SPFL Attacking : First 8 Games
As we are still in the international break, a checkpoint on the SPF attacking stats to date.
A Troubling Start?
On a recent A Celtic State of Mind bulletin I gave a bullish response to the question of what the data showed regarding how are Celtic performing.
Here is some more context to that.
Although Celtic sit 6th in the SPFL, 6 points behind the leaders with 3 defeats already, there is a huge amount of context missing from those bald facts.
Celtic are obviously building a whole new squad and a new system of operation under Postecoglou. Players have arrived from multiple locations and take time to adapt to lifestyle, culture and environment.
His start has been continually disrupted by injury and unavailability. Already 98 matches worth of appearances have been lost to injury amongst the first team squad.
And, dare I mention it, refereeing has not been kind with game changing errors in the defeat to Heart of Midlothian (a game leading goal wrongly disallowed) and Dundee United (a sending off not given in the first half). As I described in the Honest Mistake analysis so far this season Celtic have 1.4 less points than expected without refereeing errors, and The Rangers 1.24 more, a difference of 2.64.
Those are not excuses, again just facts to balance the league position.
And remember, as Expected Goals has taught us, the league table DOES lie and xG based analysis is a better indicator of performance, and predictor of future performance. Ask Matthew Benham.
The Celtic Analysis account on Twitter was ridiculed by some for producing a league table based on xG outcomes. This is the same methodology used by Modern Fitba for the last few seasons which correctly showed that The Rangers were closer to Celtic that the league table said in 19/20 and showed early the gulf between the clubs in 20/21 that manifest as a 25-point gap by seasons end.
The methodology is sound. Currently is shows Celtic top of the league on an xG basis.
What I will balance that with is that Celtic have had two massive xG chance fests in hammering Dundee and St Mirren. In a small sample of 8 games that will skew things.
Without those two matches, Celtic have still “won” the xG battle in every match with an aggregate of 12.73 – 4.65.
SPFL Attacking Data
Let’s take a closer look at the SPFL attacking stats so far.
xG
Celtic are way ahead on xG and Goals scored. If fact, nearly double the nearest challenger for xG. Remove the 6+ Dundee and St Mirren blowouts and Celtic’s xG average would still be 2.12 and easily the highest in the league.
Despite a dearth of attacking talent to chose from, Postecoglou has Celtic creating ample chances to win games.
Indeed, Celtic are the only ones in the top 6 apart from Dundee United underperforming versus their xG:
Hearts are seeing significant positive variance to their goals scored versus expected. Whilst Celtic are underscoring per game at the rate of almost 1 goal versus expected (0.86), which surely cannot continue!?
In terms of average shot quality:
Celtic are taking significantly better-quality shots than the rest of the league – averaging 0.16 xG oer shot. This is 0.05 better than most / shot. Take 20 shots in a game that equates to 1 more xG which is obviously significant in a low scoring sport.
Finally, Post Shot xG – this is the xG the goalkeeper actually faces – i.e. the xG of all shots on target or that are goals.
By this metric, again Celtic are nearly 1 goal per game higher in shot quality than their nearest opponent – Hibernian.
Shots
Why do I mention 20 shots per game?
Celtic are averaging nearly 3 more shots that The Rangers, and double half the league. They have 2.23 more on target and whilst The Rangers have the best In / Out box ratio for shots, Celtic have a higher volume of shots inside the opponents box.
Hearts, in particular, are overperforming relative to their attacking output.
Opposition Goalkeeper Saves
This shows the number of saves the opposition goalkeeper has to save on average:
Celtic are forcing the opposition keeper into 1 more save per 90m than any other team. Both Dundee clubs may reflect they need to stop hitting the ball at the custodian too!
Summary
Based on xG and shot data, Celtic are performing far more effectively than the rest of the league despite attacking talent deficits. This bodes well for the rest of the season when number of attackers available increases.
If this performance level can be maintained, and you would think this has been the worst of Celtic so far, then fans can expect a lot of goals.
And finally, don’t “happy clapper” label me! It was studying the same metrics as above that showed early last season Celtic were in real trouble relative to The Rangers, and why on the Huddle Breakdown (please subscribe), we were voicing concerns from August.
And that wasn’t popular either!
I will consider the league Creativity and Defensive data shortly.