By James Dailey
Saturday’s 5-0 result versus St. Mirren was a continuation of domestic dominance this season. Advanced analytics and related statistics have more value over larger sample sizes, as normal variance in a single game, or even a group of games, can deviate significantly from underlying metrics. Clearly a 5-0 result is better than a 3-1 result when playing a comparable opponent, but results are not the same as performance levels. Over time, results tend to mean revert towards underlying performance metrics. Today I am going to do an in-depth comparative analysis of Saturday’s performance with the recent 3-1 victory over Kilmarnock.
I’ve been consistently on record about my preference for playing Callum McGregor in place of Scott Brown at the central defensive midfield position, but this shift is only partly about that direct impact. As I’ll document, I believe it is just as important for Celtic’s midfield to play Ryan Christie as a box to box attacking 8.
Kilmarnock / St Mirren
First, let’s review the team-level stats for the two games:

While the resulting 5-0 score suggests a dominant performance, the underlying data (excluding penalties) and metrics are quite clear that the performance in the Kilmarnock game was significantly better. I think it is noteworthy that the dominance was comprehensive with complete control of the game, both in attack and defensively. Besides the obvious disparity in xG (for context the 7-0 drubbing in 2016-17 to Barcelona was 2.91 to 0.05), I think the Passes Per Defensive Action (PPDA), which is a proxy for pressing intensity, is a real standout, and is reflected in the higher level of Interceptions and Recoveries as a team.
Brown / McGregor
Here are metrics comparing the performances of Scott Brown in the St. Mirren game versus Callum McGregor in the Kilmarnock game. While Neil Lennon has declared Brown fit, in fairness I think the following comparison should be considered in the context of the potential that declaration is not accurate:

As you can see, outside of Recoveries, the rest of the data pretty much speaks for itself. Perhaps the most telling stat to me is the contrast in number of passes straight back to Jullien. Also, McGregor excels at making himself available with elite movement, which is reflected in his incredibly high number of passes received.
McGregor / Christie
I do not conduct these analysis or comparisons to “tear down” players, but rather try to assess how our squad and players can be better utilized. This is highlighted in comparing McGregor’s performance as an attacking 8 versus St. Mirren with Ryan Christie fulfilling largely the same role versus Kilmarnock. I think McGregor is a terrific player no matter where he plays in midfield – the question is where should he play to have our best midfield as a unit?

What stands out to me and is likely not a surprise to anyone who regularly watches Celtic, is that Christie runs around like a lunatic – and I mean that in a good way. His work rate and “engine” are tremendous, and I believe perfectly suited for a box to box midfielder. His volume of engaging in duels and creative activity attacking the opponent’s box are excellent. His 9 defensive duels are higher than Scott Brown’s career per game average of 8, just to place that into perspective. His 11 Progressive Runs compare to Stuart Armstrong’s career best at Celtic of 6. His running and defensive cover are superior to McGregor’s which offers a combination where he and McGregor as a 6 are better defensively than the combination of Brown and McGregor.
Conclusion
While my preference is for a fit Olivier Ntcham to play as the right sided 8 alongside McGregor at the 6 and Christie as the left sided 8, I believe the key is the McGregor-Christie combination. McGregor is a very good attacking 8, but Christie is better. Brown, at this stage in his career, is a good defensive midfielder, but McGregor is better. I’d love to see the two play in the best positions for overall team performance at Ibrox – a Bhoy can dream…maybe next season!