What is the exact reason quark-gluon plasma behaved like liquid instead of gas as predicted? I have searched all over the net (including here) but most references and answer  were either old like written in 2006 (or around 2010) or they didn't directly address it why the QGP behaved like liquid. Anyone know what is the exact reason or latest theory why quark-gluon plasma behave like liquid instead of gas as predicted before the first RHIC experiment? Thank you.
 A: The difference between a liquid and an interacting gas is a question of coupling strength. If the interaction strength between particles is weak, then they'll be essentially free particles, and so the bulk will behave more like a gas. However, if the interaction between particles is strong, then their motion will be much more restricted, and the bulk will behave much more like a liquid. In the case of quark-gluon plasma, the matter created in heavy-ion collisions is currently described very well by a theory that's designed for use in situations where the interactions between particles are very strong. It makes use of a theoretical tool called the AdS-CFT correspondence to describe the evolution of the plasma in terms of a 5-dimensional black hole, and predicts its viscosity to within experimental uncertainty.
As to why it was originally thought that QGP would be more like an interacting gas, it was essentially a reasonable extrapolation based on what we knew about quantum chromodynamics (QCD). QCD exhibits "asymptotic freedom": it gets weaker at higher and higher energies. Considering that at RHIC and the LHC, the quarks and glouns inside each nucleus are imbued with hundreds of times their ordinary energy, it made sense to assume that their interactions would be much weaker. But the nucleus (and non-perturbative QCD in general) is an extremely complicated system to study, and is therefore not well understood, so in a way it's not entirely surprising we were wrong.
For a more in-depth, but still accessible review, I recommend this one, written very recently: https://arxiv.org/abs/1802.04801
