The treacherous nature of playing teams half way through their season, in July, was amplified as Celtic struggled against a well organised Norwegian champion. Off the field events had dominated the headlines in the build-up, but on the field, there were worrying echoes of the Deila reign in Europe as the game started slowly, then got worse.
By the end, Celtic looked like they were swimming through treacle against limited but sprightly opponents. Some familiar traits came to the fore.
The Horseshoe of Futility
With a nod to 11tegen11, Celtic’s ultimately comfortable 2-0 win at Linfield was characterised by many circuits around the horseshoe of futility. That is, recycling the ball across the midfield and down the flanks, and back again. Like Linfield, Rosenborg deployed a low, deep block defence in a 4-5-1 shape. They kept the “Lord” Bendtner up front on his own. Rarely leaving the width of the penalty area, and unconcerned about isolation from midfield colleagues, the ex-Arsenal man managed to distract the Celtic defence without pressing, nor moving much at all really.
Again, note this shows all Pack passes only – i.e. passes that took at least one opponent out the play. It does not show ALL passes. Celtic passed the ball 445 times in the first half without creating a single chance of note. The thinnest lines above represent 1 completed pass in the connection. Rogic managed to connect with more players than Griffiths or Demeble tend to do, but dropped deeper to achieve that. Brown sprayed passes hither and thither. Lustig and Tierney were generally available, but time and again turned back and recycled the ball backwards. The normally telepathic Tierney and Sinclair connected 3 times to any effect. Armstrong, Rogic, Forrest and Sinclair, the creative leads, completed 77 passes between them in the first half compared to Brown’s 89.
With Ajer, Simunovic and Brown completing 207 of the 445 Celtic passes, Rosenborg forced Celtic into a pattern of regressive play. Such concentration of possession in the middle, deep territory is usually a “bad thing”. Reminiscent of the Brown and Bitton axis. Out of possession, the Norwegians were rigid and disciplined in their shape.
Hang on, wasn’t that last week?
By half time, Celtic had crafted a single half chance for Forrest to head weakly on target, and an Ajer header from a corner. The dogged Norwegians held their shape and blocked – 6 shots in the first half, 4 by the redoubtable Skjelvik. It wasn’t just the lack of a focal striking option, Sinclair and Forrest were isolated. The one ray of light saw Ntcham take out 3 defenders and all 5 midfielders to find Armstrong in the box, but he was soon crowded out.
What A Flogged Horse Looks Like
Surely the second half would be better? Rodgers will work it out. It wasn’t. The diagram above shows the full 45 minutes. The meaningful connections all but dried up with Sinclair virtually redundant – he completed 4 passes from 5 attempts. Hayes was tried at striker and possibly Sinclair later on – it was hard to tell. Of 121 passes that took out at least one opponent, 42 took out only 1. Usually it was the ambling Bendtner, who watched glibly as he waited to fight for the next high ball. This is a tad unfair, as in Jensen, Konradson and Midtjso, Rosenborg had three midfielders capable of sliding passes through the Celtic defence when not hounding Armstrong, Ntcham and Rogic and preventing meaningful forward progression. Only 17% of Celtic’s passes took out an opponent whilst the guests managed to “pack” Celtic players from 26% of their attempts.
Rosenborg were a significant step up from Linfield (think Albion Rovers), but their overall stats would be equivalent to a St Johnstone or Aberdeen on a good day at Celtic Park. They completed 209 passes and their PEI of 66% compares with Aberdeen averaging 196 passes and 57% PEI last season against Celtic. The difference their match fitness made, and Celtic’s apparent lack of depth off the bench, meant Celtic regressed as the match wore on. Ntcham, Benyu and Hayes are bound to take time to integrate. Game changers are needed from the bench if Champions League standards are to be reached. Still, it was strange not to see the direct running of McGregor utilised.
Why We Are Here
The use of so-called “advanced” statistics, like xG and Packing/Impect, highlight the misleading nature of many “traditional” statistics. If we consider some of the common metrics used to illustrate this match, it would appear Celtic were über dominant.
