Loser Buys Lunch

THE SPARK | a simple bet

Christian and I were curious about the annual revenue of another jewelry company. Instead of instantly looking it up, we each wrote down a guess, compared notes, then checked the actual number.

He won. I bought tacos.

That was fun. Not the losing part, but the thinking part. We’d taken a simple question and elevated it to a test of our business intuition. When he approached me with “Want to make our prediction contest a quarterly thing,” I immediately said, “IN!” His idea triggered a wonderful memory.

THE ORIGIN | my grandfather’s Olympics

When I was a kid, my grandfather and I created our version of the Olympics. Each of us would list 20 games we wanted to compete in – board games, pitching pennies, card games, casting contests. Any game on both our lists made the final cut. Then we’d take turns eliminating the other person’s suggestions until we had 20 total events.

We’d negotiate the rules together. Checkers was best four out of seven. Gin was first to 500 points. We created a list, and whenever I visited, he’d say, “Pull out the Olympics plan.” We’d pick an event, compete, and track our records.

It wasn’t about winning, although I certainly cared about that. It was about having a shared structure for engagement.

THE FRAMEWORK | why bother making predictions

Forty-five years later, Christian’s suggestion brought back the same structure, only this time, instead of competing at checkers, we’re competing at seeing the future clearly. We’re each creating 20 predictions for Q1, comparing our lists, and building our own Olympics of forecasting. Some will be obvious – Super Bowl, Oscars, and, of course the Winter Olympics. But we’re also reaching into technology, business, and science for predictions that will expand our thinking.

Making predictions makes you sharper in ways setting goals alone can’t.

Goals are about what you can control. Run 1,000 miles this year, launch a new product, get that promotion, sell $1M.

Predictions are about what you don’t control but try to understand. Which country wins the most Olympic medals. Whether AI regulation passes congress. What will the price of gold be.

Making predictions is a strong exercise that offers several benefits:

• Sharpens thinking and accountability. Predictions force you to clarify your assumptions and reasoning. The discipline helps you think more rigorously. When I predicted that company’s revenue, I had to ask: what’s their store count, average ticket, conversion rate? That drove me to drill down into their business model.
• Creates a feedback loop for learning. They provide concrete data points to evaluate and refine your judgment over time. If I predict the Oscars wrong three years in a row, I learn that my model of best movie wins is flawed. The Academy values different things than I do.
• Reduces overconfidence. Tracking outcomes tends to calibrate you toward more realistic assessments.
• Improves decision making. The practice leads to better strategic choices because you’re constantly testing your understanding of cause and effect.
• Builds strategic foresight. Prediction exercises train you to think probabilistically rather than deterministically. You stop thinking “X will definitely happen” and start thinking “X has a 60% chance based on these factors.”

THE PRACTICE | how to get better

You don’t have to develop a structured practice like Christian and I are doing. Here are five simple practices you can do to develop better prediction skills.

1. Start tracking (foundation). Write down specific predictions with probabilities and dates. Then review what actually happened. Use a simple spreadsheet or notebook. Recording creates accountability and makes patterns visible over time.
2. Explain your reasoning (depth). The prediction is less valuable than your explanation of why. What was your mental model? What evidence were you relying on? What assumptions are you making?
3. Make lots of predictions, not just big ones (volume). Practice on everyday stuff. Will this meeting run over, will this proposal get approved on the first pass, how many people will show up for the training session. High volume gives you faster feedback loops.
4. Do post-mortem on both hits and misses (learning). When you’re right, figure out what you understood correctly. When you’re wrong, identify where your model failed.
5. Find prediction partners (sustainability). Share predictions with someone who’ll follow up and review them with you. The social element adds useful pressure.