Play is described as children's work, which is a nice phrase and obscures the specific developmental functions that different kinds of play appear to serve.
Physical play
Develops motor control, balance and spatial awareness.
Rough and tumble play, in particular, appears to support social skills — learning about force, reading signals, and managing arousal.
Which is why it exists across mammal species, and why suppressing it entirely may remove something useful.
Pretend play
Involves representing one thing as another, which requires holding two ideas simultaneously.
Which is associated with development of the ability to understand that others have different beliefs and knowledge.
The causal evidence is weaker than the correlational evidence, and the association is consistent.
Language use during pretend play is typically more complex than during other activities, which is a documented finding.
Construction play
Building with blocks and similar materials.
Studies have found associations between block play and spatial reasoning, and some intervention studies have found effects on later mathematical performance.
Which is among the better-supported specific claims about play types.
Risky play
Play involving height, speed, tools, and being out of adult sight.
Research in this area finds that children seek it deliberately and that it appears to serve a function in learning risk assessment.
Which conflicts with the trend toward removing risk from play environments.
Studies comparing playgrounds with different risk levels have found no increase in serious injury with greater challenge, and some evidence of reduced injury as children develop competence.
The distinction that matters is between risk, which the child can assess, and hazard, which they cannot see — a height they can judge against a rotten branch they cannot.
Free play against structured activity
Children's time in structured activities has increased substantially over decades in several countries.
Which reduces free play, and research associates free play with development of self-direction and executive function.
Studies comparing children with more and less structured time have found differences in self-directed executive function, with the direction favouring less structure.
Outdoor play
Associated with physical activity, vitamin D, and in several studies with reduced rates of short-sightedness.
The myopia finding has reasonable support, with time outdoors appearing protective independently of near work, and light exposure being the proposed mechanism.
Adult involvement
Guided play — adult-supported but child-led — has evidence for learning outcomes exceeding both free play and direct instruction for some objectives.
Which suggests the useful adult role is scaffolding rather than either directing or withdrawing.
Commenting on what a child is doing, asking open questions and extending their idea are the behaviours associated with this.
What is not required
Expensive or educational toys.
Studies comparing play with simple open-ended materials against electronic toys have generally found more language and more complex play with the simple materials.
Which is a consistent finding and it is the opposite of what the market suggests.
Boredom
Periods without provided activity are associated with self-directed play and, in some research, with creative output.
Which is a modest literature and a plausible one, and it argues against filling all available time.
Children accustomed to constant provision generally take a period to adjust before generating their own activity.
Loose parts
Materials without a defined use — boxes, fabric, sticks, containers — support more varied play than single-purpose toys.
Which is a well-established principle in play provision, and it is why the box frequently outlasts interest in what came in it.
Playground designs incorporating movable materials have shown increases in play variety and duration.
Play and disability
Access to play is unequal, with playground design and social barriers both contributing.
Which has driven inclusive design standards covering surfacing, equipment and sensory considerations.
Inclusive design generally improves provision for everyone rather than only for the intended group.
Screen-based play
Varies enormously in what it involves, from passive consumption to construction and collaboration.
Which means treating it as a single category obscures the differences, and the same distinctions that apply to physical play apply here.
Play and language
Adult talk during play predicts vocabulary development strongly.
Which is about quantity and quality — describing, extending, asking open questions rather than testing.
Studies measuring conversational turns rather than word counts find turn-taking to be the stronger predictor, which suggests interaction rather than exposure is the mechanism.
Play deprivation
Studied in institutional settings and in highly restricted environments, with substantial developmental effects.
Which is at an extreme not relevant to most families, and it establishes that play is not optional.
Age-appropriate expectations
Parallel play, where young children play alongside rather than with each other, is a normal stage rather than a social difficulty.
Cooperative play develops gradually, and expecting it too early creates unnecessary concern.
Space and time
What children need most is unstructured time and somewhere to use it, which are the resources that have declined.
Which is a provision question as much as a parenting one, and access to safe outdoor space varies enormously.