League Cup Final: Preview
This special edition reviews the current questions either Glasgow club needs to ask themselves ahead of the League Cup final on Sunday.
Please Note: This post may be too long for Gmail users. If the post is cut short due to your email provider, the rest can be viewed at stephenrussell.substack.com. Alternatively, use the Substack app here:
Contents
A Message of Solidarity
League Cup Final: Preview
Celtic 4-0 Aberdeen
Livingston 0-3 Rangers
Hibernian 2-0 Kilmarnock
Motherwell 2-0 Hearts
Dundee United 1-2 St. Johnstone
St. Mirren 1-0 Ross County
League Update
A Message of Solidarity
Content Warning: Transphobia
As the trans community has already been attacked relentlessly from media outlets, high profile celebrities, and several targeted campaigns, I was extremely upset to see such transphobia demonstrated outside of Celtic Park before the Aberdeen match. Hiding behind the guise of ‘protecting’ single sex spaces, a transphobic group felt it appropriate to canvas the stadium with explicitly bigoted messages that sought to deny the existence of the trans community.
I contacted Celtic’s Supporter Liaison Officer with my concerns. His reply reassured me that Celtic were not asked permission, would not have given permission if asked, and subsequently asked the group to leave. Read in full below.
This newsletter exists for tactical analysis, and this is the first inclusion of a non-analytical message. However, I’m an analyst second to being a person. I’m a person who is upset and enraged at the continued marginalisation of the trans community and, thankfully, I’m not the only one.
While the group themselves claimed ‘overwhelming support’ on social media, I think it’s clear from myself, the club, and a long list of match attendees that this was not - and will never be - the case. Transphobes are not welcome in society, much less at Celtic Park, and my support for the trans community will forever remain unwavering.
Solidarity, comrades 🏳️⚧️
League Cup Final: Preview
Celtic and Rangers both cruised to victory last week, extending their mutual runs of consecutive wins in all competitions since the derby draw in January. One run must end on Sunday, though, as the pair face off in the League Cup final. This analysis will outline the context surrounding each team, and then detail the questions each club needs to ask in their preparation for the match.
Celtic
The most prevalent selection topic surrounding Celtic ahead of this match is the right sided #8. Despite his reduced number of starts recently, Matt O’Riley’s inclusion over Aaron Mooy is worth arguing for. The first halves against Livingston and Aberdeen, his two most recent starts, resulted in some of the best football Celtic have played since the last derby. It would be remiss to attribute this solely to the quality of each respective player, but the impact on surrounding players must be considered. This previous edition of this newsletter analyses the impact on Celtic’s fluidity and incision on the right hand side.
O’Riley once again proved his value in this role against Aberdeen. Receiving the ball frequently in the right halfspace resulted in him having a positive Expected Thread (xT) received of 0.60 - the highest in the match. For comparison, Hyeon-Gyu Oh received 0.53 - the next highest - followed by Reo Hatate’s 0.37. Through the season, O’Riley has received 0.35 positive xT per 90 minutes played (p90) contrasted with Mooy’s 0.14, of which only several defenders (and James Forrest) score lower.
While receiving the ball more centrally is beneficial to the players around him, O’Riley also passes into dangerous areas frequently. For the season, O’Riley has generated 0.48 positive xT p90. Mooy comes in at 0.38. Both players take set pieces with differing contexts so, with these removed, O’Riley scores 0.33 - slightly ahead of Mooy’s 0.30.
Positive xT is used to not punish players for back passes. Possession value models see these as an action that negatively impacts a team’s chance of scoring but, as highlighted in this past newsletter, this is still sometimes the correct decision. However, good data analysis is about bringing together information to collectively paint a picture. When negatives are included, O’Riley’s xT p90 drops by 0.48 to 0.00 - essentially breaking even with ‘productive’ passes weighed against the rest - while Mooy’s drops 0.52 to -0.13.
On the pitch, this overall supports the ideas that have been previously highlighted. O’Riley is less passive in possession, but - most critically - receives the ball in significantly more dangerous areas, to the team’s overall benefit. Against Rangers, a team specialising in compactness and restriction of space, it appears that O’Riley is the more suitable choice.
Another debate this newsletter has recently discussed is the presence of Jota on the right. Especially considering the above context of Celtic’s rigid right hand side, his dynamism is a welcome commodity. The Aberdeen match also highlighted this, with him and O’Riley starting the game with incredible fluidity and incision. This made Celtic much more balanced and not reliant on their left, as they are quite frequently.
While several factors play into Celtic’s left-bias - left footed Callum McGregor having a more natural inclination to favour that side, generally having a right footed centre back on the left, and Kyogo Furuhashi preferring that side to drop out and receive the ball - one factor is simply the quality of players on that side. Removing outliers boosted by low minutes (namely Sead Hakšabanović, Alexandro Bernabei, and Oh), Celtic’s top three players through the season for open play positive xT p90 are Jota (0.53), Greg Taylor (0.41), and Hatate (0.33). The progression abilities of all three are undeniable, with this season bringing some electric interplay between the trio.
From this, the question becomes which wingers Celtic should start and on which side. Borna Barisic still doesn’t seem to have realised that Liel Abada likes to arrive in the box to cutbacks from the left and has punished Rangers for this repeatedly, but the Israeli winger offers very little outside of his directness. He ranks the lowest of all players that have played for Celtic in the league this season for xT p90 (-0.57) and this still doesn’t compete with Celtic’s other creators when negatives are removed (0.18). Daizen Maeda, for contrast, scores 0.04 and 0.29 respectively.
