I believe that is an urban myth. Has there been a controlled study showing this? The cooling & freezing rate has a lot to do with the volume of water and the container, and possibly impurities in the water.
This actually works. I read about it many years ago and had to try if for myself. Try it in your freezer, with a bowl of hot water versus a cup of cold water. The hot water will actually freeze first. However, there will be less ice in the hot water cup than the cold water bowl. As I understand, the hot water evaporates a much faster rate and thus transfers out its heat at much faster rate, also you end up with less water to freeze in the hot bowl, so it freezes faster. If you use shallow containers with a lot of open area, the effect becomes more pronounced.
Now if you cover both cups with some plastic, the water can't evaporate, and the cold water will freeze faster.
Interesting website. From what it says the closed containers can also exibit this effect. While I don't remember that happening in my very uncontrolled experiment, that makes the whole phenomenon even stranger.
It is partially a myth. It has everything to do with the observed rate of heat loss, the ambient tempurature and the characteristics of the water during the period of heat loss, nucleation and freezing.
