The Celtic and Rangers back to the future battleground and the "expected trophies" model - James Dailey

Welcome to the football world of “Moneyball” and analytics. My weekly column will be focused upon using the skills acquired over 25+ years as an analyst to analyze and comment on all things Celtic.  

First things first: generally speaking, the analytics community is terrible at persuasion and education.

I struggle with it myself as well, so please try to be charitable and patient with me.  

“Expected” whatever is probably not a good way to label these metrics for anyone who is not a hardcore stats person. A simple way to conceptualize all these metrics is to think in terms of probabilities. If a shot is assigned an Expected Goals (xG) value of 0.25, then an easy translation is that out of tens of thousands of similar shots, as a group they went in about 25% of the time.

It is not all that mysterious or complicated in concept. There are different versions and models of all these metrics, with their own assumptions and data sources, but they all share that same simple idea. Think probabilities. 

In addition, statistics and data are not deterministic and all encompassing. If consumed and utilized within an effective and robust analytical process, they can enhance understanding and help answer questions.

There are critics who correctly state stats and data can be used to confirm biases and support desired narratives. My job as an analyst is to try and remain as objective as I can, offer my views in good faith, and do my best to get as much “right” as I am able.  

Which leads me to the topic for my column today: “Expected Trophies.”

Many football fans, coaches, executives and players understandably talk of winning trophies. Celtic have justifiably celebrated “unprecedented domestic success” over recent seasons. But my analytical mind has asked the question, “How many trophies should a club win?”   

Various studies have suggested that a club’s wage bill relative to its domestic peer group is responsible for between 80%-90% of results. This idea was popularized as part of the 2009 book, Soccernomics. I have run my own regression analyses across various leagues over various seasons, and have consistently seen this theory hold up.

As a result, I have come up with a simple football attribution framework, which goes something like this: 

Wage Bill(80%) + Recruitment/Player Development(15%) + Game Decisions(5%)= Team Performance 

These descriptions and percentages are not to be taken literally or intended to be all encompassing - rather a conceptual framework to be used when thinking about the components, or factors, which drive a football club's competitive output on the pitch. Of course, “Team Performance” does not equate to results or points won! As we have all seen and experienced, the variance in football is significant in single games, as teams which create ton of scoring chances can lose to teams that score on a single or just a couple of chances. Teams regularly drop points despite performing significantly better than the opposition. Here is a simple formula to conceptualize: 

Difference in Team Performances + Variance = Results 

Why am I going through all of this? It is a framework with which I will be analyzing the upcoming season for Celtic, but it also offers a way of assessing historical performance levels, results, and measuring them relative to what can, or could have been, “expected” – think probabilities! 

On a long-range basis, the 16-season period from Tommy Burns’ last in charge as manager, the 1996-1997 season, up through Rangers’ last prior to their collapse in 2011-2012, offers a glimpse into a likely future.

Why that period? The 1996-1997 season was the first in which Celtic’s average home attendance was able to approach that of Rangers, and where financial resources became ample enough to begin to empower a legitimate challenge. This emerging threat exploded onto the scene the following season, with Celtic obviously preventing Rangers’ 10th consecutive league title.  

Over that 16-season period, the two clubs split league titles evenly.

While it is difficult to have specific and accurate data on Rangers’ spending levels during the years EBTs were used, I estimate wage bills were close enough during almost of those seasons where relative factors such as recruitment, coaching, and good old fashion luck, resulted in a competitive and compelling 2 team league race most seasons.  

In fact, over the 16 seasons, total points earned were 1,471 for Celtic vs 1,424 for Rangers. I estimate that the aggregate in wage bills over the period was approximately £481 million for Celtic vs £476 million for Rangers. 

Relative to domestic cup competitions, Celtic won 10 and Rangers 15 of the 32 available over the period. Given the large advantage the Glasgow clubs enjoy against the rest of domestic competition, I think a reasonable haul for cups is about 40% each and 20% for the remainder of the clubs. 

If we review Martin O’Neill’s tenure at Celtic as a mini-case study, his teams won 3 of 5 leagues and 4 of 10 available cups. Over those 5 seasons, wage bills were roughly at parity in two, Celtic about a 10% advantage in one season, and Rangers with about 20% advantage in two.

Given the impact of inheriting Henrik Larsson, who was on his way to becoming one of the best strikers in the world, I think a reasonable argument can be made that O’Neill’s teams won the number of trophies that they should have- they basically matched their “Expected Trophies.”  

Brendan Rodgers' spell was an unprecedented trophy success

On the flip side, Brendan Rodgers’ Celtic side, and then the first season in Neil Lennon’s second tenure, exceeded Expected Trophies by winning every domestic cup for 4 straight seasons!

However, what was the cup record over the 26 seasons starting in 1996-1997? Winning 20 out of 52 cups is right near the 40% level – the great success in recent years has basically enabled Celtic to catch up with what should probably be expected. 

With that walk down memory lane completed, what about the future? The wage bill structure of the Scottish Premiership is such that Celtic and, starting two seasons ago, Rangers enjoy such a large gap over the likes of Aberdeen and Hibs, that the league is most likely a two horse race at least in the near term. We cannot know prospectively what wage bills will be for each team heading into the season, as that information is only available retrospectively through company filings. However, I maintain pro forma projections in order to provide estimates.  

The question of potential European qualification and associated revenue remains uncertain, so it is too early to have a firm grasp on the wage bill landscape. For example, in a nightmare scenario in which Celtic do not qualify for even the Europa League and Rangers qualify for the Champions League, the likely impact on each club’s spending level on players is likely to be material.  

With Celtic’s board having gambled £30+ million in cash reserves to back Neil Lennon and try to secure 10 In A Row, Celtic and Rangers are “Back to the Future,” with financial resources being close enough to make other factors even more important. In many ways, the competitive landscape has returned to being comparable to that 16-year period.  

Much of the fun in being a supporter is in debating, discussing, and arguing about the other 20% - transfer rumors, player trading, academy players emerging with the first team, and endlessly second-guessing lineup and tactical decisions of the manager. But the dreams and aspirations to win trophies is a foundational goal for a club like Celtic, and now you know the general process with which I will be analyzing the team over the coming weeks and months. I am looking forward to the journey.