The Champions League Draw: When the Rules Outgrow Manual Assignment

The Champions League Draw: When the Rules Outgrow Manual Assignment
At 18:00 CET this Thursday, Monaco hosted the league phase draw for the 2026/27 UEFA Champions League. What you see on screen is familiar: a ball is pulled, a club name appears. But what happens next is the part that matters more to a business owner than to a supporter. A button is pressed, and software instantly determines that club's eight opponents and splits them into four home matches and four away.
That is not a television gimmick. It is a large institution admitting that the assignment it runs has grown past what a person can do by hand. The same problem exists in your company, at a smaller scale and under a different name: who works which shift, who visits which customer, which order goes out with which driver.
What made the manual draw impossible
The league phase format puts 36 clubs into a single table, split across four pots of nine. Each club plays eight opponents: two from every pot, half at home and half away. On top of that come the restrictions. Clubs from the same national association do not meet at this stage, and no club faces more than two opponents from any other single association.
Read that last condition again. Every ball drawn does not settle one fixture, it changes what remains possible for everyone else. An early pairing can leave the final club with no valid opponent at all, forcing the whole draw to restart. Reporting on the format estimated that running the current draw fully by hand would need roughly a thousand balls across thirty-six bowls and take more than three hours on stage.
The lesson is not about football. As constraints multiply, manual work does not merely get slower. It starts running into dead ends that nobody notices until the very end.
The same problem in your company under other names
What the software did in Monaco is constrained assignment: placing items into slots or time windows without breaking a rule. That is precisely where operations managers spend their weekly hours:
- Shift rotas: who works when, given the skill each post requires, rest between shifts, approved leave and minimum coverage at peak hours.
- Dispatching work orders: who is nearest, who holds the certification for that equipment, who has the part in the van, and which job is closest to its contractual due date.
- Customer appointments: which slot to offer, so nothing is double-booked and no dead gaps open between visits.
- Delivery runs: which orders travel together, in what order, within what vehicle capacity.
- Storage locations: which item sits on which rack, based on turnover, weight and separation requirements.
The only difference between that list and the Monaco draw is that UEFA knows its problem is hard, while many companies assume theirs is easy because one manager solves it every Thursday on a sheet of paper. It is not easy, it is quietly expensive: an hour of supervisor time each week, an error that surfaces on Sunday, and a staff member who says the rota is unfair with nobody able to prove otherwise.
The precondition for automation: write the rule down
The software in Monaco works for one simple reason. The rules it enforces are written and published before the draw. You cannot automate a rule that lives in a supervisor's head and only comes out when somebody challenges it.
So the first practical step is not buying a system. It is a short session where you write your rules on paper, sorted into two kinds:
- Hard rules: never breakable. No back-to-back shifts. No technician without the equipment licence. No slot booked for two customers.
- Soft rules: preferred and measured. Spread night work evenly across the team. Keep consecutive sites close together. Keep a customer with the technician who visited last time.
That separation is the whole job. Systems usually fail not because they are weak, but because nobody told them which rules cannot bend and which are merely preferences. And when you write the list, you will often discover something useful before any system arrives: two of your rules contradict each other, and the supervisor has been resolving that contradiction differently every week.
Why the ball stayed on stage
Notice that UEFA did not drop the physical draw, even though software could generate the entire table in a second. The ball stayed because people need to see the step that cannot be manipulated. Automation took the complex part, and the visible part stayed visible.
This is exactly where companies get automated scheduling wrong. The system produces a mathematically correct rota, and the team rejects it because it is a closed box. Three things make an automated schedule acceptable:
- A visible reason: the system shows why this job went to this person, even in one line.
- A logged override: the supervisor can change the result, provided the change and its reason are recorded. Blocking overrides entirely pushes people to ignore the system altogether.
- A measurement afterwards: after a month, compare what the system proposed with what actually ran. The gap is your list of missing rules.
The Origami view
What we mostly meet in operations projects is not a company that needs advanced scheduling software. It is a company that has not written its rules down yet. So we always start with rules before screens: sit with the supervisor, extract what they actually do each week, then split hard from soft. In most cases it turns out that half the lost time comes from one ambiguous rule, not from missing software.
Our second principle is that automation starts with the most repeated case, not the hardest one. Automate the ordinary assignment that recurs daily and leave exceptions to the supervisor with a clear, logged decision. Anyone trying to automate every exception on day one ends up with a complex system nobody trusts, which is worse than a paper rota everybody understands.
