The “big picture” shows European football after the New Year. All macro objectives met. But our own eyes tell us Celtic’s performances have not always met the highs of last season. Celtic lost 5 matches all season in 16/17 and have lost 7 already this. To stand still is to inevitably go backwards. We seek continual, even if gradual, improvement in all aspects of the club. What does the performance data tell us, and how do we improve after the winter break?
The first half term report considers Celtic performances in European football. There will be a further article considering domestic league form.
Unfamiliar terms are beautifully explained in the family friendly Glossary.
Champions League Qualifying
With the Scottish Cup Final on May25th 2017, and the first qualifying match on 14th July, Celtic came back to preseason training on 19th June. However, Scotland played England on the 10th June alongside other World Cup internationals, meaning many of the squad had less than two weeks break.
The upside to this was that the squad was match ready by the time the qualifiers started. The raw results were certainly more impressive than the previous year.
More goals scored, less conceded, and overall 16% more successful than the previous season.
The 16/17 qualifying campaign saw Rodgers manage the side for the first time mainly with Deila’s squad. This was characterised by experimentation in defence with the likes of Janko, O’Connell and Ambrose featuring. It was also a period when Rodgers was trying to bring in and settle new signings. There can have been few evenings as fraught as Hapoel Be’er Sheva away. Celtic got over the line and the Rodgers reign had momentum.
The 17/18 campaign was not without personnel difficulties. There was no fit striker to start the Rosenborg ties. And central defensive experimentations continued, with Bitton struggling badly in Astana. Injuries can be down to misfortune, but Celtic must reflect on the squad management that allowed such crucial matches to be dogged with such uncertainty in key positions.
However, when comparing the fundamentals across both campaigns, a strong case can be made for year on year improvement.
The Rodgers possession based model is noticeable here. Possession overall went up on average by 10% which is significant. The sheer volume of passes completed went up by 38% which is a huge uplift. Meanwhile the number of passes incomplete went down leading to a 5% increase in pass completion, from 84% to 89%. This is in line with Celtic domestic performance which indicates the level of opposition to an extent. It all lead to greater control for Celtic in these games.
Not much use if you don’t create chances mind.
* I don’t have Expected Goals for the Lincoln Red Imps games.
The average number of xG per match increased slightly as did the total number of shots taken. However, despite 7% lesser shot accuracy, Celtic converted 10% more of their shots on target and 2% more of all shots. This is a significant indicator as we shall see. Celtic also benefitted from good old-fashioned luck in the 17/18 campaign. 3 of 15 goals were Own Goals helping lift Celtic’s average of 2.5 goals per match well above the xG model.
Meanwhile the opposition managed greater shot accuracy from an average of 1 more shot per game, but converted 8% less on target shots, and 4% less from all shots. 6% overall conversion is low and given that Celtic did not restrict the opposition to less shots, either Gordon was exceptional (can he sustain it?) or the opposition forwards were poor (will Celtic get as fortunate next time?) or they were poor quality chances? Or a combination. What would be good to see is a reduction in overall shots allowed, removing the skill of goalkeepers and forwards from the equation. But Celtic can expect a better on target conversion rate from opponents next season.
Defensively, there were several indicators of improved performance. The reduction in average Clearances per match is inactive of two factors: 1. Philosophical approach – playing out from the back and 2. Being under less pressure. Similarly, the number of defensive errors halved in this campaign. Gordon, Lustig, Tierney and Simunovic were virtually ever present in the 6 matches whilst the second centre back position was rotated around. Simunovic missed the defensive debacle in Astana.
A couple of points of interest. There are slightly less physical challenges this season as Celtic have had more possession control. This is probably a positive as we saw in the recent Derby where Celtic struggled when the game became fractured and a fight for possession every few seconds. Secondly, the opposition tend not to commit as many defensive errors as Celtic.
The Astana away game was the outlier but a real concern. No goals had been conceded up to that point. Suddenly, with qualification to be confirmed and a makeshift back 4, errors emerged and at 4-1 down genuine doubt as to whether qualification could be sealed. This was a troublesome performance. Given the maturity and solidity going away to Rosenborg and winning 1-0, it was a skittish and open Celtic. A mature, experienced European side would have closed that game out. With Ajer and Bitton holding the centre, they should have been given maximum protection. Brown had his worst match of the season, and it is telling how much of a hole this leaves in Celtic’s defensive capabilities. The Brown dependency is a real key man risk.
In summary, despite significantly more possession and more accurate passing, Celtic did not create significantly more shooting chances, nor were the chances of better quality than the previous season. Better finishing plus deflection-fortune saw Celtic through.
Defensively, Celtic were overall stronger as 5 clean sheets shows. However, the Astana away game is a worry not only in terms of individual performances, but the openness with which the team played in far away and hostile conditions. It was not a mature display individually or tactically.
But objective met and the balance sheet per 90m stats will never have looked better!
Champions’ League Group Stage
The strategic objective of achieving European football after the New Year was met. Following Celtic for as long as I have, you appreciate that. The club having achieved post New Year European football 6 times in the last 40 years. The Scottish press like to portray not qualifying in this regard as failure, yet clearly it is a relatively rare event.
Simplistically, 3 points won in the Group matches last year’s total.
