When it made things WORSE (and that's still data)

What you tested

Cutting my toddler's afternoon nap cold turkey to fix a 9pm bedtime. One change, ran 5 days.

Target

Time from lights-out to asleep at night, baseline ~40 min; plus number of evening meltdowns.

Data: before → after

Baseline: asleep ~40 min after lights-out, ~1 evening meltdown/day. No-nap week: asleep FASTER (~15 min) BUT evening meltdowns jumped to 3-4/day and a 5:15am early waking appeared on days 4-5.

What it means

The change 'worked' on the narrow target (faster to sleep) and clearly backfired on everything around it. He was overtired — quicker to crash, but a wreck by 5pm and undersleeping overall.

What's next

Reverting the cold-turkey cut. New single change: shorten the nap to 45 min and cap it by 2:30pm instead of removing it. Re-run and re-count.