There remains quite a bit of debate on which outcome in the Hearts v Rangers best suited Celtic’s title chances - a fertile topic for this week’s podcasts.

Here, we continue on from last week’s journey through the moments, randomness, chaos, and inches which were part of how winners and losers were determined.

This exercise is not intended to be forensic or suggest any specific team was hard done - simply to point out how much such close games can come down to fine margins.

Read more

Of course, as we always warn, single game data and related metrics are to be taken with a grain of salt, as there are material limits to what event data measure, and what the associated modeling can help infer.

To highlight these limits, I created four videos to review all the shots for each of Celtic and Falkirk, then Rangers and Motherwell.

For each shot, the expected goal and post-shot expected goal values are shared, with my eyeballing of where keepers’ positions are relative to modeling an optimal angular position at the time of each shot.

Read more

Read more

As we look forward to the Cup semi-final, figured since I prepared would share and supplement the analysis provided by Alan and Martin:

While the second half in the league fixture Sunday was even more reflective of the unusual inversion of playing style dynamics, I clipped this continuous clip from relatively early in the first half.

Shared these two graphics on the early-week podcast and sharing again to connect the game-data with the video above.

Celtic generated two shots from pressing St. Mirren high which compares to a season-long average of just under 5. Neither team were able to generate a shot from counter attack or clear sighted on either keeper.

Read more

What had been a material positive with Callum McGregor’s performance profile has deteriorated this season. His ability to control tempo in buildup and drive ball progression from the defensive third through the middle third and into the attacking third was important and valuable.

I can only speculate as to why he has become so risk averse this season, but with a temporary departure during the 33 days of Wilfried Nancy’s tenure, it has changed materially.

Read more

We can see from this graphic that the non-penalty goal scoring rate (all metrics are per 90 minutes) is down from 0.589 average in the prior 8 seasons to just 0.251 this season. Notably, xG and post-shot xG (PSxG) are down less, with xG down the equivalent of one relatively poor quality (0.057 xG) chance per 90 minutes.

Given non-penalty xG for the team is down almost 17% versus last season, that 11% decline suggests that “getting onto the end of chances” has not been the major issue, overall. While not shown, xG per shot is actually up 4% versus the long-term average.

But before I get into more specifics, I created this video to walk through Tomas Cvancara’s four shots in Sunday’s victory over Dundee as a way to contextualize the metrics in the first graphic:

Read more

Some contextual factors to consider for this analysis:

1. I use some comparative metrics between the 2012-2013 to 2015-2016 and then 2016-2017 to 2025-2026 periods. The logic behind those periods are to examine what Celtic were doing in the transfer market for the period after Rangers’ liquidation but before the analytics revolution kicked off, and then the season I use as the starting point when data and analytics began to become materially impact player trading across the industry in 2016-2017. That is my discretionary call, which is why this entire analysis is characterized as an “exercise.”

2. The football industry experienced severe inflation over this period. I attempt to capture that factor to an extent, but there is not singular calculation that can quantify “inflation” in a uniform way.

3. I use Transfermkt as my source for data. It is not 100% reliable. As we often say, data is often imperfect, all models are flawed, and our goal is accuracy over precision. The euro is used as the base currency.

4. The ratios I generate are generated as part of my analytical process and are not intended to reflect or align with player trading data reported within official club financial reports. This is not an exercise in financial reporting using IFRS or UEFA’s regulatory regime via FSR, etc.

On to the first graphic, which given the nature of the past 6 months and the contentiousness between the Celtic board and many supporters, is what I call the Board to Supporters Ratio:

The formula I used to generate the ratio is shown at the bottom of the image, with stadium capacity data taken from Transfermkt under each club’s home page on the site.

Read more

Alan and I have been very critical of how Celtic have managed injury risks in recent years, while noting the positive outlier during the 2024-2025 campaign, when players seemed to be rotated in what appeared to be a relatively coherent fashion.

Stephen McGowan’s January 9th column published in the Herald provided a plethora of related reporting as to the extent to which the legacy culture manifested. It is a piece of reporting I am likely to return to as the autopsy of the current season is done once over.

Accustomed to elite coaching standards under Brendan Rodgers, then Martin O’Neill, first impressions were underwhelming. Rumours of players and backroom staff being unimpressed with the new manager surfaced in no time at all.

The description of Martin O’Neill as offering “elite coaching standards” was notable early in the piece. John Kennedy had departed along with Brendan Rodgers, and none of the people remaining at the club or brought in as interim coaches had CV’s which suggested “elite coaching standards.”

Despite 18 years as Celtic’s goalkeeping coach, Stevie Woods was challenged to prove himself all over again to the new regime. Rumours of a departure became a weekly soap opera with more twists and turns than River City.

