We’re back because this is important. Calls for the sacking of a manager, especially one as decorated and with the storied history with Celtic that Lennon has, deserves special consideration.
Celtic’s form this season, by anyone’s eyes, has been disjointed and laboured. Recent results, a supine derby defeat, a rudderless first half against AC Milan, and finally a badly managed draw at Pittodrie have sent many to the conclusion the managers time is up.
That’s a big step.
Is it justified?
A reminder this site was set up, amongst many reasons, to provide an alternative to the binary viewpoint common within the confines of social media – a player/manager/team is either totally crap or world beaters.
Secondly, I focus on performance not results. The reason for this is performance is predictive. Results, in a very low scoring sport, less so.
Thirdly, I will present A LOT of graphs in this article. No apologies because Lennon deserves thorough consideration. I ask you to focus on the trends not the detail.
I have mapped key performance indicators at team level going back over three seasons.
We start our analysis at the beginning of Rodgers season two (17/18). The trends are 6-match averages from the start of 2017/18 season until now covering 190 matches.
Another important point – this data presented for THIS season DOES NOT include the last 3 matches This is largely irrelevant to the direction of the trends, however!
Do the numbers suggest Lennon be sacked?
Attacking
xNPG
Expected Non-Penalty Goals is a mouthful but essentially is the expected goals (xG) minus penalties which are although quite rare (for most teams at any rate) can distort total xG due to their high value.
Here is the 6-match trended average going back to 17/18:
As you can see this goes up and down a lot – really it is dependent on strength of opposition in the main. You can probably spot the European game tranches. What is key is the trend shown by the black dotted line. This is a clear upward trend in Celtic’s xG.
Let’s look at this season:
Even without the last 3 matches where we know shots on target were like hen’s teeth, the trend for this season is regression.
Shots On Target
Shots On Target are the most important kind as obviously without it you rarely score.
Trend over the last 3 seasons:
Not surprisingly, given xNPG, the trend for Shots On Target is a similar pattern- gradual improvement in the 6 match average over the time since Rodger’s second season.
This season:
This season, up to The Rangers match, this was pretty flat.
Touches In The Box
This measures the number of possessions by Celtic players in the opposition box per 90m as a 6-match trended average.
Once again, the trend under Lennon has been for Celtic to be getting more touches in the opposition box per 90m over time.
And this season:
This season, the trend is the opposite to that desired albeit still slightly above Rodgers season two levels.
xG / Shot
xG per shot details the quality of the average shot Celtic are taking. If, for example, there were more shots but more from outside the box then you’d expect this to go down.
Possibly a surprise to many (and me!) under Lennon xG per shot has been rising. Despite Rodger’s team seemingly not shooting until the perfect chance presented itself, and Lennon seemingly encouraging his side to shoot on sight, you might imagine the trend going the other way. That’s why I’m here!
This season?
This season, once again, the trend is reversed with average shot quality per shot reducing over time.
Creativity
We’ll now consider a number of measures that indicate the creativity of the team. Again, we’ll look at 6 match averages since 17/18 then look at this season.
Impect
Impect measures the ability of the team to break opposition line – that is take opponents out the game with passes and running with the ball. The higher the number the more the team are breaking through opponent’s organisation.
Again, this is highly variable based on quality of opposition. So, again, focus on the trend line (black dot) which is that Celtic have been increasingly able to break down opponents over Lennon’s tenure.
This season:
This season, again caveated with the fact this doesn’t include the last 3 games, is the Celtic’s ability to break down opponents’ lines has thus far maintained a modest improvement.
Chances Created
Very simply this is the number of passes per match that result in a shot at goal.
This is a more modest trend but nevertheless the number of chances created per 90m has shown increase over Lennon’s tenure.
This season:
Chance creation is pretty flat to slight improvement as a trend but as we can see was falling off and would have been a worse picture with the last three matches included.
Celtic’s PPDA
Celtic’s ability to beat a press and counter defensive organisation can be assessed using Passes Per Defensive Action. It’s a complex one to untangle as playing many passive passes in front of a team will result in a high PPDA whilst trying to force passes through a low block resulting in many interceptions and turnovers would result in a lower PPDA. So, I don’t immediately interpret this as good/bad based on high/low.
In essence Celtic are completing less passes per opposition defensive action This either suggests Celtic are less successful at countering the opposition press OR it could suggest Celtic are taking more risks to counter e.g. low blocks. I am minded to believe, with Chance Creation holding up, this is a stylistic measure that reflects Lennon’s more direct style preference.
