Why the sleep chart doesn't match your baby
Updated: Sep 26
You found the chart at some point in the first few weeks. Everyone does. Ages down one side, hours across the other, tidy little boxes telling you what should be happening.
And then your baby did something else.
Where the numbers actually come from
A sleep chart is a summary of what a group of babies did. Researchers collect data from a population, work out the middle, publish the range.
It is useful information. The National Sleep Foundation's ranges came from an expert panel reviewing hundreds of studies, and they tell you what is broadly typical, which is how we know that a newborn sleeping 16 hours and a newborn sleeping 14 hours are both unremarkable.
The panel also published a second set of hours that may be appropriate for some individuals, sitting outside the headline range. The people who made the chart are telling you the chart has edges.
What it cannot do is describe an individual. The average is a description of a crowd. Your baby is not in the crowd, they are in your house, doing their own thing, on their own timetable.
The bit that does the damage
The information is not the problem. The problem is what happens when a population average gets turned into a personal target.
That happens constantly. A chart becomes a schedule. A schedule becomes a thing you are behind on. And then a normal Tuesday, where your baby napped for forty minutes instead of ninety, arrives feeling like a failure rather than a Tuesday.
Nothing went wrong. Your baby is not late. They were never on the timetable, because the timetable was made of other people's babies.
What prediction apps add to this
Plenty of tracker apps take that same population data and go a step further. They run your logs through an algorithm and hand back a prediction: here is when your baby should next sleep.
It sounds like help. In practice it hands you a clock to watch, and when your baby does something different, and they will, it lands as though you got something wrong.
You did not. The prediction described an average baby. There isn't one of those in your house either.
This is why there is no AI in Littles. Nothing predicts, nothing suggests, nothing quietly scores your day. That is a deliberate choice, not a missing feature.
What to look at instead
Your own baby's pattern, built from your own baby's days. It takes about a week of logging to see. Not a perfect week. A real one. Once you can see it, a few things change:
Short naps stop being an emergency. You can see where the sleep went instead.
You spot your own baby's rhythms. Often nothing like the chart's, and completely consistent for them.
You get something solid to bring to a professional. A real record of the last fortnight is worth far more than a best guess made over a cup of tea you didn't finish.
And a quieter one: you stop watching the clock. A surprising amount of the exhaustion in the early months comes from monitoring rather than from the actual broken sleep.
A fair question
Is any of this saying charts are useless? No. Ranges are helpful for spotting when something is well outside what is typical, and that is worth knowing. If your baby is sleeping far more or far less than the broad range, or something has changed suddenly, that is a conversation with your GP or health visitor.
The point is narrower than that. A range tells you roughly where the crowd sits. It cannot tell you where your baby sits. For that, you need your baby.
Littles is free to download, and logging feeds, sleep and nappies stays free for as long as you want it.
Your baby's own 24 hour sleep needs, and the rest of the analysis screens, are part of premium. Two weeks free to try, then £6.99 a month or £50 a year.
Source: National Sleep Foundation sleep duration recommendations (Hirshkowitz et al, Sleep Health, 2015).
Related reading: How much sleep does a baby need in 24 hours?
