The last international break in October saw the benefits for Celtic of having a manager with a coherent plan given time with the players on the training ground.
Since that break, Celtic have six wins out of seven in all competitions.
Most strikingly, Expected Goals Against (xGA90) have dropped from an average 1.27 to 0.63 – that is, halved.
Yes, daft goals are still being conceded as seen in both Hungary and at Dundee, and Starfelt’s injury is another in a long list, but the time spent refining the defensive shape has borne fruit.
It is also true to say that Rangers, domestically at least, are improving their fluency at the same time. Three wins and two draws over the same period but a big increase in goals (17 in 5 matches).
So, where does that leave Celtic in the crucial league title race?
Expected Points
The excellent @TheGersReport on Twitter produces a league table projected from Expected Goals for and against. This is broadly the same methodology Matthew Benham developed to build his betting empire and wealth before taking over Brentford. Turns out Expected Goals are more predictive than real ones.
When people say, “the table does not lie”, it turns out, it does!
@TheGersReport saw Rangers’ big SPFL winning points lead coming early in the season based on underlying performance data.
This season, current projections expected Celtic to have an 11 point lead by the end of the season.
Proven methodology or voodoo hokum? Time will tell. It is a model; they are not always correct!
Let’s look at SPFL performance specifically to see if we can assess how the rest of the season might pan out.
Attacking Performances
As mentioned above, the gap is closing in attacking performance indicators given Rangers' recent six and four goal victories.
Celtic average almost 1 xG extra per game. This could translate to an additional 38 goals over a season, so it is a significant gap.
Indeed, Postecoglou's side are ranked 1st in all these attacking metrics except Shots in the Box and Crossing where they are second. It is a feature of the Gerrard/Beale reign, influenced by Liverpool analytics, to be patient and await the right moment to cross or shoot.
That being said, Celtic are easily ahead of the league in average shot quality (xG per shot). Their average of 0.15 per shot against a league average of 0.1 may not sound much. But consider, Celtic are averaging 10 shots more than the rest of the league outside Rangers. That means an extra 1.5 xG per game!
What is noticeable is how small the gap is between the teams across these metrics. Except for Expected Goals.
At the October break, these metrics showed a bigger gap with Rangers misfiring in the early part of the season. That performance gap is closing whilst, of course, their points gap in the league table is still a healthy 4 points.
Defensive Performance
Pundits like to question the Celtic backline. We all know the structural weaknesses inherent in the 20/21 squad.
The current cadre are a mixture of what we thought were cast offs (Ralston), loanees (Carter-Vickers), those struggling to adapt (Starfelt) and those fitting in where they can (Juranovic). Add in none of them can claim to be much over 6 foot, and it doesn’t seem a recipe for solidity!
Yet, (whisper it and don’t tell the papers) Celtic have conceded but 9 SPFL goals, 3 less than Heart of Midlothian and St Johnstone.
Do the underlying metrics support the strength of this patchwork defensive curtain?
And a fascinating picture it is!
Celtic do not have the dominance in defending that their newly forming attack has.
But the headline is they have the lowest xG against in the league at just 0.63, compared to 0.69 goals against. Celtic’s Post Shot xG (the xG value for all on target shots) is also the lowest.
They also let the opposition have the least number of Big Chances (defined by Opta as “A situation where a player should reasonably be expected to score, usually in a one-on-one scenario or from very close range when the ball has a clear path to goal and there is low to moderate pressure on the shooter. Penalties are always considered big chances.”).
Rangers' superior organisation and shape means they limit teams in terms of total shots, and those in the box and on target. They are also forced into the least Goalkeeping Saves. Shape and structure is the key to a good defence and they limit the opposition crossing more than other sides.
But there are a couple of outliers. Rangers are only ranked 4th on Big Chances given up, and following from that, the xG per shot they do concede is only the 9th lowest in the league.
What Now?
It is fascinating to compare what is essentially a brand-new team plus Callum McGregor versus a well-stocked, well drilled opponent.
Credit is due to Postecoglou for quickly crafting an incisive attacking style – over 3 xG per game. Celtic’s xG Difference (xG For minus xG Against) is 2.39 whilst for Rangers it is 1.26. Celtic’s average xG differential should be enough to account for unfriendly variance (e.g. defensive errors; the Livingston factor; Willie Collum) whilst a differential closer to 1 tends to result in the impacts of that variance leading to dropped points over the long term.
Rangers can rightly point to continuing defensive solidity, and they lead on most metrics across the league.
But Celtic are improving with much headroom for further enhancement.
Based on the last international break, we may see the fruits of that as early as 20th November.
Comments: Our rules
We want our comments to be a lively and valuable part of our community - a place where readers can debate and engage with the most important local issues. The ability to comment on our stories is a privilege, not a right, however, and that privilege may be withdrawn if it is abused or misused.
Please report any comments that break our rules.
Read the rules here