet of Rodgers’s first two seasons. In the 2016-17 season, Rodgers’ first, Celtic scored an incredible 29 goals from outside the box. A conversion rate of six per cent effectively 'beat' the xG model that season. Stuart Armstrong scored seven such goals including in the Scottish Cup Final.

In the following season, only 14 goals were scored from range and the conversion rate halved to 2.8 per cent. Armstrong did not score a single goal from outside the box in 60 efforts.

As I said, a feature to be used wisely.