One hallmark of Rangers’ system under Michael Beale is very high fullbacks, with no other players offering width. That creates the opportunity for Celtic’s wingers to become very dangerous in transition. If you can allow Abada the license to remain high and be ready to run onto the ball, he may be able to partner Furuhashi in a direct role. However, the 2-2 rest defence that Rangers use to cover for the fullbacks pushing high means a good ball carrier able to drive centrally, receiving wider and deeper, and engage the block could be more useful. This is more suited to Maeda.
This leads onto whether Jota should start on the left or right. Maeda’s pace poses more of a threat to the declining James Tavernier, and Jota on the right (with O’Riley) creates a much more balanced and fluid attacking machine. However, if the aim is to emulate Abada’s late arrivals into the box when play builds up on the left then a switch could actually be useful. Maeda’s box incision is astute, having played similar roles in Japan, and Jota paired with Hatate and Taylor on the left has been the setup for some of Celtic’s best patterns of play this season.
Another consideration is which winger the management team prefer to be inverted. As mentioned in the previous newsletter discussing this, Jota is more selective of dribbles/shots when he isn’t coming in on his left from the right. The counter-argument would be that Maeda’s style suits getting to the line to provide cutbacks and playing on the right may be more suitable for that.
Overall, the case for Jota and Maeda to start together is very strong - and likely given recent lineups - and also provides a great deal of tactical flexibility based simply on them both being comfortable on either side.
Rangers
There are two key debates regarding Rangers’ team selection: who starts in the two pivot roles, and does Malik Tillman earn a position somewhere.
John Lundstram went off injured against Ross County at the start of the month and it is unclear whether he could be fit to start. He started alongside Ryan Jack (also in a race to be fit) in that match, and - in the first ten minutes before he went off - a couple of interesting decisions could be seen.
To understand the pivot situation, it’s important to know what Rangers are currently doing ahead of the pivots. Recently, Rangers’ formation in build up looks similar to a 4-2-2-2 shape. Fashion Sakala and Alfredo Morelos appear as two forwards and Cantwell and Ryan Kent sit behind them. The latter two play in extremely fluid roles; Cantwell drops back and aims to aid with link up play in build up while Kent seeks space and then looks to progress the ball in a more attacking way if received. What this fluidity also does, however, is open passing lanes to the forwards. If Cantwell and Kent are marked closely, the forwards find space, but if they aren’t then two of Rangers’ better creators are able to find gaps within the opposition shape and can cause chaos. The below graphic provides an overview of how this can look.
This means that Rangers’ pivots need to be comfortable evading the front line of a press while also able to play progressive passes well. Both Lundstram and Jack are generally fine at this - 0.25 and 0.23 positive xT p90 respectively - and Glen Kamara (while ranking the highest) is only marginally better at 0.27. Lundstram generally holds up well in transitions, generally dictating the direction of play and defending well, but - if missing - it isn’t a problem to be swapped out. In fact, Rangers’ extreme compactness removes from individuality in a lot of situations so even players that might be worse in duels are fine as they have immediate support; controlling space is key.
Nicolas Raskin, however, started in Rangers’ most recent match. With only 95 minutes in the league, it’s difficult to draw any conclusions from his statistical performance so far. Similarly, it’s difficult to see how he fits into the Rangers team against Celtic. While he appears to favour progression and seems attacking in nature, Rangers’ build up against Celtic has often been much more direct with Beale’s involvement. The aim seems to be more on gaining territory and being able to have attacking players swarm the target to pick up the second ball/counterpress. This might have been personnel related, and Raskin might be trusted to progress the ball more in his natural style if he plays, or this could be an interesting test. Regardless, his high energy style might help Rangers to keep up with Celtic’s quick ball movement when Rangers defend which could make him a starter regardless of Lundstram and Jack’s fitness.
Another interesting point from the County match is that Kamara did not replace Lundstram in a like for like sub. Beale instead allowed Tillman to enter the match, and Jack ended up essentially a single pivot supported by Tillman, Cantwell, and occasionally Kent dropping back. While it would be a huge shock if Rangers took such a cavalier approach against Celtic, it highlights the other big decision to be made: should Tillman play?
Ignoring the low minutes Adam Devine and Bailey Rice, Kent and Cantwell are Rangers’ second and third highest for open play positive xT p90 (each on 0.40) - behind Barisic, the league leader for xT (although traditional fullbacks are generally favoured by possession value models due to the high volume of crosses). Kent - utilised much more effectively under Beale and a manager favourite - is unlikely to be dropped.
In contrast to Kent and Cantwell, Tillman scores quite poorly on open play positive xT p90 (0.14). This figure, however, is much more in line with Rangers’ strikers; Morelos scores 0.13 and Sakala scores 0.11. This similarity highlights the essence of the debate: do Rangers want a direct player to support the attack (Tillman) or a link up player to support build up (Cantwell)? The former might be able to engage Celtic’s Greg Taylor in 1v1s - a weakness of his - and be another facet of incision in transitions to disrupt Celtic’s two centre backs while the latter’s movement and awareness would help Rangers retain possession should they decide not to favour the usual direct approach adopted against Celtic.
While the pivot situation may be forced by injury, the Tillman decision will be a good indicator of how Beale wants to approach the final, and how he views Celtic’s weaknesses.
What a preview and review!