When fermions are interdependent within a very small volume, we call them 'composite particles'. Their bosons are confined: emitted and re-absorbed within the structure of the composite. Such composites are detectable by their residual or vacuum interactions.
Baryons feature an ongoing sequence of three nominal quark states; a QCD singlet {A, B, C}, fig.4.
Although all interactions are fundamentally identical, there are three notable modes of interaction:
Each quark has a colour and an anti-colour, both before, and after their fermion event. e.g. the gluon constituents of quark A are:
In terms of path lengths, red + green = yellow, i.e. quark A links to quark C in two ways:
Quarks B and C have similar constructions, and we can find other equivalent formulations of the gluon octet, including flux tubes for added concurrency and complexity.
Supposing there were no vacuum interactions in the anti-colours, then red + green would be the same length as a one-step yellow without vacuum (imposing a triangle shape, flattened into a straight line). The geometric consequence is that the hadron would travel at the speed of light, with all positional solutions on the same directional vector.
Conversely, with vacuum interactions, the anti-colours (yellow, magenta, cyan) have opportunity to split their journey into 'away' and 'return' segments, allowing its length to equal the corresponding pairs of colours (two from: red, green, or blue), and therefore enabling the baryon to have an aggregated velocity of between zero and light-speed (noting that a full positional cycle for the baryon spans two singlet cycles). This also allows change in the classical directional velocity vector, vacuum interaction, boson substitution, and charge currents, all dependent on the ratio of red+green : yellow (and likewise for the other combinations).
Fig.4 is simplified, and there are other variations that result in working QCD singlets. If, for example, A, B, C, or the vacuum events, were shifted a little, symmetry breaking would occur as shown in fig.3.