The claim everybody repeats, and why it cannot be right
You will meet the number 18 constantly in this category. A steel conducts heat 18 times better than a stone. A steel stores 18 times more energy than a stone. Both sentences appear, sometimes on the same page. They cannot both be true, because conductivity and stored energy are different quantities with different ratios.
So we computed all three of the ratios that could plausibly be meant, from published material constants:
| Property | Carbon steel | Cordierite | Ratio |
|---|
| Thermal conductivity, W/m.K | 50 | 1.3 to 1.7 | 29x to 38x |
| Volumetric heat capacity, MJ/m3.K | 3.71 | 1.73 | 2.1x |
| Thermal effusivity, W.s^0.5/m2.K | 13,625 | 1,612 | 8.5x |
| Thermal diffusivity, m2/s | 1.35e-5 | 8.7e-7 | 15.6x |
Nothing lands on 18. The closest is diffusivity at 15.6, which is a property almost nobody quoting the figure could name. The stored-energy version of the claim is out by a factor of nine.
Credit where it is due: NerdChef states 20 times the conductivity and twice the heat capacity, as two separate figures. Their heat-capacity number matches our independently derived 2.1x almost exactly, which is why their spec sheet ranks second here on a criterion that has nothing to do with the plate itself. We checked their conductivity claim as well, and it turns out to understate the published constants - the full NerdChef review is here.
The number that actually decides your crust
Effusivity. It is the square root of conductivity times volumetric heat capacity, and it governs the contact temperature when two bodies at different temperatures touch. That is precisely the physics of a 70 F wet base landing on a 500 F plate.
8.5x
Carbon steel's thermal effusivity against cordierite's
Not 18x, and not 2x. Effusivity is what sets how far the contact patch cools when the base lands, and it is the only one of the four ratios above that is about crust rather than about the plate.
High effusivity means the surface barely notices the dough arriving and keeps pushing heat into it. Low effusivity means the contact patch drops and takes time to recover, which is the mechanism behind a pale, slightly steamed base.
A steel is not faster to preheat. It is slower.
This is the most useful thing on this page and it is the opposite of the marketing. Internal conduction is not the bottleneck in a preheat - the Biot number for a quarter-inch steel is 0.001, which means heat crosses the surface roughly a thousand times more slowly than it travels through the plate. Mass is the bottleneck, and steel is dense.
| Surface | Published mass | 95% of oven temp | 99% | Energy to 500 F |
|---|
| Old Stone 16" cordierite | 7.3 lb | 25 min | 38 min | 0.18 kWh |
| Lodge 15" cast iron pan | 9.9 lb | 23 min | 36 min | 0.15 kWh |
| Baking Steel Original, 1/4" | 15 lb | 27 min | 41 min | 0.21 kWh |
| Baking Steel Pro, 3/8" | 27 lb | 42 min | 64 min | 0.38 kWh |
| NerdChef Ultimate, 1/2" | 32 lb | 55 min | 85 min | 0.45 kWh |
Two things fall out of that table. The quarter-inch steel takes two minutes longer than the stone, not less. And Baking Steel's own published guidance for the Original - 45 to 60 minutes at 500 to 550 F - brackets our computed 41-minute figure for 99 percent. Their number and our number were arrived at completely independently and they agree, which is the strongest validation available to a site that owns none of this equipment.
What we did not do
We have not baked on any of these. Every guide in this category says it tested a dozen stones; we would rather say plainly that we did not. What we did instead is on this page in full: published specs with their sources, thermal arithmetic from material constants with the working shown, and the makers' own manuals read rather than paraphrased.
What we would skip
Half-inch steel, for almost everybody. 32 pounds, 55 minutes to 95 percent of oven temperature, and no visible benefit on a single pizza over a quarter-inch. Buy it only if it is going onto a bottom rack permanently and you make four pizzas in a row.
Stainless steel plates sold as pizza steels. Austenitic stainless has roughly a third of carbon steel's thermal conductivity. It will not rust, and it will not perform like the plates on this page.
Any steel you cannot comfortably lift. The surface you actually use beats the surface that wins on paper, every time.
