Connect the Dots

Help participants experience the difficulty of estimation in complex systems first-hand. Demonstrate why seemingly simple tasks resist accurate prediction, why asking clarifying questions up front is critical, and how iterative learning improves forecasting (but never eliminates uncertainty entirely).

.Before the game

Gather the group. Explain the setup: Your mission is to connect numbered dots on a paper. Similar to the art & craft you might have played as a kid (if you are old enough). You want to connect as many as you can within a 30 second round. You will play multiple rounds.

Decide how many dots to connect. The recommended number is 20.

🌐 Online: Share the game link in the chat. Participants open it in their browser and join the session.

Step 1 - Estimate

Before the first round, ask each player to estimate how many rounds they think they need to connect all the dots (e.g. the recommended 20). Lock in the estimates before anyone plays, and note it on the chart.

👉🏼 Before you let people start, ask if everyone understands, or if there are any questions. Don't volunteer information - but answer honestly if they ask. Most groups jump straight into estimating without asking a single question. That's part of the lesson.

Step 2 - Play the rounds

Start the first round. Players have 30 seconds to connect as many dots as they can, in numerical order. When time runs out, the round ends.

Ask players to write down how many dots they connected and then ask everyone to turn their papers up side down. Next, everyone plots the results on the charts.

Play as many rounds as needed. Between rounds, deliberately pause to discuss what happened before rushing into the next one. Maybe ask if they can share any assumptions they made, or questions they should have asked.

Step 3 - Debrief

Use the charts as anchor for your debrief.

Key debrief questions:

1. What were your initial assumptions before round 1? Most people assume it will be trivially easy. Explore where that confidence came from and why it was wrong.
2. What clarifying questions could you have asked - but did not? How many dots? Can I see them first? Most teams estimate blind. Connect this to real-world planning.
3. Did estimation get easier or harder as you played more rounds? Connected lines that obscure remaining numbers might make it harder.

Facilitator tips:

• Don't explain the purpose up front. Let the experience teach. If you say "this is about why estimation is hard" before round 1, you give away the punchline.
• Ask if there are any questions before estimating - but don't prompt specific ones. If nobody asks anything, that silence is a powerful debrief moment: "You all estimated without asking a single question. How often does that happen in real life?"
• Vary the dot count for difficulty
• Give participants different charts. Give some people charts with five iterations and others with ten, and see if the chart affects people's estimations (for example that the estimated average is higher for people with ten iterations available - are they biased)
  
• Also works in pairs. Then let pairs discuss their strategy for the next round in-between each round.

Connect the Dots builds on a well-known insight from complexity science and agile practice: that estimation accuracy depends less on skill and more on the nature of the system being estimated. In the Cynefin framework, tasks in the "complex" domain resist upfront prediction because the relationship between cause and effect can only be understood in retrospect.

The exercise also draws on the idea that the questions you ask before estimating matter more than the estimation technique itself — a principle emphasised in agile coaching and elaborated in works like Douglas Hubbard's How to Measure Anything.

This digital version was created by FacilitatorKit.co to bring the exercise into online and hybrid workshop settings with built-in timers, automatic burndown charts, and zero setup overhead.

Source: FacilitatorKit.co — Connect the Dots Guide