In my last piece, I reported on the Celtic trends relating to expected goals difference over Ange Postecoglou’s two seasons and last under Brendan Rodgers and explained why that matters.

For this one, I will focus on trends in pack passing. Pack passes are forward passes that remove opponents from the game – that is, they are now on the “wrong” side of the ball relative to defending their goal. Firstly, is there a relationship between expected goals (the quality of chances created) and pack passing (the degree to which you can “play through” your opponent's lines of defence?

xG and packing

Comparative data for both goes back to the start of the 2019-20 season.

You can see that the line shapes are very similar as regards the peaks and troughs. The r-Score indicates the correlation between the two variables (six match rolling xG difference and six match pack passing score rolling average) and is 0.46. That is a moderate correlation.

Unpacking the trends in terms of what happens on the pitch. Under Neil Lennon the xG trended above the packing score. Simplistically, a more direct style of football would serve to suppress the pass pack score. Although “direct” implies getting through the lines quicker, the result of this style is less accurate passing and fewer pack passes overall. The successful pack passes may occasionally generate a higher pass pack score given the greater directness.

The emphasis on line-breaking pack passing was evident as a key component of “Angeball”. Pack passing scores have increased as did xG in the last two seasons. Under Rodgers, the pack passing levels have broadly been maintained but the xG generated was more volatile. Hint: the quality of the wing play, and the failure to optimise attacks to serve Kyogo Furuhashi’s strengths.

Read more: 

Another consideration is the increasing trend amongst opponents to adopt a low block versus Celtic. This suppresses pack passing given the relative lack of space between the defensive lines. Occasionally it suppresses xG also, but usually not.

In the 42 matches Celtic has a lower pack pass score than their opponents, their record is Won 10 Drawn five Lost 27. That’s a 28 per cent points won rate. In 64 matches where Celtic has a lower xG than their opponents, their record is Won 20 Drawn 11 Lost 33. That’s a 37 per cent points won rate.

In 351 matches Celtic has posted a higher pack passing score than their opponents, they have lost just 38 of those matches – 11 per cent. In 329 matches Celtic has a higher xG than their opponents, they have lost only 32 matches – 10 per cent. xG is more volatile than pack passing, given there are far fewer shots per game (17.8 by Celtic) than pack passes (71 by Celtic).

More study would be needed, but based on Celtic history, winning the pack passing battle MAY be more predictive of success than winning the xG battle. So, long story short, pack passing matters.

Pack passing trends

Here are the trends for and against in terms of pack passing score over a six-match rolling average in the SPFL.

Both for and against, pack passing scores have been increasing over time. This may be a football trend thing. I know Celtic are aware of the packing concept internally, and it may be something clubs track. The power of it, in terms of measuring both player effectiveness and team performance effectiveness, may result in more teams focussing on achieving favourable packing outcomes.

Celtic’s pack passing scores since 2017 have increased far more aggressively compared to the average opposition scores. Aside from a clear dip during the nadir of 2020, pack passing was becoming more prevalent, peaking as a cornerstone of “Angeball”. It plateaued under Rodgers last season with a slight tweak in emphasis on control and possession over aggressive direct passing.

Opposition packing declined under Postecoglou due to sides adopting low blocks therefore being too deep to effect line-breaking moves. Also, Celtic’s counter-pressing has become ever more aggressive. As mentioned often on the Huddle Breakdown podcast, opponents were overwhelmed by “Angeball” and could not mount sustained attacks in the rare gaps between attacking onslaughts.

Opposition packing success is as productive now as it was at the end of 2020-21 under Lennon. Celtic do need to become harder to play through.

Pack passing difference

The difference between Celtic and their opponents in the SPFL is highlighted below.

The overall trend appears positive with the difference in pack passing score peaking at around 375 per game on a six-match rolling average basis under Postecoglou. Under Rodgers in 2017, that average difference could be as low as 50 to 100. By the end of last season, the average had fallen back to below 200 difference.

I am not sure this is down to a change in emphasis, but more likely down to the quality of personnel. Midfield is the key area here. Celtic lost Aaron Mooy, a talented passer with a terrific range. Paulo Bernardo was noticeably weak on pack passing and Matt O’Riley’s role emphasised getting on the end of those passes, not initiating them. Reo Hatate, a creative passer, was missing for large parts of the season. Finally, Callum McGregor’s creative performance data is declining year-on-year with age.

Summary

The warning signs are there. Celtic’s pack passing relative to their domestic opponents, had fallen back to pre-Postecoglou levels. I don’t believe there is a huge difference in playing style between Postecoglou and what Rodgers wants to achieve.

Celtic needs to upskill in the pack passing department, mainly in central midfield but also from the defensive positions.