I was grabbing his failures before he could feel them

7:14 AM. Backpack on, homework still sitting on the counter. Almost said "grab it" — caught myself. Turns out productive failure is a thing in education research, and stealing the failure was maybe the bigger mistake.

7:14 AM. Homework on the kitchen counter, where it landed last night.

"Grab it" was about to come out. I caught it.

Last night, he said he'd take it himself. If I say it now, that's me doing the inventory check again, not him.

Backpack on, "bye dad" at the door.

Homework still on the counter.

The 3-second pause

I'm trying this lately: when I notice he's forgetting something, or messing something up, or stuck on something — I wait three seconds before saying anything.

Three seconds, and usually one of these happens:

• He notices on his own, comes back, grabs it
• He doesn't notice, leaves without it (and something downstream happens)
• He still doesn't notice and I'm the one who ends up dealing with it later

One of those three.

If I jump in, the first option goes away. The half-second before he would've caught it himself — if I fill that with my voice, his brain logs it as "dad reminded me." Not as "I noticed."

That afternoon, he came home crying

The homework day.

3:30 PM. Door opens. Eyes red.

"The teacher got mad at me."

"Oh — you didn't turn it in." "No."

He cried for five minutes, I think.

Then, quiet: "Next time I'll check the counter."

If I'd said "grab it" at 7:14 AM, that line never gets said.

"Learning from failure" is also a research thing

Education research has a name for this.

Productive failure (Kapur, 2008)¹.

Short version: kids who get handed a problem first and struggle with it on their own end up with deeper understanding than kids who get the solution explained first. The failing itself is part of the learning.

Getting stuck, getting it wrong, hitting a wall — that becomes the scaffolding for the next attempt.

When a parent jumps in early, the scaffolding doesn't get built. It's like only showing kids the worked example, then handing them the real test cold.

This is also kind of how we think about the Todo apps at Lyco. The "you got partial credit for partial effort" thing in TodoFarm came from this — you don't have to finish everything to have built something. Mistakes are fine too. Our app is a Todo app, not a Done app.

"Just let them fail" isn't the rule either

I don't think "let them fail at everything" is the actual move.

• Anything dangerous (running into the street): stop it
• Anything that can't be taken back (hurting a friend): stop it
• Stuff that just keeps repeating without becoming a lesson (sensory-related, etc.): scaffold it

What's left is non-catastrophic, recoverable failures where the next-time delta is real.

For us, that looks like:

• Forgot the homework → leave it (gets corrected by the teacher)
• Forgot the jacket and it's chilly → leave it (he checks weather next time)
• One sock missing in the morning → leave it (he packs the night before next time)
• Running into the parking lot → stop, every time
• Shoving his little brother → stop, every time

The line moves. House to house, kid to kid, even day to day. Cant fix it once and call it done.

The guilt doesn't go anywhere

Honest part: after a "I let him fail" day, the guilt doesn't vanish.

That 3:30 PM, watching him cry — yeah, "should've said it at 7:14" crossed my mind.

But I'm starting to think carrying that guilt is the work. The scaffolding his learning sits on is made out of my guilt — overstated maybe, but kinda.

Snatching the failure away from him is, maybe, the actual mistake. Once in a while, he forgets something, and learns something.

While he's still at an age where the failures are small enough to recover from — that's the window. Our job is just to keep that window open. Failing isn't bad. There's nothing to be mad at him for.