So what?
And now we’ll consider a different set of metrics:
That’s more like it!
A far more balanced view emerges. By the xG metric, Celtic were lucky to escape. Rosenborg had 5 shots on target to Celtic’s 3, with the xG of 0.64 – 1.01. The Norwegians created the one Big Chance of the game when de Lanlay scooped the ball over the bar from 8 yards. The xG statistic is a useful antidote to being fooled that possession of 81%-19% renders the ball-hoggers truly dominant.
Whilst Celtic managed 10 shots inside the box, Rosenborg managed 8. Sure, Celtic created more shooting opportunities reflected in the xA (Expected Assists) of 0.92 – 0.62 but 10 of Celtic’s 18 shots were blocked. This reflects good defensive organisation and positioning, as well as poor shot selection / execution.
There is enough here for Celtic to feel they were dominant – the ability to maintain possession overall reflected in the PEI. Are Rosenborg encouraged by their chance to possession ratio, or disheartened not to have scored?
Captain Marmite
There were audible disgruntlements on several occasions when Brown played passes too long to the flanks, and twice, when he intercepted long kick outs despite better placed colleagues sweeping behind. With Bendtner content to watch Celtic try their worst to break the defence down, and no obvious number 10, Rosenborg ceded possession to Brown especially. Sinclair and Forrest struggled to find any space, and Rogic tended to drift to familiar territory and so the central of their defence could remain compact.
It was left to Brown to set the tempo and try and move the defence. I have never recorded such a dominant individual performance in terms of the amount of possession Brown had. He represents 19% of all Celtic’s possession events. His 145 completed passes are the highest I have recorded for a single player in 4 seasons. The next highest was 89 and he was 63% higher than that. This is not a positive as Brown is not the creative force of the team. But he manfully tried to switch the play, as well as undertaking his usual defensive midfield jobs.
He was the team’s most successful defender, and beat more opponents with his passing than any other player. On the other side, he also inadvertently set up Rosenborg’s best chance with a loose back pass.
But overall, he led the team as best he could to try and get forward, even having a shot on target in the last minute. Having to play riskier passes than he normally attempts was not his fault given the failure of others to break the lines consistently.
He was my Bhoy of the Match, by miles.
The Norwegian Spruce
Despite only 30 minutes of Celtic first team action, at home to Lincoln Red Imps last season, Ajer slotted into the side with ease. Tall, upright, certain. Perhaps emboldened by playing his fellow countrymen, there were no nerves defensively. He led the team committing 3 fouls, the victim of cunning opponents, height and overzealous reaching for headed challenges!
Such was the ease of gait, his relaxed air and balanced passing game, the only regret is he did not stride out from the back more, and vary the angle and intensity of the initial attacks, giving his captain a breather. He covered his full backs well, especially the roaming Tierney, who gets caught in some terrible defensive positions occasionally, by blocking 2 crosses. An accomplished passer of the ball, it appears there is a lot more to come from the 19-year-old Norwegian.
Keep the Faith
A confident Rodgers seems sure that Celtic will be better in the second leg. No goals conceded at home renders the tie evenly matched for next week. Another week of fitness, and the discovery of a striker in the squad will assist. They have certainly lived to fight another day. Celtic must be reborn with an intuition of immortality.
Patrick says
Your point about them sitting off Brown is an important one. I remember last season when Caixinha’s Rangers were roundly criticised in the media for allowing Brown to run the game in their 2-0 defeat at Hampden. The consensus seemed to be that getting tight to Brown could derail Celtic. In fact, if you’re going to be on the back foot against Celtic, allowing Brown time on the ball is probably the way to go. He’s not particularly good at picking passes, playing long switches, or shooting. As much as Bitton gets (deserved) criticism, I think he’d have done better on Wednesday night, because he’s much more of a threat- both passing and shooting- when given time.
