What is HDR?

I see it mentioned around here but I haven't the faintest idea what it is?

In simple layman terms: High Dynamic Range. The ability to take a series of different exposures of the same scene and combining them in post processing so basically all the image is properly exposed.

yeah, basically most camera's are not capable of recording the entire dynaimc range of light in that exists in the real world, capturing only about 5-7stops in good quality detail. HDR's are created my taking multiple LDR shots exposed for the highlights, mid-tones and shadows seperately and then merged into a single 32-bit image which has a dynamic range up to 14-15 stops.

Nearly all 32-bit images are impossible to view on a monitor or print out so you need to compress the tonal range down to 8 or 16bit by tone mapping.

HDR - High Dynamic Range, 32bit image produced by merging multiple images, then tone-mapping down to 8/16bit.

IDR - Increased Dynamic Range, produced by blending 2 or more images to extend DR

LDR - Low Dynamic Range, single shot images from any camera.

Have a look here: http://www.hdrsoft.com​/

Is one RAW image enough to produce the 3 different exposures or should you actually shoot 3 separate exposures? Is that even possible with action photography?

^You can take a single exposure and make it into 3 exposures using Adobe Camera Raw.

2) Open the RAW file in PS and adjust the exposure to -1 or -2, depending on how much dynamic range you need to capture. Save as a tiff file.
3) Open the RAW file in PS again, adjust the exposure to +1 or +2, and save as a tiff.
4) Open the RAW file in PS a third time, adjust the exposure to 0, and save as a tiff.
5) Open all three tiffs in Photomatix.
6) Tonemap the files in Photomatix to suit your tastes.

That strategy is best if you don't have a tripod (or steady enough hands for hand-held HDR) and/or are photographing moving subjects. It's best to take three separate exposures, if possible.

HDR is by definition the extension of the dynamic range portrayed in the final image beyond what the camera can capture in one shot. The camera has a 11 stop range but your final image has detail from elements in the original scene that were 18 stops apart. One RAW from that camera will have 11 stops worth of data, no matter what you do to it. You are not God nor is your computer, you can't create new data from nothing. However, blending multiple conversions from a single RAW into what is called a "fake HDR" can be an excellent way to get the maximum milage out of the data you do have. Another way is to convert to a single 16 bit tif, to retain almost all the data, and work on it with something like PS's Highlight/Shadow tool. But the three conversion method does have a certain advantage in that you are pushing and pulling linear RAW data before it gets bent out of shape by the gamma correction of the conversion.