G: Yes. The overarching principle I would use is this:

A recipe is not merely a ratio of ingredients. It is a ratio of ingredients being subjected to a particular physical process in a particular piece of equipment.

So when you make 2×, 3×, or 5× as much food, multiplying the ingredient list is only the first operation. The more important question is whether you've preserved the conditions under which the original recipe worked. That distinction shows up repeatedly in professional recipe-scaling guidance and food science. citeturn387584search12turn387584search36turn329628search5

I'd break what you need to think about into about nine categories.

1. Ingredient quantity: start with multiplication, but don't assume everything ends there

For the bulk of a recipe—meat, vegetables, flour, sugar, dairy, stock, etc.—ordinary multiplication is the correct starting point. And once you're scaling significantly, weight becomes much better than volume. The error in “three cups of chopped something” gets magnified when three cups becomes twelve; 400 g becoming 1,600 g is unambiguous. This is why professional baking and commercial scaling work overwhelmingly in weights and percentages. citeturn329628search4turn329628search5turn387584search36

But I would resist Internet rules like “always use only 1.5× the salt when doubling.” You'll find that advice everywhere, but it is too crude to treat as a law. If you're making twice as many individually seasoned pork chops, they still need roughly twice as much salt. If you're making a giant pot of soup whose evaporation behavior changes, blindly doubling the final seasoning is much more questionable.

So the better general rule is:

Structural ingredients usually scale mathematically. Sensory ingredients should be regarded as a calculated starting point and then corrected by taste.

That especially means salt, chile heat, concentrated spices, acids, extracts, very bitter ingredients, and similar “high-leverage” ingredients. With anything you cannot take back out, reserve some until late in the process.

2. Geometry: preserve the food's shape and depth, not merely its total volume

This may be the single most useful concept.

Suppose a casserole normally occupies a 9×13 pan at a depth of 1½ inches. Doubling it into a gigantic vessel that keeps approximately that depth is quite different from doubling it in the same 9×13 pan and producing a 3-inch-deep casserole.

The quantity doubled in both cases. The cooking problem did not.

Depth governs how far heat has to penetrate. Surface area governs evaporation and browning. The amount of food touching hot metal governs searing. Surface exposed to air governs moisture loss.

This is especially obvious in baking: King Arthur points out that even the apparently modest jump from an 8-inch round pan to a 9-inch round changes capacity by about 27%, and deeper batter takes longer for heat to reach the center. citeturn329628search0

But it applies just as much to lasagna, roasting vegetables, gratins, meatloaf, braises and sauces.

A useful mental instruction is therefore:

When scaling, try to reproduce the original thickness, exposed surface, and contact with the cooking vessel.

Often that means two normal pans are superior to one enormous pan.

3. Heat transfer: your stove or oven has not become twice as powerful

This is where many scaled recipes fail.

If you put twice as much cold meat into the same skillet, you've doubled the thermal load on essentially the same burner and pan. The pan temperature crashes. Water comes out of the meat. Instead of searing, it steams.

Similarly, putting four sheet pans into an oven designed around one or two changes airflow, radiant exposure and recovery time every time you open the door.

Consequently, capacity is not just “will it physically fit?” It means:

Can your burner supply enough energy?
Can your oven circulate air around it?
Can your pan recover its temperature?
Can your mixer actually work that amount of dough or batter properly?
Can you stir the bottom of the pot adequately?

This is why professional-scale cooking often becomes batch cooking rather than simply “bigger cooking.”

If eight steaks fit physically in a skillet designed to sear four, you do not have an eight-steak skillet. You have a four-steak skillet containing eight steaks.

4. Evaporation and concentration: liquids are especially geometry-dependent

This is the category people often don't anticipate.

Imagine one quart of sauce simmering in a pot with 50 square inches of exposed surface. Now triple the sauce but keep approximately that same exposed surface. You haven't tripled the evaporation capacity. You've merely made the liquid much deeper.

So the larger batch may:

reduce much more slowly,
remain thinner,
concentrate seasoning differently,
cook its solids longer while you wait for the liquid to reach the intended consistency.

Or, if you solve the volume problem by moving to an extremely wide vessel, you can go too far in the opposite direction and increase evaporation.

Commercial recipe-scaling guidance specifically warns that larger batches can develop unexpected evaporation changes and that longer process times can alter ingredient behavior. citeturn387584search36

This means that liquid amounts sometimes need to be treated functionally rather than arithmetically.

Ask what the liquid is doing.

If it is simply “there”—say, milk providing moisture to a batter—the ratio may be critical.

If it is a cooking medium that will later be drained, like pasta water, there is no reason whatsoever to multiply it mechanically.

