well, there is always SOME friction, just very low amounts on oil and/or ice.
On ice (I'm assuming glass-smooth) since the amount of surface of your ball that would touch the ice is very small, the amount of friction created would be miniscule so. . . likely it would skate for . . . forever.
Friction will increase with a decrease in speed, yes.
Friction is increasing due to multiple factors. Speed is one. Change in lane surface is another.
The ball's internal properties do NOT impact friction (EXCEPT for ONE thing - making the ball flare which brings "fresh" surface to bear) - friction is the amount of resistance to movement between two surfaces. That's it. The equation for it is F=F(n)*u where u (actually micro, but I don't know how to get that on here) is the coefficient of friction, and F(n) is the normal force (mass * gravity, basically)
In reality, it's NOT a simple math equation. There is a coefficient of friction which is basically the "stiction" between the two surfaces - which in our case is impacted by the oil (how viscous it is, how thick it is), the lane surface (how hard and smooth it is), the ball's surface (grit, whether it gets coated with oil itself, hardness) etc. There is gravity (which we can use as a standard) and the weight of the ball - which gets us the "normal" force.
The axis and tilt of rotation only gets us the VECTOR that the ball wants to travel. They impact what the ball does while sliding (the amount of flare) and what happens once the friction between ball and lane increases to the point that the ball's rotation vector overcomes the imparted vector (the slide direction) - how gradual the directional change is and how much of a directional change it is.