Object statistics describe the data stored in the database. For example, they tell the query optimizer how many rows are stored in tables. Without this quantitative information, the query optimizer could never make right decisions, such as finding the right join method for small or large tables (result sets). To illustrate this, consider the following example. Let’s say I ask you what’s the fastest method of transportation for me to get home from a particular place. Would it be by car, by train, or by plane? Why not by bike? The point is, without considering my actual position and where my home is, you can’t arrive at a meaningful answer. Without object statistics, the query optimizer has the same problem. It simply can’t generate optimal execution plans.