There are a couple ways that people define bonding and antibonding orbitals. One of the simplest is to just look at the energy of the MOs formed compared to the AOs they were formed from: if the MO is lower in energy, it is bonding and if the MO is higher in energy, it is antibonding.
Like I said, this way of doing thing is relatively simplistic, but you will see it pretty commonly in textbooks on the matter. The method of course isn't perfect, in particular because it essentially assumes that all local bonds are affected equally by a particular orbital.
We can see this is not the case by looking at the $\pi_2$ orbital you show. While the orbital is listed as bonding, the interactions of the p-orbitals where a positive lobe is next to a negative love will locally decrease the bonding between those two atoms (ie the MO is antibonding with respect to those local bonds).
So, for a simple first approximation, you can use the energy of the MO compared to the AO, but for a more complete sense, you should look at the symmetry of the MO and see what effect it has on a given local bond.
In terms of the ordering of the MOs themselves, assuming you don't have experimental way to determine the ordering like ionization energies, you can heuristically use an approach like you suggest where you compare the number of nodes and less nodes will mean lower in energy, more nodes will be higher.
In terms of comparing to the original AOs, it can be tricky without experimental data, however there are again general qualitative ways of ordering the MOs with respect to their original AOs. For example, if you have 2 MOs formed from two overlapping AOs, one will have to be higher in energy and the other will have to be lower.
When you get more orbitals overlapping, the arrangements can vary, but the sum of the energies has to remain the same. So, for example, if you have 3 overlapping AOs, you could form 2 bonding and 1 antibonding MOs or vis versa so long as their combined energy is the same as before.
Beyond this, you really just have to read up on MO theory to get a sense of how qualitative energy ordering is done. There is a lot more to it than what I've written (or could write) and it's probably better to build your understanding of the theory from the ground up rather than getting it piecemeal.
