My Son Failed the Marshmallow Test (And I'm Glad He Did)

The Setup

I placed one marshmallow on a plate in front of my 8-year-old son, Elliot.

I said: "You can eat this marshmallow now, or you can wait 15 minutes and I'll give you two marshmallows."

Then I left the room.

This is the famous Stanford Marshmallow Test, originally conducted by Walter Mischel in 1972. The children who could wait were found to have better life outcomes decades later—higher SAT scores, lower BMI, better social skills.

It's become the gold standard test for self-control and delayed gratification.

Elliot lasted 47 seconds.

He ate the marshmallow.

The Standard Interpretation: My Son 'Failed'

By traditional marshmallow test standards, this was a complete failure.

Most parenting books would say:

• He lacks self-control
• He can't delay gratification
• He needs to practice willpower
• This predicts poor outcomes later in life

And then they'd recommend interventions: teaching mindfulness, practicing waiting, rewarding patience.

But here's what I did instead:

Nothing.

I let him experience the natural consequence of his choice.

What Happened Next (The Part Mischel Didn't Study)

When I came back into the room 15 minutes later, Elliot was sitting at the table with an empty plate.

Me: "Did you eat the marshmallow?" Elliot: "Yeah." Me: "Okay. So you don't get a second one." Elliot: "I know." Me: "How do you feel about that?" Elliot: (shrugs) "It's fine. I wanted the marshmallow."

At this point, most parents would lecture. Or explain why waiting is better. Or express disappointment.

I did none of that.

Instead, I called in his 10-year-old brother, Nathan.

Round 2: The Control Comparison

I gave Nathan the same test.

One marshmallow now, or two in 15 minutes.

Key difference: Nathan had just watched his brother take the test.

Nathan waited the full 15 minutes.

I gave him two marshmallows.

Elliot watched this happen.

Round 3: The Retest (One Week Later)

One week later, I ran the test again.

Same setup. Same rules. One marshmallow now, or two in 15 minutes.

This time, Elliot waited.

Full 15 minutes.

He got two marshmallows.

Me: "What changed?" Elliot: "Last time I ate it and then Nathan got two. That sucked. So this time I waited." Me: "Was it hard to wait?" Elliot: "Yeah. But I kept thinking about getting two instead of one."

What the Research Actually Says (And Doesn't Say)

The original Mischel study is famous. But most people misunderstand it.

Myth 1: The Marshmallow Test Measures Innate Willpower

The 2018 replication study by Watts, Duncan, and Quan found that when you control for family background and environment, the predictive power of the marshmallow test drops dramatically.

Translation: Kids who wait aren't necessarily more disciplined. They might just come from more stable environments where waiting has historically paid off.

Myth 2: Failure Means the Child Lacks Self-Control

Elliot didn't lack self-control. He lacked information.

He'd never experienced the regret of choosing immediate gratification when delayed gratification would have been better.

Once he had that data, he changed his behavior.

Myth 3: You Should Train Kids to Wait

This is the one that bothers me most.

You can train a child to wait through sheer repetition and reward. But you're not teaching delayed gratification—you're teaching compliance.

Real delayed gratification requires understanding the trade-off and choosing the delayed reward because it's worth it to you.

The Behavioral Economics of 'Failure'

Here's what Elliot's 'failure' actually taught him:

Lesson 1: Opportunity Cost is Real

When you choose one marshmallow now, you give up two marshmallows later.

This is opportunity cost. It's abstract until you experience it.

Elliot didn't understand opportunity cost until he watched Nathan get two marshmallows while he had zero.

Lesson 2: Regret is a Teacher

Regret is an underrated learning mechanism.

Research by Zeelenberg and Pieters (2007) shows that regret is one of the most powerful motivators for behavior change.

Elliot felt regret when Nathan got two marshmallows. That regret motivated him to wait the next time.

No lecture required.

Lesson 3: Future Rewards Need to Be Concrete

Eight-year-olds don't have fully developed prefrontal cortexes. Abstract future rewards don't compute.

