What to cook In the mood Workshop Scenarios Actually easy Techniques Dishes Ingredients Diets Recipes
Home · Scenarios · Muhallebi — milk pudding
Scenario

Muhallebi — milk pudding

Milk, starch, sugar. Three ingredients and one threshold: nothing happens until the starch sees 62 °C, and once it has there is no going back. The smallest scenario in the set, because the only variable is a ratio.

Amounts

Change them and see

Ratio 8.0%

Properly set

A spoon leaves a line and the line stays. It firms further as it cools, so a pudding that looks right while hot ends up too firm.

The timeline

In order, minute by minute

Not a list of steps but a timeline: each one carries the minute it happens at and the temperature it happens at. Changing the order changes the result, and this shape is why the engine can say so.

  1. 0 minutes

    Slake the starch into a little of the cold milk until no lumps are left.

    Starch granules dropped dry into hot liquid gelatinise on the outside instantly and stay dry within. That is what a lump is, and it does not break up later.

    What to look for The slurry should pour smooth, like cream.

  2. 2 minutes 60 °C

    Warm the rest of the milk with the sugar over a medium heat, stirring throughout.

    Sugar competes with the starch for water and pushes gelatinisation a couple of degrees higher. A sweeter pudding sets a little later.

    What to look for Steam at the edges, no boil.

  3. 8 minutes 62 °C
    Critical point

    Add the slurry to the warming milk and do not stop stirring.

    At 62 °C the granules start taking up water and swelling; towards 70 they burst and build the network. Stop stirring in that window and you get scorching underneath and lumps on top.

    What to look for The back of the spoon coats, and a finger drawn through it leaves a line that stays.

  4. 12 minutes

    Divide into bowls, press film onto the surface so no skin forms, and cool.

    Most of the body arrives as it cools, when the starch chains reorder. What looks right while hot comes out one notch firmer.

    What to look for Get it near room temperature before it goes into the fridge.

What you get

The numbers that decide it

Texture

Properly set

Read straight off the starch-to-milk ratio and nothing else.

Sweetness

80

Twelve to eighteen per cent of the milk is the classic band. Below it the milk itself comes forward; above it the sugar masks the chalkiness of the starch.

Appearance

A pudding that never reached 62 °C stays dull and cloudy; gelatinised starch is what gives it that translucent sheen.

What is happening in the pot

  1. 62–70 °C · no way back Starch gelatinised The granules soaked up water, swelled and burst into a network. It has thickened, and it will firm further as it cools. Held in the system
When it goes wrong

What happened, why, and what to do

Some of these can be saved and some cannot. Writing up the second kind as though it were the first costs somebody another twenty minutes.

01 It went lumpy
Why

The starch went into hot milk dry. Its outside gelatinised at once and its inside stayed dry.

Can be saved

Blend it or pass it through a sieve. The texture never comes all the way back, but it is perfectly good.

02 It never set
Why

Either it never reached 62 °C, or the starch was under 5% of the liquid.

Can be saved

Back on the heat, stirring, up to 70 °C. If the ratio was low, slake another spoon of starch in cold milk and add it.

03 It came out rubbery
Why

The starch passed 14% of the liquid. The network built too tight.

Cannot be saved

It cannot be saved. Adding liquid and reheating does not loosen the network; it gives you a watery rubber.

04 It caught on the bottom and smells scorched
Why

Stirring stopped after gelatinisation began. The network settled on the base and went past 100 °C there.

Cannot be saved

Lift the unburnt top into another pan and do not scrape. Once the scorched taste is stirred through, it stays.

What could not be checked

The safety check does not know the following. That is not the same as nothing being wrong — it means nobody looked, and anything meant to keep at room temperature has to be measured rather than assumed.

  • Water activity
How firm is this

This is a model estimate rather than a measurement. How far each threshold behind it has been checked is below; fourteen of the fifteen still have no source on file, and saying so is better than not.

Starch gelatinisation · Held in the system Boiling · Held in the system

engine 1.0.0 · ruleset 2026.09.1 · confidence medium

Sources and method →