Where to start this week
You do not need a large project to test the idea. Run this sequence on a single assignment decision:
- Pick one recurring decision: the shift rota, work order dispatch, or appointment allocation.
- Write its rules on one page and mark each one hard or soft.
- Measure today's cost honestly: hours spent arranging it weekly, how often it was amended after publication, and how many complaints it produced.
- Apply the rules manually for exactly two weeks, as written, with no improvisation. Whatever breaks in those two weeks is the most valuable input to any system you later buy or build.
What you watched from Monaco today is not a technology story about football. It is a reminder that every organisation eventually reaches the point where its rules outgrow the paper. The difference between crossing that point calmly and being stuck at it for years is simple: are your rules written down, or still in one person's head?
Sources
- UEFA Champions League official site — league phase format, draw date and venue.
- Wikipedia — 2026/27 UEFA Champions League — number of clubs and the season's stages and dates.
- Sky Sports — 2026/27 draw, dates and format — pot and association restrictions and matchday one dates.
- Report on the Swiss model and the digital draw — the estimate of balls, bowls and time a fully manual draw would require.
Frequently asked questions
Do I need artificial intelligence to automate a shift rota?+
Usually no. Most assignment problems are solved by a rules engine that applies written conditions: required skill, rest between shifts, minimum coverage. AI helps later with demand forecasting or suggesting improvements, but the foundation is clear rules, not a language model.
What is the difference between a hard rule and a soft rule, and why does it matter?+
A hard rule can never be broken, such as no back-to-back shifts or no slot booked twice. A soft rule is a measured preference, such as spreading night work evenly or keeping consecutive sites close. The split matters because a system that treats every rule as hard will find no valid solution at all, while one that treats them all as soft will produce schedules nobody can actually run.
Our rules are unwritten and only the supervisor knows them. Where do we start?+
Start with an extraction session: sit with the supervisor while they build the schedule and ask why behind every decision until the full list fits on one page. Then apply that list as written for two weeks with no improvisation. Whatever breaks in that period is your set of missing rules, and it is the most accurate specification you can hand to any system later.
How do I get the team to accept a schedule produced by a system rather than a manager?+
Three things: the system shows a reason for each assignment even in one line, the supervisor keeps the right to override with the reason recorded, and after a month you compare what the system proposed with what actually ran. Resistance is usually not to automation itself but to a decision offered without explanation.
Follow Origami in Google
Pin Origami as a preferred source and our articles will surface first for you in Google Search and Top Stories.

Related articles
- AutomationNo-Code Automation (n8n, Zapier, Make) vs a Custom Build: Which and When?When to pick no-code automation (n8n, Zapier, Make) and when to build custom — a practical decision guide for Saudi business owners on cost, control, and scale.
- AutomationWorkflow Automation Across CRM, ERP, and AccountingManually copying data between your store, accounting, and CRM consumes hours and breeds errors. Workflow automation makes your systems talk and move on their own. A practical guide to what it means and where to start.
- Business SystemsHR and Payroll Systems for Saudi Businesses: What You Actually Need and How to ChooseA practical guide to choosing an HR and payroll system in Saudi Arabia: what it must cover, how it connects to Mudad, GOSI and Qiwa, and when custom beats off-the-shelf.
- Business SystemsThe Rodri Transfer Lesson: Why Performance Does Not Move With the Asset You BuyBarcelona are closing in on Rodri, and the question occupying analysts is the same one facing every manager who buys the best system on the market or hires the strongest candidate: does performance travel with the asset, or is it a property of the system that produced it? A technical and managerial read of a deal that is not done yet.
- Business SystemsKnowledge Management: Why It's Your Company's Most Valuable Asset — and How to Stop It LeakingKnowledge is your company's most valuable asset, yet the only one that walks out the door every evening. Knowledge management keeps your company's expertise available to your team instead of trapped in people's heads — and with AI it's more powerful than ever. A practical guide for business owners.
- Artificial IntelligenceWhat Is Goose? The Open-Source AI Agent and What It Means for Your BusinessGoose is an open-source AI agent that runs on your machine and works with any model. Learn its features, how it compares to Claude Code, and why it gives your business powerful automation at lower cost and higher privacy.
Weekly newsletter
The latest articles that matter to business owners, once a week. Just your email.
Have a project in mind?
We build custom systems, apps and websites for your business. Tell us your idea and we will give you a straight answer on it.