If it is supposed to evaporate by half and create a sauce, then the pan geometry and endpoint matter more than whether you began with exactly 2.000× as much stock.

5. Time: never multiply cooking time by the recipe multiplier

A doubled recipe virtually never takes twice as long to cook.

Instead, time depends primarily on the dimensions of the individual thing being cooked and the thermal environment around it.

Six identical chicken breasts don't individually need more time than three—provided you have enough pan/oven capacity to cook them under equivalent conditions.

A meatloaf twice as massive but formed into two identical loaves should behave much like the original.

A meatloaf twice as massive and formed into one enormous loaf absolutely will not.

Same principle for cakes. If you distribute twice the batter between two identical pans, the recipe's original baking time is a reasonable starting point. Put twice the batter into one deeper pan and you're solving a different heat-transfer problem. King Arthur and Serious Eats both emphasize pan area/depth and judging doneness rather than multiplying bake time. citeturn329628search0turn329628news24

So think:

Time scales with geometry and temperature, not with servings.

And once you've departed substantially from the original geometry, use doneness indicators—temperature, tenderness, browning, viscosity, probe resistance, reduction level—rather than the clock.

6. Technique itself may have a maximum useful batch size

Certain operations depend on physical manipulation, and there is a point where simply doing more at once makes the technique stop working properly.

Consider:

Sautéing/searing: limited by hot surface area.

Frying: limited by oil temperature recovery.

Whipping: limited by bowl geometry and mixer capacity.

Kneading: limited by mixer motor, dough-hook geometry and dough temperature.

Emulsifying: limited by shear and the rate at which one phase can be incorporated into another.

Folding delicate batters: a giant mass takes longer and more strokes to combine, potentially knocking out more air.

Reducing: limited by exposed liquid surface.

Roasting: limited by space between pieces and airflow.

This is the point at which an amateur cook starts thinking more like a professional kitchen:

“Can I scale the recipe, or do I actually need to repeat the process?”

Those are different questions.

For many recipes, 4× ingredients cooked as four 1× batches produces a much better result than one 4× batch.

7. Chemical and biological processes deserve extra caution

Baking is less forgiving because you're not merely heating ingredients; you're constructing something while chemical reactions are occurring.

Leavening, gluten development, egg-protein setting, starch gelatinization and gas expansion all have to happen in the right relationship. Bread adds fermentation on top of that.

For bread, baker's percentages solve the ingredient-scaling problem beautifully: flour is defined as 100%, and everything else is expressed relative to it. citeturn329628search5

But even then, a giant dough mass can retain fermentation heat differently from several small ones. A huge mixer can develop dough differently. Proofing containers can change the geometry.

Likewise, cakes may mathematically contain the right amount of baking powder and still fail because the batter is now too deep or sat too long before reaching the oven.

So I'd classify breads, cakes, soufflés, emulsions, candy, preserves and other reaction-dependent foods as “scale conservatively and preserve the process.”

For these, repeating a proven batch is often smarter than improvising a mega-batch.

8. Sequencing and prep suddenly become part of the recipe

There's also an organizational threshold.

Chopping one onion while butter heats is reasonable.

Chopping twelve onions while butter heats is a way to burn butter.

A recipe written for four servings may casually say:

“Meanwhile, dice the vegetables.”

At 5× quantity, that sentence may now describe twenty minutes of labor.

Likewise, adding flour gradually, draining something immediately, folding something while warm, or transferring cooked food before the next step can become logistically impossible if you haven't scaled your mise en place along with the ingredients.

So for larger cooking:

Prep increasingly needs to precede heat.

Have the ingredients measured, pans available, landing places cleared, sheet trays ready, thermometers out, serving containers prepared, and ideally the scaled quantities written down rather than doing mental multiplication while cooking.

This is one of the less glamorous but more consequential differences between cooking four portions and cooking forty.

9. Food safety becomes a scaling issue too

Large quantities heat and cool slowly.

A gallon of chili put into the refrigerator in one deep stockpot does not behave like four cups in a small container. Its center can remain warm for a long time.

USDA specifically recommends dividing large quantities into smaller, shallow containers for rapid cooling rather than putting a large pot of soup or stew directly into refrigeration. citeturn685323search0turn685323search1 The FDA Food Code's professional standard is cooling cooked temperature-controlled food from 135°F to 70°F within two hours and then to 41°F within another four hours. citeturn685323search36

Likewise, serving a large group means food may sit around much longer than dinner for four. Holding temperature, serving batches, refrigerator space and leftover containers therefore need to be part of the scaling plan.

