Regression to the Mean: Why Your Best Day is Probably Followed by a Worse One

The Coach's Mistake

A basketball coach notices something weird.

Every time a player has an amazing game, their next game is more mediocre. So the coach stops complimenting them — doesn't want to jinx it.

Every time a player has a terrible game, the coach screams at them. And sure enough, their next game is almost always better. So the coach becomes convinced that yelling works.

The coach is wrong about almost everything. But statistically, they're seeing something completely real.

Welcome to regression to the mean — one of the most counterintuitive ideas in math, and one of the most important ones for how you think about literally everything in your life.

What's Actually Happening

Here's the key insight: extreme results are usually a combination of skill and luck.

Your best-ever test score probably reflected both your real knowledge and a lucky streak — you slept great, guessed well on the tough questions, got questions that happened to be your strengths.

Your worst-ever score probably reflected your real knowledge and bad luck — you were tired, stumbled on a question you usually get, the specific topics that tripped you up were overrepresented.

Neither extreme is purely "you." Both are you plus random noise.

And random noise? It doesn't stay extreme. It averages out.

So after your best-ever performance (high skill + high luck), the luck component will very likely be lower next time — pulling your result back toward your actual average. After your worst-ever performance (normal skill + bad luck), the luck component will likely be less bad next time — pulling your result up.

This "pull toward the average" is regression to the mean. It's not karma. It's not the universe punishing success or rewarding struggle. It's just... math.

Why This Messes With Our Heads

Humans desperately want explanations. When something gets better, we want to know why it got better. When something gets worse, we want to know what caused it.

So we invent stories:

• "I congratulated them and they got worse. Praise causes complacency." → Wrong. Regression to the mean.
• "I yelled at them and they improved. Pressure works." → Wrong. Regression to the mean.
• "I tried a new pre-game ritual and then had my best race. The ritual works!" → Probably wrong. You had an extreme result. Mean was coming for you either way.
• "I changed my study method after bombing that test and my grades improved." → Maybe it helped. Or maybe you just had a bad day and were due to bounce back. You can't know without a control.

This is why bad coaches, bad teachers, and bad managers often think punishment works better than reward: they tend to apply punishment after unusually bad performance, which regresses upward naturally. They take credit for the math.

Real-Life Hall of Fame

The Sports Illustrated Jinx

Athletes who appear on the Sports Illustrated cover often have worse seasons afterward. Eerie, right? Except: you end up on the cover because you just had an exceptional run. Exceptional runs contain luck. Luck regresses. There's no jinx. There's just mean.

Miracle diets and supplements

People try a new diet supplement right after hitting their highest weight ever. They lose weight. "It worked!" Maybe. Or maybe extreme high weight is partly extreme circumstances (holidays, stress), and things were going to normalize anyway. The supplement gets the credit.

Investment fund performance

Funds that beat the market one year are marketed heavily. They usually underperform the following year. Partly this is luck: the best-performing funds in any given year were partly lucky. Luck regresses. This is why financial advisors saying "past performance indicates future results" are saying something statistically misleading.

Your actual life

You had the best day ever — everything clicked, you were on fire, everyone liked your posts. Then tomorrow is just... fine. Normal. And it feels like a letdown.

It's not. Your "best day ever" was partly you being great and the random universe being kind. Both won't peak at the same time again immediately. You're not declining. You're just averaging out.

How to Use This

Stop taking credit (or blame) for regression:

After a high point, prepare for normal — not because you failed, but because that's how averages work.

Give things time before judging:

New coach, new teacher, new routine — if you try something new right after a low point, of course things will improve. That's regression, not the new thing working. Try to evaluate across a long stretch of normal performances.

Be suspicious of "it always works after a rough patch":

This is one of the most common ways luck masquerades as strategy.

Understand variance:

The more volatile something is (mood, sports performance, test scores), the stronger regression to the mean gets. High variance = more extreme highs and lows, and more dramatic regression.

The Challenge

Think of a time something in your life went from extreme (very good or very bad) to more normal.

• Did you attribute it to something? A decision, a change, an action?
• Could it have been regression to the mean instead?
• How would you actually test whether your explanation was right?

Write it out. Bonus points if you can think of a case where someone else misread regression as evidence that their action worked.

The universe is not punishing your good days or rewarding your bad ones. It's just doing math. Understanding that is genuinely freeing — and it makes you much harder to manipulate.