Which three factors influences the optimizer's choice of an execution plan?

Cardinality estimates - the expected number of rows returned at each step - are a core input to the CBO cost formula, directly determining which join methods and access paths are selected.

The OPTIMIZER_MODE parameter (ALL_ROWS, FIRST_ROWS_n) sets the optimization goal and directly influences whether the optimizer favors throughput or response time when comparing candidate plan costs.

Fixed SQL plan baselines mandate the use of a specific plan rather than influencing the optimizer's cost-based selection; they bypass the optimizer's choice rather than serving as a factor in it.

The type of database connection (dedicated server, shared server, etc.) is a session configuration detail and has no effect on the optimizer's cost calculations or plan selection.

Object statistics stored in the data dictionary (row counts, column cardinality, histograms, index clustering factor) are the primary data the CBO uses to estimate the cost of each candidate execution plan.