Read more

I was skeptical that the club would have the depth and staying power necessary to successfully deploy what appeared to be Jens Berthel Askou’s vision. In the interest of brevity, I will call it a Bodo/Glimt-adjacent style of play and game model.

Not only has staying power been a non-issue to date, Motherwell have been getting better.

Of the top five clubs above, only Motherwell has played in 9 vs 10 league games so far in calendar 2026. Depth and rotation in their forward line and midfield, as well as no European or late Cup runs, have probably kept the team relatively fresh.

As Alan’s most recent title race update showed, their rolling average has been trending upward.

Read more

As pressure continues to mount in what may be a title race for the history books, how individuals and collectives respond under increasing pressure is sure to be interesting.

This past weekend offered some intriguing examples. Whether it was Derek McInnes going to 4 CB’s with a 1-goal lead at home to Aberdeen, Celtic going unusually direct for large stretches of the derby (and at Aberdeen), or Rangers being unusually passive in a “must win” home game, what might happen when the pressure REALLY ramps up?

Looking specifically at the Celtic vs Rangers game, the latter setup in possession as we speculated in last week’s Extra Time preview, with a narrow 4-2-2-2. Selection seemed coherent within that context. They played direct and looked to negate risks from Celtic pressing while they had possession.

Out of possession, Rangers seemed to be trying to target Dane Murray like they had Auston Trusty in the League Cup semifinal, but otherwise were very passive and man marking was inconsistent.

In contrast, Celtic went unusually direct, often under little or no pressure, during the first half.

Our central midfielders were generally able to make passes when under light pressure or none at all.

Read more

1. The timing and circumstances of Wilfried Nancy’s appointment were poor.

2. It was more important for Celtic to hire Columbus Crew’s head of analytics than Nancy and they should do both, ideally.

3. Nancy’s intermediate to long-term outlook at Celtic offered promise ala Ange Postecoglou’s tenure, but that the short term was perilous due to #1.

4. Sticking with Nancy at least through February was the least bad option given the risks likely to be unleashed by sacking him and Paul Tisdale given the importance of having some semblance of coherence in the January transfer window.

5. While open to the possibility that Tisdale was not up to the job, I could discern no clear evidence from the prior transfer windows given the lack of visibility on who was responsible for what, and would reserve judgement until analyzing the January 2026 window. The only evidence available was the signing of Araujo, the targeting of now-English Premiere League player Jocelin Ta Bi, and links to Niko Sigur.

Obviously, the club decided to go down a very different path, so today I review what credible evidence there was/is to support those decisions.

I will not review the Martin O’Neill timeline, as his CV is well known.

However, Shaun Maloney’s seems less so outside of relatively surface-level information on where he has worked.

Read more

This graphic is a scatter plot of non-penalty shots conceded and shots from open play conceded in all league games since the beginning of the 2018-2019 season. As is probably unsurprising to most, the greatest volume of shots conceded have generally been in derby matches at Ibrox, including the outlier from January 2025 in the truly awful 3-0 defeat.

For the non-derby games I have highlighted above, here is some additional information on each in chronological order:

As has often been the case over recent seasons, Celtic often get results versus overmatched domestic opposition even in games where performances are not very good.

Read more

The specifics from that clip are obviously different, but the underlying question is commonly apt, “What’s the problem?”

Since the departure of Kyogo last January and the diabolical summer transfer window, many have understandably considered the striker position the biggest problem at Celtic…kinda like replacing Giambi was for the scouts in that clip?

The striker position has been A problem, but has it been “THE problem?”

Subscribe now

Through last week’s game versus Falkirk, Celtic have an average non-penalty xG of 1.71 through 24 games. In the nine league games under both Martin O’Neill tenures, that falls to 1.64.

The image above lays out various performance measurables related to goal scoring by Celtic strikers in recent seasons.

Read more

Subscribe now

Research suggests that through the lens of interpersonal survival, these Thin-Slices are pretty accurate in determining whether a new encounter is a threat or friend.

But how about for the domain of footballers’ performance measurement?

Cognitive biases are a feature and not a bug of evolutionary psychology. Performance measurement and related analysis is an attempt to account for such things.

Read more

The first edit was to change the subheading from “Avoiding Anderlecht-ification” to unfortunately laying out why I believe that scenario is, by some distance, more likely than not.

What does that made up word mean? I referenced it a bit in Part I, but it is my way of trying to use an analogy to introduce the idea that Celtic may be in the process of transitioning from an extended period of domestic dominance to be a club who routinely finish 3rd or worse.

Such a scenario likely seemed implausible to many as recently as 11 months ago, as Celtic took Bayern Munich relatively late into the 2nd leg of the first knockout round in last season’s Champions League. However, that was around when Anderlecht-ification had already become my base case.

