[Omin]

The fixed ratio doesn't scale all that well, because some of the water is absorbed by the rice and some just evaporates. Even with a lid on water still evaporates. The evaporation portion depends on the type of pot, how long you cook and how much heat you put in.

So, if you only vary amount of rice, you really need some_factor * amount_of_rice + constant_amount and the knuckle rule has that characteristic.

[iainmerrick]

Huh, OK!

I'm surprised that there would be significant evaporation loss in a pot with a close-fitting lid for 20 minutes on low heat. If you heated the water on its own I wouldn't expect a large amount to boil off... Although now I say that, I'm not so sure, maybe I should run that experiment.

I would have guessed the imprecision of 'knuckle depth' being uncalibrated with the volume of rice and the size of the pot would be greater than the imprecision of a pure volume ratio due to evaporation, but I can see that the knuckle depth thing might make sense if you're always cooking roughly the same amount of rice in roughly the same size of pot.

I've had good success cooking rice by the 2:1 ratio in both large and small amounts in various pots, so I'm still happy to vouch for that method.

[Omin]

Another aspect is that there is a large span in the amount of water you can let the rice absorb. A 1:1 ratio of water to rice (not accounting for evaporation) is all you need to cook rice no matter if it's short grain or long grain, but a lot of rice will also happily absorb more, like 2x its weight, if you give it enough water and time. It sounds like you are using long grain rice given the large cooking time and high water ratio.

[busssard]

i have cooked all different kinds of rice with this method, except for risotto rice all have measured up to this ratio

[blooalien]

I have also had great success with this method ever since first being told about it. Didn't believe it could be that easy until I was shown for the first time, and sure'nuff it seems to work pretty nearly perfect every single time.