Is entropy a real change in disorder? This is what I read from the famous book 'Concepts of Physics' by HC Verma :

A block moving at a speed v on a rough table eventually stops and the table and the block warm up. The kinetic energy of the block appears as the internal energy of the table and the block. Can the reverse process be possible? That is, we heat the block and the table and put the block on the table. Can the bodies cool down and the block start sliding with speed v on the table converting the internal energy to kinetic energy?

Well, obviously, according to what we've all learnt and seen, the answer is no.
But isn't it right to think in the following way?
Induced motion of the block as given in the hypothetical situation can cause an increase in disorder (as motion is disorder) and cause positive entropy, which is spontaneity? In that case, it should be possible. Now that we know physically that it is not possible, isn't this a violation of the fact that entropy of the universe is always increasing?
 A: If the block starts sliding with velocity v, then this motion is mighly organized, since all construents of the body move with the same velocity $v$.
By contrast, the internal energy gained by the table when the block stops is highly disorganized motion, since the construents of the table move with totally uncorrelated velocities.
If the table would cool and the block would start sliding, it would imply that the particles of the table all suddenly aligned their velocities, which is highly improbably.
With that said, I think the best way to understand entropy is through statistics, and entropy is a measure of disorder in the following intuitive sense:
Entropy measures how much information would you gain if you suddenly knew the instantaneous microstate of the system. In the case the system has only one allowed state, you gain no new informaton (entropy = 0), if the system has infinite amount of possible states, then you gain infinite information (entropy = $\infty$).
This can be directly applied to the case of your block. Assume the characteristic particle number is $N\approx 6\times 10^{23}$. When the block stops and the particles of the table move in an uncorrelated fashion, to specify the microstate of the system you need to specify $3N$ velocity components.
On the other hand, when the block moves with velocity $v$ (and we view the particles of the table as being at rest), then only 3 numbers are needed to specify the microstate (the 3 components of $v$). $3<<<<<<<3N$ so entropy is much lower in this case.
Obviously this example is highly exaggerated, the table construents have disorganized movement even in the case the block moves with velocity $v$, but if we look at only the distribution of energy that used to be contained in the block's motion early on, this should still be somewhat accurate.