Read more

Entering the 2025-2026 season, my concern was that Celtic faced a likely inflection point - kind of like an important decision within a choose-your-own-adventure book, which was as style of children’s literature popular in the 1980s when I was a kid.

A quick summary of what I foresaw:

1. Celtic faced with material known and controllable risks dealing with a manager in the last year of his contract, an antiquated football operating model, a thin squad with a poor underlying level of quality, and a controlling shareholder seemingly resistant to modernising.

2. Hearts, Hibernian, and Rangers in the nascent stage of modernising using different operating models pushed through new ownership participants.

3. A club and country club coefficient backdrop poised to shift in an extremely negative way against Scottish clubs and Celtic in particular.

4. A return to the perils of a Champions League qualifier.

I could delve into more and deeper into each, but have written and spoken about these at length since February 2025.

I have used circumstances at Anderlecht within Belgium over the last decade as an analogy for what has unfortunately been my base case for Celtic over the next 3, 5, and possibly 10 years. As I often say, analogies can be useful when introducing something that is unfamiliar.

For example, if I was to meet someone who had never seen a zebra before and was not familiar with the animal, I could describe it as being “like a horse.” They are obviously two distinct animals with very important differences, but the comparison gets the picture to within the directional neighborhood.

A primary variable within my base case outlook was whether the plethora of risks facing Celtic this season would trigger a kind of “avalanche” like what happened in 2020-2021, and then what the details of the avalanche would look like.

Read more

That analysis concluded with the following:

Unfortunately, Celtic have the greatest degree of uncertainty given my analytical framework, with skew titled towards the negative. Yes, some increased efficiency with the arrival of Nancy could occur, but the timeline and risks of getting to that point given the context of his arrival remains highly uncertain. It is probably unrealistic for him to wave a magic wand and acquire omnipotence amidst the shower of fixtures forthcoming.

Throw in the lack of depth, injury issues, and uncertainty on how well the January transfer window will be managed, and I could 100% see a scenario where underlying performance levels converge with Hearts, and possibly even Rangers.

This is where the art of analysis enters the picture, as people ranging from Niels Bohr to Yogi Berra once said:

“It’s tough to make predictions, especially about the future.”

Resiliency is a key concept I use when making predictions about the future, and I see far more of that at the title contenders than I do at Celtic pending proof that the combination of Nancy and Paul Tisdale can stabilize amidst a potential burgeoning avalanche.

Could we reasonably consider the past two weeks a kind of “avalanche” within the support and results?

Read more

Orderly linear systems are typically easier to analyze, where as complex systems in some degree of disorder/chaos can be exceptionally difficult.

Welcome to Glasgow Celtic FC as of December 15, 2025 - complexity and emergent chaos.

Subscribe now

The setup, or initial condition, for this season became evident to me starting in February 2025. While the specifics were different, I saw a similarly perilous setup in spring 2020 heading into the 10-in-a-row season.

Read more

Subscribe now

It truly has been an interesting league to cover and analyze so far this season, with the trials and travails at Celtic one of many storylines.

As I continue to explore various topics as to what has actually changed this season, some things jumped out for this week’s column.

The first of MANY graphics this week compared the number of possessions and fouls per game for each Premiership team starting with the 2018-2019 season. I added a simple linear regression line which suggested some statistical correlation between the two variables.

I would argue that makes common sense, as higher turnovers in possession logically present more situations where fouls are more likely to be committed. We saw the Dundee United side from 2020-2021 and Celtic from 2024-2025 were two relative outliers within this sample.

The second graphic added the 2025-2026 season-to-date to the same data from the first, where we saw how crazy the entire league has apparently gone - CHAOS BALL!!!

Read more

But is this likely to continue?

Will the major changes at Celtic and Rangers lead to improved performance levels? Will Derek McInnes continue being Derek McInnes? How resilient will Hibernian be despite the Black Knight divorce?

Subscribe now

To explore these topics in this column, I want to start by placing Celtic’s performance against St. Mirren into some context, as I did following the Midtjylland performance on Extra Time.

Through Saturday’s game, Celtic have had 270 league games since Rangers reached some semblance of financial relevance beginning with the 2018-2019 season. The image shows the xG for Celtic for each game on the vertical y-axis and the non-penalty xG difference on the horizontal x-axis. Needless to say, but these measures, Saturday in Paisley was no bueno.

I added some labels to the image showing the date of each game merged with the opponent. Probably unsurprisingly, the lower left corner of sadness is populated mostly by games versus greater-resourced opponents such as Rangers. Please forgive the Americanized formatting on dates.

Read more