I was hoping you could explain
A> Definition of a 'phase',
B> purpose of a phase and
C> variances between phases...
As your generator runs, each coil produces a voltage that varies cyclically, repeating the same pattern over and over. One ideal case would be a coil spinning in a uniform field, in which case the voltage pattern would be a sine wave, and real generators tend to approximate that. But even if the pattern is some odd shape it still repeats identically.
The "phase" is how far along this cycle (from some arbitrarily-chosen reference point) the waveform is at this moment.
If you have multiple coils in your genny, spaced such that they encounter the magnets at different times, they will produce identical waveforms, but will be at different phases along the waveforms at any given instant. In that ideal generator case it would be like having several coils rotating, but at different angles to each other (like the arms of an asterisk).
In such a situation you pick one of the coils as a moving reference point and refer to the "phase" of each coil as the fraction of a cycle it is ahead of or behind the reference. You express this fraction in degrees, as a fraction of a circle. (Back in that imaginary perfect two-pole generator this would actually be the angle between the coils.)
The number of phases you have is the number of DISTINCT phase angles you get from your various coils (including the zero-degree phase angle of the one you picked for a reference). Note that in a multi-magnet, multi-coil genny some of your coils may be producing identical waveforms, so the number of distinct phases may be much smaller than the number of coils.
In a polyphase alternator the phases are typically evenly distributed "around the circle". So a three-phase alternator would have x=zero degrees, y=120 degrees, z=240 degrees.
Any single phase produces a varying output (and a varying drag on the shaft) during its cycle. But with multiple phases these varying outputs and drags occur at different phases of the cycle, so they add up to something much more smooth, and when one phase is producing little or no power another will be near its peak. (In fact, two or more phases with a resistive load can be made to add up to a continuous load on the shaft that doesn't vary with angle.)
Does that help?