This season:
My reading of this is that Celtic have increasingly been faced with low blocks this season and the increased number of passes reflects more “passing around at the back”. I am not suggesting it is either a “good” or “bad” thing and would refer back to Chance Creation as the trend to follow. I include as it is interesting stylistically!
Defending
How have Celtic been limiting the opposition over the same periods being analysed? We can look at the same metrics for the opposition.
Opposition xNPG
Again, we exclude penalties. By this basis a league table formed of xNPG for and against see’s The Rangers 5th (that’s a joke – lighten up people!).
Over Lennon’s tenure, the opposition xNPG is decreasing. Anyone surprised?
Let’s look at this season:
You might also be surprised to see it has been reducing prior to the last week. Well, I was surprised and it’s my data so feel free.
Opposition Shots On Target
As it says – shots on target from the opponents trended over 6 matches:
This is all within a fairly narrow average of between 1 and 5 shots on target per 90m. It does highlight that Celtic’s main opposition are of low quality with sprinklings of top-quality opposition in Europe. Bit nevertheless, and not surprisingly, in line with xNPG, shots on target against is trending down.
This season:
It may also surprise that shots on target from the opposition was, prior to the last week, trending down this season as well, quite significantly albeit it is a small sample.
Opposition Touches in the Box
Have the opposition been increasingly able to get possession is Celtic’s box?
Collections of European matches can easily be discerned from the peaks here. But once again it is s pleasing picture of modest but decline in opposition box possession over time.
This season:
Before this last horrid week, the trend on this had been downwards also.
Are Celtic easier to play against?
Opposition Impect
As mentioned, this indicates the ease with which the opposition can pass or run (mainly pass) through Celtic’s defensive lines:
So, this would be my first long term trend yellow flag. Celtic are becoming slightly easier to play through as time goes on. Definitely one to nip in the tactical bud.
This season:
Even before the horror week, the upward trend in Celtic being easier to play through continues.
Opposition xG / Shot
Are Celtic allowing the opposition better quality shots over time, or not!?
Another indicator Celtic are becoming easier to play against. Over time, the quality of the shots Celtic allow is increasing. That is, the probability of each opposition shot going in is going up#. Another yellow flag given the modest increase.
This season:
The trend this season however is to reverse that trend. Even in the last week, The Rangers, AC Milan nor Aberdeen had that many shots against so I would not expect this trend to be bucked considerably.
Stay with me, we are nearly at the end of the chart-fest. Neil deserves it!
Opposition PPDA
This indicates the efficacy of the Celtic press – are Celtic limiting the number of passes the opposition manage before a defensive action ensures?
Celtic are gradually allowing more opposition passes per defensive action over time. The press is less effective. The changes are modest – we are in the between 3 to 4 passes range. But still.
This season:
In line with the 3-year trend the opposition are completing more passes before any defensive action and gain modestly so.
Conclusion
No apologies for the length. What I have presented is a Neil Lennon team trended health check of some vital team level metrics.
Before I conclude, I want to be transparent about a couple of things.
- Before researching this my perception bias was that the numbers over Lennon’s tenure would show decline and the current form would simply be an acceleration of that.
- Before pulling together this data I have, privately, and probably sign posted on Twitter, that it wasn’t looking good for Lennon and perhaps a change would be for the best.
- As with all managers, there are aspects of Lennon’s tenure that annoy the fan in me:
- Lack of the Brown succession planning I flagged up 2 years ago
- A certain surliness around notions of modern methods and data in football
- A predilection for what I call players who “run around a lot” – i.e. visible effort over quality of decision making e.g. Elyounoussi, Duffy.
- This season there are some real (non data driven) concerns for me:
- Body language is awful from the manager – he looks tired and lacking fight
- There appears to be little or no interaction between the coaching staff on match days
- The team appears to lack an identify on how Celtic are playing in terms of attacking and defensive shapes
- The major domestic opposition are consistent in style and execution under Beale and have improved further
What I hope I have shown you is that Celtic, under Lennon, have largely improved over the last 3 years. Lennon sceptics will hate that. Rodgers acolytes will scoff. I am just reporting the data that surprised me too.
What is also true is that there are some chinks of regression especially around how Celtic are easier to play against, the press is less effective, and are giving up slightly better chances. This definitely needs addressing but if Celtic’s play was falling off the performance cliff I’d call it. It isn’t.
That’s the long-term trend. In the main it reflects well on Lennon.
But what of this season as football moves at such a pace?
There is evidence of regression on a larger scale albeit it isn’t uniform and a couple of sore results on the bounce tend to focus attention.