"You'll get two marshmallows later" is abstract.

"Nathan just got two marshmallows and you didn't" is concrete.

Once Elliot saw the concrete outcome, he could mentally simulate it happening to him next time.

The Extended Experiment: Money Edition

Two weeks later, I ran the same experiment with money instead of marshmallows.

I gave Elliot a choice:

He chose Option B.

One week later, I gave him $10.

Then I ran it again:

He chose Option A.

Me: "Why'd you take the $10 now?" Elliot: "Because two weeks is too long. And $5 more isn't worth waiting that long." Me: "Interesting. So there's a point where waiting isn't worth it?" Elliot: "Yeah."

This is called a discount rate.

Everyone has one. It's the rate at which you value future rewards less than present rewards.

Elliot's discount rate is:

• 1 week for 100% return ($5 → $10): Worth it
• 2 weeks for 50% return ($10 → $15): Not worth it

He's not being impulsive. He's calculating.

What This Looks Like In Real Life

Three months after the marshmallow experiment, here's what changed:

Change 1: He Asks Different Questions

Before: "Can I buy this toy?"

Now: "If I wait until next month, how much more money will I have?"

He's not asking permission. He's calculating opportunity cost.

Change 2: He Negotiates With Himself

Last week, Elliot wanted to buy a $15 Lego set.

He had $12.

Elliot: "I could buy it now if I use my savings. Or I could wait two weeks and buy it with my allowance." Me: "What are you going to do?" Elliot: "Wait. Because if I use my savings, I won't earn interest on it."

He's 8 years old and he's factoring in compound interest.

Change 3: He Identifies 'Worth It' vs 'Not Worth It' Delays

Not all delayed gratification is good.

Sometimes the immediate reward is actually better.

Elliot now asks: "Is waiting worth it for this?"

Examples:

This is sophisticated decision-making.

And it started with failing the marshmallow test.

Why 'Failure' Was Essential

If Elliot had waited the first time, he wouldn't have learned anything.

He would have gotten two marshmallows and thought: "Cool, I got more marshmallows."

But because he failed, he learned:

1. What it feels like to choose immediate gratification
2. What it feels like to regret that choice
3. What it feels like to see someone else get the better outcome
4. How to make a different choice next time

The failure was the education.

How to Run This Experiment at Home

If you want to try this with your kids:

Step 1: Run the Test (But Don't Intervene)

• Offer: 1 treat now, or 2 treats in 15 minutes
• Leave the room
• Don't lecture if they eat it
• Just observe

Step 2: Create a Visible Contrast

• If they ate it: Show them what they would have gotten ("You would have had 2")
• If possible, let them watch a sibling or friend take the test successfully
• Make the opportunity cost concrete and visible

Step 3: Retest (Don't Force, Just Offer)

• Wait 1-2 weeks
• Run the exact same test
• See if behavior changes
• Don't pressure them to wait

Step 4: Apply to Real Money

• Offer: $5 now or $10 in one week
• Whatever they choose, follow through
• Let them experience the outcome
• Don't say 'I told you so'

The Data After 6 Months

I've now run this experiment monthly for 6 months. Here are Elliot's results:

Pattern: He's not blindly delaying gratification. He's calculating whether the wait is worth the reward.

That's not impulsivity. That's rational decision-making.

The Bottom Line

My son failed the marshmallow test.

And I'm glad he did.

Because that failure taught him:

• Opportunity cost is real
• Regret is a teacher
• Delayed gratification isn't always better—it depends on the trade-off
• He can make different choices when he has better information

If I'd prevented him from failing—if I'd coached him to wait, or rewarded patience before he understood why—he wouldn't have learned any of this.

He'd have learned compliance, not decision-making.

Now, at age 8, he runs cost-benefit analyses in his head before making purchases.

He asks: "Is waiting worth it for this specific thing?"

Not: "Should I always delay gratification because authority figures say so?"

That 47-second failure was worth more than a thousand lectures.

And I wouldn't change a thing.