If you animate it in N-1 frames, just one less frame!, it suddenly looks like it's happening much to quickly, without any regard for momentum.
Many is the burgeoning animator (myself included) who has fallen into despair trying to get around this seemingly impossible conundrum. As with everything else, the key to getting your movement to look good is actually a trick (or a principle, the difference between the two is marginal).
The basic problem is this: you want the movement to go -really- fast, so that it looks like it has a lot of momentum. However, if the whole movement goes really fast, the audience doesn't have time to realize what is going on, and so the movement looks incoherent, and is perceived as happening "too fast". If you make the movement last long enough so that the audience can tell what's happening, it doesn't look like it has momentum.
So, what you really want to do is to have the audience realize what is going to happen -before- it actually starts happening. That way, the super-fast movement can occur when they already know it's going to occur, and they can understand it and appreciate it at the same time.
This is the reasoning behind the principle of Ancitipation and Snap. These are two terms that are really part of the same principle: anticipation is the movement that keys the audience, and gets them ready to perceive a fast action. Snap is the actual fast action that they then perceive. The two terms don't really have any use apart from each other, in the same way that "Slow-In" doesn't make a lot of sense without "Slow-Out".
In fact, there's a third term, Overshoot, which is also tied up in this principle. It is the slower action which happens after the Snap, and gives the audience time to process what happened during the fast Snap action.
However, even knowing the principle of Anticipation and Snap, you're likely to use it incorrectly. Why? See the next page.