Are there mitigating factors?
Yes, there are.
- The annual centre back “who’s available” tediously continued this season
- Once again, more strikers have been injured than fit
- Once again, a new ‘keeper needs bedding in
- The impact of injuries has been disproportionate compared to rivals
- The impact of COVID19 has been hugely disproportionate compared to rivals
Lennon has ben backed to a far greater extent that I thought especially as no major player was sold. He got his major signing targets. He can have no complaints.
I believe taking a data driven approach what I have shown is that the long-term trajectory for Celtic is upwards, Lennon has been backed, is improving over the long haul and deserves the chance to rectify this short-term road block.
But there needs to be changes especially in defensive organisation and setting a template and framework for how the team is set up fundamentally. Only Lennon knows if he has the hunger and drive to continue. Is he getting the support from the other coaching staff? I couldn’t know. But a change in attitude and approach is needed.
Does that mean sacking Lennon? I don’t see the evidence for that for now.
David Malloy says
I have not enjoyed any game this season it has been awful about 4 games we have been lucky to win the team looks disjointed and plays players out of position and then they get slated for not producing ??
celticbynumbers@btinternet.com says
David I don’t disagree but my point is Lennon deserves the chance to fix. or do you disagree?
Duncan says
I think you can only judge him thoroughly this Season once he has the key men back and 100% fit.
The impact of Christie Griffiths and Rogic was plain to see and no shock to most of us.
On the long term planning for Brown I think Turnbull was that last Season but his medical put that back 12 months.
Captaincy wise then McGregor is the obvious man to take over long term despite what many think I think he sees his future at Celtic long term and the reason I say this is simple.
The Club stood by him after he made the personal mistake of drink driving.
He’s never forgotten that I don’t think .
W dare more open to counter attack as this is the only flaw with 3-5-2 it does leave teams exposed out wide if the wingbacks don’t do their defensive duties properly.
Ideally you would have both Right and Left CB’s experienced at the fullback role like Elhamed is and Lustig-Mulgrew were previously under Lenny.
Getting Elhamed Bitton Jullien Édouard Forrest amd Ajeti 100% fit will make a massive difference to the current squad and as Leigh Griffiths pointed out today we are only 2 points short of the total reached last Season at this juncture despite the injury issues the Manager and team have had to cope with.
Our nearest rivals haven’t had any serious injury issues yet and haven’t also been affected by Covid restrictions.
Far too early to be calling for the Manager to be sacked it is a disgrace so many are on Social Media.
Terrific effort on these stats btw but as someone who looks beyond the performance or result they really don’t surprise me.
Ryan Malone says
Would be interesting to compare our own stats vs Rangers. Are we unlucky? Are they over performing or do they have higher than expected goals?
How do the above stats compare to actual goals scored and conceded? Does the above mean we’re conceding more than we should be – a sign of poor goalkeeping performances?
celticbynumbers@btinternet.com says
Not really over last 2 seasons in fact League Cup win was jammy as hell but i don’t feel guilty cos Goram and Tommy Burns. But seeming lack of ability to fathom Beale’s The Rangers is exasperating in the extreme.
Duncan says
They are performing well this Season.
There has been a major improvement under Gerrard especially in Europe where the real test is.
If you look at their start to this Season it has matched anything we have done in the last 9.
Whether they can continue to play like they have all Season remains to be seen.
Paul says
Have to say, really refreshing to hear a more positive spin on things Alan. Like you (and most people) I would have said the trends would all have been negative but I’m pleasantly surprised to see that’s not the case.
Thanks a lot for this, really appreciate it. Definitely has made me reconsider how strongly I believed it was time for the manager to go.
celticbynumbers@btinternet.com says
Paul – i started from a “he must be sacked” perspective but as trite as it sounds i do let the data lead me
Tommy B says
Great analysis as always. I was definitely surprised. Always thought we had regressed tactically but advanced in forward movement.
Would be interesting to look at the identical stats excluding Europe.
Thanks for your efforts in this. Insight always helpful when emotions are fraught.
John says
If I was Celtic management I would hire you ,put you up in the gantry to give the manager an extra set of eyes and full match analysis report no pulled punches
SFTB says
A very welcome counterblast to the “My gut tells me that we’re playing much worse and my gut never lies” belief that is prominent amongst the support kust now.
Nonetheless, the data does not include the last 3 games- 2 defeats and a draw- I expect that will make a mild impact on the overall trend. I “look forward” to an article on the Sevco match stats, if you are up for it, to lay clear where the danger areas were for us and why we made such little impact going forward.