Celtic played two financially loaded giants last season (Barcelona and Manchester City) and two this (Paris Saint-Germain and Bayern Munich). No wins were expected against such opponents containing some of the most talented (and well paid) players on the planet. Progress depended on games against the middle classes (Borussia Moenchengladbach and Anderlecht).
Having not qualified out the groups with 9 points before, Celtic will be grateful of qualifying with 3. Did they perform more effectively, or are Anderlecht weaker than the Germans?
Champions League football really is a whole new ball game for Celtic. In the qualifiers, they must initially face part time teams of the standard of e.g. Albion Rovers. Then the opposition immediately gets better than any team Celtic face in Scotland. But Celtic got through posting numbers commensurate with domestic performance.
Against the best teams in the world, Celtic go from 70% possession to under 50%, and completing 600+ passes to around 400. Number of passes completed, and completion rates went up slightly. Also, Celtic’s overall possession average increased. Evidence of greater control? Kind of. Against Paris Saint-Germain Celtic never reached 40% in either game. Against Bayern Munich it was a respectable 48% in a well fought game at Celtic Park and 40% in Munich. Against Anderlecht it was 60% in Belgium and 41% at home. Again, a low-key completion to the campaign influencing feelings of disconsolance.
In the below chart, the xG’s were calculated using different models. The one I use now has much lower overall values. That explains most of the differences.
Celtic managed one more shot on target than in 16/17, and managed to restrict their opponents to less shots overall and on target compared to the previous campaign. The worry is Celtic conceded more goals and as can be seen below, a whopping 44% of on target shots resulted in goals in both campaigns.
This is as stark a reminder of the quality differential to the SPFL as there can be. Celtic and their SPFL opponents score from 11% of all shots with 37% being on target. In the Champions League the conversion rate doubles from the opposition. Celtic are converting less than half that number, whilst having less shots as well. This is the challenge for many Champions League participants from non-Elite leagues. This is what Elite level strikers can do, the Neymars, Cavanis, Lewandowskis, Mullers and so on.
In a reverse of the Qualification campaign, it is Celtic’s conversion rate that collapsed to 6% of all shots and an 8% drop off in on target conversion. For Celtic to win at this level they need to over perform in converting the fewer chances, whilst keeping out the Elite strikers. Stats like above will result in 3, 5, 7 goal hammerings.
Defensively, Celtic force the opposition into 1/3 less Clearances, which is another 24 out of 29 times Celtic lose possession. The other obvious discrepancy is the defensive error rate once again. Of 16 defensive errors by opponents in the Group stage, 10 were by Anderlecht, 9 in Belgium. Therefore, the average for the two Elite clubs was considerably lower. But the overall average is consistent with the qualification opponents. Meaning Celtic’s defensive error rate is abnormally high. Whilst failing to put the opposition under similar pressure.
Tactically Celtic often lacked clarity of purpose. Rodgers wants his side to press high, and in Scotland that is achievable with high reward side outcomes due to Celtic having better players and the majority of possession. In Champions League, the likes of Motta, Alves, Silva were content to play around a high press with ease. That is not to say pressing is taken out the armoury. More that game intelligence and decision making of the first order are needed to decide when the press is on. For example, Celtic picked their moment correctly away in Paris to force the error that brought the opening goal.
At home to Munich was the high-water mark of Celtic performances. The balance between a deep low block defence, and appropriate pressing was closest to perfection in that match. Yet it wasn’t good enough. Poor game management at 1-1 saw defeat against weakened opponents.
At home to Paris Saint-Germain, Celtic made the game easy for the French side. They neither pressed aggressively nor fell back into a deep block. Motta completed over 140 passes and played through Celtics lines with ease. The nightmare can be reviewed in gory detail in Gone in 45.
Then, here is my opening line from Brutal Reality in Bavaria; “Celtic escaped a complete dismantling at the hands of the Billionaires Club of Bavaria, but a 3-0 reverse was a tough watch. Positives were difficult to find, and Rodgers must reflect on defensive shape and organisation once again.”. Rodgers must decide whether he is overly optimistic about his sides abilities and whether the players are capable of the tactical flexibility needed to combat the best opponents.
Celtic’s Group stage adventure was a mixed bag when compared to last season. The overall result was clearly superior, but familiar failings in defence make the improved possession redundant. Add to that the failure to deal with Elite level finishing, whilst Celtic’s own conversion rates fell off the cliff, meant an uphill struggle against opponents that are leaving middle class European sides far away in the rear-view mirror. Many fail to grasp how much of a gap there is between teams like Paris Saint-Germain with many truly world class players, and teams like Celtic from impoverished leagues. And yet football, as a low scoring game, partly takes its appeal from providing opportunity to lesser sides to battle equally with so called better sides. Celtic need to learn to be harder to beat and less open. Two years of hammerings should have made this obvious – there were echoes of the 0-7 reverse in the Nou Camp in both Paris matches and in Bavaria.
Ultimately Celtic progressed thanks to a 3-0 win against a dour, defensive and error strewn Anderlecht under a caretaker coach. To say there is work to do to improve team tactical organisation at this level is an understatement. Whether Rodgers thinks he has Champions League squad to achieve this remains to be seen and the forthcoming transfer windows will be telling.