Here's the deal. The focal plane shutter on the camera has to mechanically open and then later close the shutter. The shutter takes time to move. Each camera has a maximum flash-sync speed. For your camera that speed is 1/200th. This is the amount of time it takes for the focal plane shutter to open completely.
It turns out the shutter has two moving halves referred to as "curtains". You'd think the two halves would meet in the middle, but this doesn't give an even exposure. Instead, they start at one edge. The first curtain opens... The second curtain just sits still. After the exposure is complete, the second curtain closes. This is how it works when the shutter speed is at or slower than the sync speed (1/200th in your case).
For faster exposure, the first curtain starts traveling across and the second curtain starts "chasing" the first curtain -- even before the first curtain can fully open. This means you effectively has a "slit" sweeping across the sensor to expose the image. Each point on the sensor will only have been exposed for the set time... but they technically don't all get exposed at the same precise moment. If the shutter speed is set to 1/1000th then the slit will expose only 1/5th of the sensor at a time.
And this is a problem for the flash. If the momentary burst of light occurs and any part of the sensor was blocked by the shutter curtain then you'll get an image with the exposed slit being bright, and the rest will be too dark. To compensate, better flashes have a "high speed sync" mode. This causes the flash to pulse repeatedly (VERY rapidly) so that each section of the sensor gets the same amount of light. The downside is that since the flash will have to fire repeatedly and very rapidly, it can't fire at full power (otherwise the capacitors would have time to recharge for the next burst.). The flash has to ration the power of each rapid pulse to make sure it can expose the whole sensor for an evenly eased shot.