As always, your analyses is worth 20 of my gut feeling posts BUT…..
I would counsel you on getting too sure on the non-data areas e.g.
Body language is awful from the manager – he looks tired and lacking fight
There appears to be little or no interaction between the coaching staff on match days
If we have no data on how his body language was previously, then this is lurching into “gut-feeling” territory.
And, unless, you are present on match days, you have no reliable indicator of how often he consults with his coaches. We are at the mercy of TV editing of pictures and whether they routinely show clips of dug out or sideline consulting. This measure is also affected by setting a good example on Covid required social distancing. It is also possible to communicate electronically through earpieces- do we and can we use such technology?
celticbynumbers@btinternet.com says
All fair. For the last 7 months like many I’ve been working from home and so reading body language and facial cues has become very important in monitoring the wellbeing of my teams remotely. But I am an amateur at that as well as the numbers ! ??
Kildave says
Thanks very much for the data. It adds to the debate. Covid has impacted on teams already this season St Mirren probably being the most impacted. However given the recent reinfection news I would presume that it is likely that all teams will be faced with difficult periods this season and it will be interesting to see how that impacts their team performance. While I am not confident things will turn round in the short term I think Lennon deserves time to turn things around even if that can be frustrating. On another point; there were questions about the attitude of players who want to leave raised by Lennon. I think it is worth saying that both Ajer and Christie’s have not lacked effort or commitment in the last couple of games. Which might suggest he Has not lost the dressing room.
Rolling_Stone says
Great article with some very interesting data, showing that things have been improving under Lenny.
However, I still think there is justification for being dissatisfied with matters and even for the calls to replace him.
Firstly, on the improvements. Neil inherited a double treble winning team and then spent circa £35m. Has any other manager received such backing and in such a short period of time? Are the improvement then not somewhat expected? I don’t think that’s an unfair comment to make.
“deserves the chance to rectify this short-term road block…But there needs to be changes especially in defensive organisation and setting a template and framework for how the team is set up fundamentally.”
On the first half of the above- these aren’t short term road blocks. Defeats at home to Cluj, the Hungarians, Copenhagen and Sevco x2 span both Neil’s season in charge. These are very poor results, with recurring themes and failure abound.
On the second half of the above- you speak of fundamental changes. I dont suspect these to arrive as we are a good way into Neil’s second season and these qualities have lacked from the off. Neil is a coach of 10 years. These should not be new skills to be learned and if he hasn’t implemented changes to date, then I don’t think he will.
More generally, my rationale for sacking Neil is based on performances in Europe and against Sevco. Much like Bottas in Formula 1, people are not going to pay attention to you by beating 80% of the field when you have vastly superior assets (car/players) and resources, rather it will be your performances against opponents on a similar or greater level that will determine your quality.
I think in that regard Neil is failing badly. I’d point to the results referenced above. I’d also reference Jamea Dailey’s article on here regarding the importance of Fraser Forster.
From memory, under BR’s reign, our expected goals conceded in Europe matched our actual goals conceded.
However, under Neil we were conceding a considerable amount fewer than expected. The conclusion being that FF was saving us, I think papering over cracks would be an appropriate description; look at the LC final, the two games v Lazio and even the Rennes home again. Time and again saved by FF.
The worrying thing is that with him gone, we are reliant on tactics, organisation, discipline and our general game plan from the manager. The result: we are losing these games.
We are punching under our weight with Lenny and we need a change.
Wait another month and the league may be gone.
Jonto says
Excellent analysis and a good corrective to the Lennon chasers. I too have been frustrated at our performances, without crossing over a big expertise boundary i worry about NL’s mental health. One analytical point when you talk of regression in this context are you meaning in the statistical sense regression to the mean or a downward indicator in performance? Also i find any measure of goals which omits penalties could distort the picture. Better teams with more attacking episodes get more penalties and given rule changes and the like that’s more of a systemic shift in the game. Maybe penalties could be weighted down to control for the higher likelihood compared to poorer opposition..
Great analysis and a refreshing change from the grousing we tend to see online. For me Lennon has a special investment for 10 and can lead us To it. Thats as long as he is looked after in terms of his own health and wellbeing.
The Cha says
I assume you started from Rodgers year 2, as the Invincibles were so unique that the rest would be so skewed to render objective comparison meaningless?
Lets hope this season doesn’t turn out to the mirror image to that one and tbf, as you’ve shown, that isn’t happening…yet.
My big concern is addressing the obvious current deficiencies so that they don’t become long term trends.
