Conservative estimates generally include the time required to take down and set up the next game, as well as any bonuses that were reached for each game. I generally figure that 75% of the estimate is a good approximation of how long the game itself should take (For shorter games, not true for ~3h+ game).