Dark energy is parameterized by the time evolution of its equation of state $w(z)$. For a very wide class of quintessence (and phantom) dark energy models, we parameterize $w(z)$ with physical quantities related to the scalar field potential and initial conditions. Using a set of updated observational data including supernova, CMB, galaxy power spectrum, weak lensing and Lyman-${\alpha}$ forest, we run Markov Chain Monte Carlo calculations to determine the likelihood of cosmological parameters including the new dynamical parameters. The best fit model is centered around the cosmological constant (flat potential), while many popular scalar field models are excluded at different levels.