I also thought Rodgers made a mistake in how he deployed Lustig. Because there was no striker, Forrest was encouraged to play higher and more centrally, and Lustig to be more of an orthodox fullback, rather than part of a back 3. This, for me, is counterproductive in two respects. It meant that Forrest was denied the chance to pick the ball up relatively deep and drive at a fullbacck with space behind him, which is the situation that best suits his attributes. And it resulted in Lustig receiving the ball high up, expected almost to play as a right-sided equivalent of Tierney, which isn’t his game because he can’t beat his man or drive to the byline.
Over all, Rodgers tried to make a virtue of necessity; gaining a central midfielder because he didn’t have a striker. This was a mistake, though, because it affected the roles of pretty much every player from middle to front. Instead of being a good team with what would’ve been a crap, makeshift, striker. Celtic became a crap, makeshift, team. Celtic would’ve been far better served, in my opinion, by just telling Jonny Hayes to spend the 90 minutes doing his best impression of a number 9; occupy the centre backs, threaten runs in behind, and close down aggressively. If he could’ve taken it with his back to goal and linked the play, or even nicked a goal, then great. But the important thing would’ve been to allow Forrest, Rogic, Sinclair and Armstrong to play something like their natural games. I also think the decision to start Ntcham, though he did alright, was just another variable in a team that was already likely to be somewhat thrown off its natural rhythm.
If Griffiths is out of the next leg, I hope Rodgers uses the template of last year’s team and plugs the gaps where they appear, rather than attempting to restructure it to accommodate the absentees.
celticbynumbers@btinternet.com says
Patrick – I’ve been away a few days and couldn’t do justice a reply to your thoughtful post on a phone. Thanks for taking the time to reply.
I agree with your overall point that you shouldn’t make lots of changes to cover one weakness. For me this especially means allowing Sinclair to do his thing in that insider left rail. Just don’t try and fiddle with that. Similarly Rogic – he is a 10, he plays in between the lines and is hugely effective receiving the ball on the half turn with a fraction of time. Forrest has more variety and is the only one we have that I believe can do a job as a striker. See the article I have just posted (Striker Light). I can’t see Hayes having the tools to do that job, but accept not seen much evidence either way.
Interesting points about Lusting. I did not really notice that. If you see my passing maps, I thought Lustig connected well with Forrest – its just that nothing happened when either Sinclair or Forrest got the ball – they were effectively marshalled.
I thought Ntcham deserved another chance and he did ok again. He is young tho and McGregor has more Euro pedigree and experience scoring away in Europe and would be a better bet for the 2nd leg. Thanks again and hope to see more contributions!
Tommy E says
Hello, Rosenborg fan here. Interesting analysis, RBK will feel hard done by not to get that away goal. Without the away goal and our best defender (Refiniussen), I fear the probability for qualifying for the next round is quite correctly mirrored in the odds – i.e. around 65 % for Celtic and 35 % for RBK.
I saw an xG analysis from @11tegen11 on Twitter with quite different xG numbers (0.93-1.6). What is your source for these stats (especially xG)?
celticbynumbers@btinternet.com says
Great to hear from you. The xG model I use is very simplistic and available to all. See the Contact Us page for the link. I agree re Reginiussen he looked solid though hardly put under pressure. It looks 50-50 to me! All the best .
Andrew says
Thank you celtic numbers for yet more wonderful coverage of Scottish football. Why we don’t have more of this type of conversation on MSM is beyond me!
It doesn’t cost the earth and it gives a better insight into the modern game than any ex pro (with vested interests) could do in 40 “it’s a game of two halfs” summations.
(I’m sure we are all aware of the usual shock jock types on both side of the border that really should be left to pasture).
Scottish football would be best minded to use more of (dare I say it) an American influence on the promotion of our national sport: “Stats and stripes” all the way for me!
Keep up the good work.
celticbynumbers@btinternet.com says
Cheers Andrew very kind. It doesn’t cost the earth because I don’t charge by the hour!!! ?