Your definition of a "Tight" vs "Loose" game

Sklansky in SSH makes the distinction of "Tight" games as those with 3-5 players average on teh flop, vs "Loose" games with 6-8 on average.

At 2/4 Limit, I have yet to encounter a loose game under those defintions. Specifically, it's rare to consistently see 6-8 callers, and I therefore find these definitions not applicable to changing my play preflop.

However, using Pokertracker and Gametime, I certainly have identified looser games in terms of statistical critieria, most readily VP$IP. I usually identify a looser game as one with VP$IP > 20%, and anything above 30% is a goldmine.

My question is, how do others define a "loose" vs "tight" table using Pokertracker/GameTime statistics, rather than simply how many callers there are? And should one therefore change starting hand play accordingly using the Sklansky model?

Thanks!