David Turnbull’s goal against Motherwell in the final league game of the season was his tenth of the season. It was a neat way to end a season that promised to be spectacular but ended in a long journey back to fitness after his hamstring gave out in the League Cup Final in December.

By the time that severe twang occurred, Turnbull had played virtually every minute of the Postecoglou era. Nine goals and seven assists suggested a bumper harvest come season's end. In reality, he made cameo appearances in the final matches, and got a cursory run-out against Armenia for Scotland.

With Celtic purchasing Reo Hatate and Matt O’Riley in January, it felt a very different Celtic side that ended the season in glory compared to the ups and downs pre-Christmas.

Is Turnbull at risk of becoming the forgotten man?

It always felt that playing Tom Rogic and Turnbull as the nominal no.8s in Postecoglou’s preferred 4-3-3 was a 'make do and mend'. Whereas the position demanded athletic box-to-box, all round ability, the Australian and the youngster from Motherwell seemed more laconic. Throwbacks to days when no hipster team was complete without a chain-smoking no.10.

I’m being slightly facile as regards Turnbull as he was definitely a no.8 at Fir Park. However, his lack of real pace, comfort on the ball and diffidence towards defensive “dig” suggests his role should be further forwards at Celtic.

What I want to explore is, now the season is done, let’s look back at Turnbull’s data in the round and ask the question “Did he improve?”. Remember, he is still only 22, and improvement is all we can ask of younger players.

  • All metrics bar % are per 90m.
  • The far-left column indicates the % change in performance against that metric
  • Green = +10% improvement
  • Amber = between +9% improvement and -9% regression
  • Red = -10%+ regression


Using a mix of Statsbomb and my own data, I will break down the information into four categories starting with creativity.

Celtic Way:

The noticeable increase is in the volume of Secondary Assists – those are the passes before the assisting pass that creates a shot.

Also, he showed great improvement in successful crosses into the box.

READ MORE: Celtic's 20-year wage spend bang for buck assessed: From O'Neill to Postecoglou

Number of passes into the opposition Danger Zone also showed slight increase. All this led to a healthy increase in overall Expected Assists from 0.37 to 0.46 per 90m. Despite this, actual assists achieved dropped from 0.34 to 0.22. Bad luck really, as he isn’t always to blame if the recipient doesn’t finish.

My concern last season was that his rate of assisting in open play was average, and it was elite-level set play abilities that boosted his overall numbers. xA from open play remains modest at 0.18. Given the bulk of his chance creation was from set play passes, and that Celtic are not an aerially dominant team, the reason for the disparity between expected and actual is perhaps clearer.

Ball Progression

Passing and running with the ball forward, next.

Celtic Way:

Significant increases here.

Some of this is team style and the demands of the manager. He wants his players to play aggressive passes through opponent’s lines, and carry the ball where possible, especially in one-vs-one scenarios.

Turnbull has improved his pack pass score noticeably, by 25 percent. He seems well suited to that aspect of the Postecoglou way.

Goal Threat

This is a big part of Turnbull’s appeal with 34 goals already in 110 senior appearances. He even managed a hat-trick for Celtic against St Mirren.

Celtic Way:

In terms of raw goals, his scoring rate decreased by 24 percent whilst his Expected Goals was steady at 0.28 per 90m. Real goals drive perception and expected goals drive predictability.

His main area of growth was in receiving forward passes (up 39 percent). This again speaks to the emphasis from the manager on aggressive forward passing but being on the receiving end of such passes is a useful proxy for movement, positioning, control, 'off the ball' awareness.

What is telling is that his overall volume of shots did not change much, but he took 13 percent more outside the box and 12 percent less inside. All this leads to a low xG per shot of 0.07 as 69 percent of his shots are outside the box.

This would continue to be the criticism of Turnbull – poor shot selection. This often happens with players who are capable of scoring from outside the box and deliver on occasion, getting the balance between when to take it on and when to resist is key.


Playing midfield in this Celtic side is demanding and with the full-backs often bombing forward, defensive capability is part of the job spec.

Celtic Way:

This is a really positive picture of all round improvement. Again, this is partly due to the demands placed on his role, but he still has to deliver to them.

Regaining the ball following team pressure is way up as are the volume of tackles on a possession-adjusted basis.

He wins the ball back more frequently and, crucially, in the opponent's defensive third.

There is a “but”.

If we compare Turnbull to bright shiny new thing O’Riley as regards defensive capabilities using Statsbomb:

Celtic Way:

See the right-hand side of the chart for the defensive actions and how O’Riley is well ahead of the Scot.

With a more athletic no.6 rumoured, meaning a move forward for McGregor, and the more defensive active O’Riley and Hatate available, what does this all mean for Turnbull?


David Turnbull, at 22, remains a player of huge potential for club and country. His injury-curtailed season should not detract from across-the-board improvements in his productivity given the demands of the role he has been asked to play.

He can continue to improve on shot selection and more direct creative passing from open play.

The main question mark is whether his improving defensive contribution is good enough for the manager going forward.

Being home-grown, young and clearly still improving, I suspect he will be given every chance to show he can fit into this Celtic side.