I think all of that can actually be compressed into one practical model:

When you enlarge a recipe, ask yourself whether you are preserving these five things:

Ratio — Geometry — Energy — Process — Endpoint.

Ratio: Did the ingredients scale appropriately?

Geometry: Is the food approximately the same depth, thickness, spacing and exposed surface?

Energy: Does the equipment have enough heating/cooling/mixing capacity?

Process: Can you still execute the same browning, mixing, reduction, fermentation, emulsification, etc.?

Endpoint: Are you judging the food by what it is supposed to become rather than blindly following the original time?

If all five are approximately preserved, doubling a recipe really can be almost as simple as doubling it.

If one changes substantially, you need an adjustment.

And if several change at once, you've crossed the line where I would stop thinking of it as “the same recipe, only bigger.” You're effectively developing a large-batch version of the recipe, which should be tested and written down as its own formulation. Commercial recipe-scaling guidance does exactly that: scale experimentally, document the changes, and treat the successful larger version as the new formula. citeturn387584search36

That last distinction is probably the semi-professional leap: don't assume scalability; validate it. Once you've learned that your chili works beautifully at 3× with two wide pots, slightly less initial stock, and seasoning corrected after reduction, write that down. Next time you're not improvising—you now have a tested 3× recipe.

料理(りょうり)の量(りょう)を増(ふ)やす時(とき)、材料(ざいりょう)だけではなく、熱(ねつ)と形(かたち)も考(かんが)える。 — When increasing the amount of food, think not only about the ingredients, but also the heat and the shape. memcite

G: Yes. Your coconut-rice example is basically the right way to think about it, and your “minimum dimension” idea is very close to a useful general rule. I would modify the terminology slightly because there are really two different physical problems hiding inside it.

For the coconut rice: if you doubled the recipe and moved to a proportionally larger saucepan so that the rice/liquid layer remains roughly the same depth, you've preserved most of the important geometry. Once it reaches the prescribed simmer and you cover it and turn the heat down, you're creating a relatively closed, humid system. Water evaporates, hits the cooler lid, condenses, and returns; the recipe isn't relying primarily on continuous evaporation into the room to cook the rice. Consequently, once that larger batch has actually reached the same thermal state as the original, the covered cooking time shouldn't increase dramatically just because there's twice as much rice.

There is, however, an important scaling effect before that point: twice the rice + coconut milk + water represents roughly twice the thermal mass. The burner hasn't doubled in power. So starting from the same ingredient temperature, the larger batch will generally take longer to come up to temperature. Once you've reached the simmer, though, you've largely eliminated that initial difference. In other words, if the recipe says something like “bring to a simmer, cover, reduce to low, cook 15 minutes,” the 15-minute clock begins after you've reached the simmer. That's extremely convenient because the recipe itself has effectively built in a thermal reset point.

Your hypothetical evaporative rice recipe is also right, with one refinement. Liquid depth is important, but exposed surface area relative to the amount of liquid is the more fundamental variable. If you double the liquid and double the exposed surface area, you've approximately preserved the amount of liquid available per unit of evaporating surface. Keeping the same depth in geometrically similar straight-sided pans accomplishes approximately that.

That's quite different from doubling the recipe in a pot that's merely taller. Same exposed surface + twice the liquid = evaporation becomes a bottleneck.

And that brings us to your minimum-dimension rule, which I think is worth keeping, but I'd formulate it as:

For heat penetration, ask what is the longest distance heat has to travel to reach the least-accessible part of the food.

For a steak, that's basically half its thickness. Width and length barely matter for cooking the center. A 4 × 4 × 1-inch steak and an 8 × 8 × 1-inch steak have radically different masses, but the center of each is still only about half an inch from a major hot surface. Under otherwise comparable conditions, their internal cooking times can therefore be surprisingly similar.

That's why doubling the number of steaks doesn't inherently increase their cooking time either. The problem becomes energy delivery at the surface: crowd the skillet and the steaks dump water and thermal energy into the pan faster than the burner can compensate. Now the surface temperature collapses and the cooking mechanism changes from effective searing toward steaming.

For something like a baked potato, conversely, the important dimension is essentially the radius of the potato. Twenty potatoes of the same size do not require twenty times the cooking time. But one potato twice the diameter is a dramatically different thermal problem because heat now has much farther to penetrate.

And that connects directly to your original oven question.

Putting twice as many cookies or potatoes into a 350°F oven doesn't change what 350°F means, but it changes how difficult it is for the oven to maintain the conditions associated with 350°F.

There are three important effects.

