Nancy Bhoy Bringeth

And give him a +1 invite

I was only slightly aware of Wilfried Nancy and his tenure at Columbus Crew prior to the recent speculation about his being part of Celtic’s managerial search. MLS is not a league I track closely, and after they placed their games behind a paywall on Apple TV, I have barely seen any games at all.

However, while taking some cursory looks at the various rumored names, some aspects of his profile caused me to do a bit more digging. Fortunately, MLS is a spreadsheet shagger’s dream, as American Soccer Analysis provides excellent open source data for the league.

All data within this column was sourced from their free web application.

Let us begin with the boring Expected Trophies foundation of relative payrolls, where we see for the 2025 season Columbus (CLB) are 18th out of the 30 teams. Nancy arrived at Columbus for the 2023 season, with the team ranked 9th for 2022, 6th for 2023, and 18th for 2024.

MLS is a closed and salary cap-structured league, similar to the J and A leagues in Japan and Australia. Similarly, the dispersion in payrolls is MUCH lower than those seen across most European leagues, and certainly in Scotland. So, despite Columbus sliding from 9th to 18th over the three seasons, the difference between those slots for 2025 is about $3.7 million- meaningful but not huge. The difference between those two slots was about $2.9 million for 2022 out of 28 teams.

For the purposes of this analysis, the importance of this context is that the team’s performance level was most likely not a product of some material increase in spending.

American Soccer Analysis produces a possession-based metric called goals+ (g+), which is comparable to what Alan and I frequently refer to from Statsbomb, On-ball Value (OBV). The models have important differences but are generally trying to do similarly, which is to value all actions within a game in the “currency” of expected goals.

The 2nd image above graphs Columbus’ xG Difference and g+ Difference for each of the past 4 regular seasons, so the one immediately preceding Nancy’s arrival and the three for which he has been head coach. Despite what was arguably a lower amount of relative payroll spending over the period, underlying performance levels increased dramatically and persisted.

There was a backslide in xG Difference a bit for the 2025 season, but g+ Difference was marginally higher than 2024, so I would have to do some additional digging to try and attribute what may have occurred to drive that disparity.

However, relative payroll spending fell, so some potential impact would not be a surprise. In addition, the club sold the third highest g+ player in MLS over the 2024 season for a record €13 million to Real Betis.

Analytical Conclusions

American Soccer Analysis produced two articles this season where analytics at MLS clubs was covered in some depth. They sent a survey to the 25 clubs that list at least one analytics staff person, and Columbus was voted as the most advanced by 10, which was 2nd highest to Colorado’s 11 votes.

Their annual State of MLS Analytics article from July listed clubs based upon the number of staff members populating their analytics departments. Columbus’ head of analytics, Alex Mysiw, has a masters degree in bioinformatics. I had to look up exactly what that entails, and here is the first part of the description on Wikipedia:

Bioinformatics is an interdisciplinary field of science that develops methods and software tools for understanding biological data, especially when the data sets are large and complex. Bioinformatics uses biology, chemistry, physics, computer science, data science, computer programming, information engineering, mathematics and statistics to analyze and interpret biological data.

Similar to my professional experience with financial markets, and with no disrepect to those working in football analytics, but Mysiw’s academic field was a lot more complex than dealing with football data and analytics.

This introduces an interesting attribution question, which is to what extent the surge in performance levels at Columbus has been due to Nancy’s coaching versus aspects driven by the club operating smarter via higher quality analytics?

Nancy managed at Montreal in the two seasons prior to his arrival at Columbus. For the 2021 and 2022 seasons, the team ranked 13th and 23rd in payroll, and finished 9th and 4th in xPts. In the following three season, Montreal slashed payroll to be towards the bottom of the league, and xPts followed, with ranks of 28th/23rd, 29th/25th, and 26th/26th in payroll/xPts rank.

This backdrop suggests to me that Nancy is likely a very good coach/manager and his style of play aligns with how Celtic have played in recent seasons, though he’s deployed some variant of 3-4-3 at both clubs.

Hiring Nancy would likely be a good decision in isolation, as I actually believe our squad, even in its current state, is reasonably well suited for something like 3-4-2-1.

However, it would not comprehensively address the systemic issues relative to recruitment and the broader cultural embrace of leveraging modern analytics tools to better run football operations…but it would be a start. None of Rodgers, Lennon, or Postecoglou appeared particularly willing and/or adept at leveraging those tools.

Better yet, Celtic should look into lifting out at least the head of analytics as well as hiring Nancy.

What is it they say about it being darkest before dawn?

Perhaps another false dawn on potential modernization, but we can dream for now until a new manager is hired.