This means everyone, including (especially?) the manager being focussed on addressing and sorting out the issues together.
Sadly, I don’t get a warm feeling in that regard, even ignoring throwing the players under the bus after Ferencvaros, as it was “so long ago”.
Recently he’s complained about leaked team sheets, a ridiculous spat with a minor Motherwell coach, complaining about refereeing decisions.
None of this appears to acknowledge our current issues, far less dealing with them and seems to simply be deflecting from them.
I hope I’m wrong and the proof will be in upcoming performances.
Uralius says
Alright Cha,
Where is the mid-field it’s like we are a 2000s American team! The gap between the lines is far too great and any decent team will see that and exploit it!
Cornelius says
Good article & thanks for the work to put to together.
Being cautious about allowing ‘gut feelings’ to convince us is valid enough but that has to work both ways. If gut feelings are to be critically questioned when the team is faltering then they should just as critically be questioned when the team has apparently “turned a corner”.
This explains the exchange of opinions on fans’ sites after games. Everyone’s opinion or gut feeling about how well we played, who was MOTM, who should have been subbed & even did we deserve to win the game at all provides plenty of scope for disagreement & even anger.
Retrospective analyses like this help us understand our own skewed & partisan view during games or runs of games. I especially liked your admission of personal biases beforehand. This analysis gives us all a chance to take stock. Also your caveats about stats that can disguise a clearer picture. I think of all the post game stats put up by Uefa where we find that CBs have the most passes completed % AND the highest number of passes: all that unchallenged passing over short distances to similarly unmarked colleagues.
I’d like to see more use of the large 1st team squad we have but NOT just in emergency situations. Bedding in newer players means that when we really need them they stand a better chance of succeeding.
I’d like to see less reliance on “Brown must start every game”. He is a legend but as such deserves to be used so that we get the best out of him. I don’t think we do him or the team any favours by making his the first name on every team sheet.
Similarly with Elyounoussi. It seems that he has an agreement to be played in more games than his form should dictate. He admitted after the Lille game that European games are his preference & frankly that shows in lower profile games against some SPFL teams. Including him automatically then has a knock on effect on the rest of the midfield & forward line that isn’t always positive.
Tactically, I question how prepared the team is before stepping out on to the pitch.
Against Sevco that seems to speak for itself.
But against Aberdeen on Sunday we looked surprised at how Aberdeen played for the opening 10 minutes. EG Frimpong played like he’d been told to close off the inside to the Aberdeen winger. When he did the winger simply strolled to the byline & crossed the ball unopposed with his left foot. He did it twice & each led to dangerous crosses that Celtic only managed to scramble away.
Was this a series of brain farts by the player or him following a coaching instruction?
My gut reaction was that it was a coaching inspired choice by him.
Uralius says
Alan,
This is a great piece! Thanks so much for doing this. What stands out to me is we could’ve reached a breaking point. Where Celtic’s less effective pressing and our oppositions greater ability to play through us has reached a certain level that suddenly we will see multiple statistics like good chances and touches in the box against us sky-rocket! The trends that we are all seeing point towards a requirement for our team to begin playing more compact, flooding mid-field out of possession with a narrow compact shape just in-front of the CBs, with our wide-players allowing the cut-in from the wide areas to just run into bodies. Unfortunately without certain key personnel we are not going to be extremely effective on the break. With the sideways pass allowing the opposition to get back behind the ball.
We should be real here, Europe and “rangers” aside our second string of players should be more than adequate to defeat teams domestically. If they can’t perhaps we should just fill our team with Scottish players.
I’m wondering if Lennon has tried to make us harder to beat, but has taken a risk in the middle of the park by not bringing in natural Defensive midfielders like Wanyama or Ayer.
Uralius says
Isn’t/wasn’t Bitton a CDM? Perhaps it’s time to try out something different if we are to insist on this formation with only 1 proper holding midfielder. Although I’d rather see Ayer moved there. If we must play our “marquis” signing.
Ian McKelvie says
Sounds like we need to tighten up the gaps between midfield and defence. (Which is hard when half the defence plays in wide midfield.)
Here is the second goal from the Sparta away game, pretty much explained by a tetris type model courtesy of the Notts County Girls coach. Around 5.20
https://www.youtube.com/watch?v=ynyeGpsksnY
Julien took the flak,and McGregor didn’t stop the cross, but surely Laxalt should be doing some defending ;0)
celticbynumbers@btinternet.com says
100% agree – the spacing / alignment and control over who is where and when is all over the place