First, thermal load. Opening the oven loses heat anyway, and then you're inserting perhaps 10 pounds of 70°F potatoes instead of 5 pounds. That mass absorbs a great deal of energy. The thermostat sees the resulting temperature drop and turns the heating elements/burner on, but recovery takes time. A commercial oven with enormous burners may shrug this off; a small domestic oven may not.

Second, airflow and radiant exposure. One sheet of cookies sitting in the center of the oven gets a particular combination of hot moving air and infrared radiation from the oven surfaces/elements. Two or three sheets can obstruct one another. The upper sheet may shield the lower one from radiant heat; crowded pans impede convection; different shelves occupy different thermal environments. This is why rotating trays becomes useful when heavily loading a home oven.

Third, and subtly, moisture. Food is constantly releasing water vapor. Twice as much food means potentially much more moisture entering the oven atmosphere. In a well-ventilated oven much of it escapes, but a heavily loaded oven can temporarily become more humid. That suppresses evaporation from food surfaces, which can retard browning and crisping. This matters much more for something like roasted vegetables or French fries than for potatoes you're merely trying to cook through.

So there's an interesting distinction:

One potato → twenty identical potatoes: penetration distance hasn't changed; oven capacity and circulation become the concern.

One 8-ounce potato → one 16-ounce potato: oven loading barely matters; penetration distance becomes the concern.

One tray of roasted potatoes → four crowded trays: both become concerns, plus evaporation and browning.

That gives you a slightly more sophisticated version of the rule you were developing. When scaling food, ask two separate questions:

How far must the heat travel through each piece?
and
How much food is competing for the available heat-transfer capacity?

The first determines a great deal of the cooking time. The second determines whether your equipment can reproduce the original cooking conditions.

And there's a nice consequence of this that explains a lot of otherwise mysterious cooking advice: cutting something into smaller pieces can reduce cooking time dramatically even though its total mass hasn't changed at all. You haven't removed any food; you've shortened the distance that heat has to penetrate while enormously increasing surface area.

So your intuition about “minimum dimension” was pointing directly at one of the fundamental principles. I'd just rename it something like penetration distance, because sometimes it's the smallest dimension (steak thickness), sometimes it's effectively the radius (potato), and sometimes geometry gets complicated enough that “minimum dimension” stops describing exactly what matters.

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: It can help a little, but there is no reliable rule like “double the food, add 10°F.” The reason is that the oven is not just a temperature setting; it is a heat-delivery system with a finite amount of stored heat and heating power. A higher preheat gives you a larger reservoir of thermal energy in the oven walls, racks, and air, so the temperature drop after loading can be smaller or recover faster. But the effect depends enormously on the oven, the mass and temperature of the food, how long the door is open, what kind of pans you use, and whether the food itself is sensitive to an initially hotter environment.

For something forgiving like baked potatoes, roasted vegetables, or a large roast, a modest “heat boost” can be useful. For cookies, cakes, custards, or anything where early surface setting matters, it can backfire by over-browning the exterior, changing spread, or setting structure too quickly. In those cases, maintaining the intended oven environment matters more than compensating aggressively for the load.

The most useful numeric idea is not a cooking formula so much as an energy-balance concept:

[ Q = mc\Delta T ]

where (Q) is the heat energy needed, (m) is the mass of what you're heating, (c) is its heat capacity, and (\Delta T) is the temperature rise.

So if you literally double the mass of food at the same starting temperature, you've roughly doubled the amount of heat the oven must transfer into it. But that does not mean doubling the oven-temperature difference. It usually means the oven takes longer to recover or must run its heating elements longer.

A practical rule I'd use is:

Don't scale oven temperature with recipe quantity. Scale your concern about oven recovery and airflow.

If you want a useful household technique, you can often preheat 25°F hotter than the target, load the oven, and immediately or within a minute or two return it to the recipe temperature. That can compensate for door-opening and a heavy load in a conventional home oven. But I would treat even that as an equipment-specific trick, not a universal law. For delicate baking, I would usually skip it.

An even better technique is to preheat thoroughly, not just until the oven beeps. The thermostat may say 350°F while the oven walls and racks are still warming. Giving the oven an extra 15–30 minutes after it first reaches temperature increases the stored thermal energy substantially without exposing the food to a higher-than-intended temperature.

So if you want something concise enough to remember:

More food → usually same set temperature, longer recovery, possibly longer cook time.
Heavy load → consider +25°F preheat, then return to target after loading.
Delicate baking → don't use the boost unless you know your oven.
No meaningful “+10°F per doubling” rule exists.

If you wanted to get a little nerdier, the closest thing to a useful scaling parameter would actually be food mass divided by oven heating power, not food mass divided by recipe size. A 2,400-watt oven handles a doubled load differently from a 1,500-watt one.