A state table holds exactly the same information as a state diagram, laid out in rows and columns.
- Each row is a present state.
- There is a column of next states for each input value.
- Outputs: a moore machine has one output column (one output per row); a mealy machine writes next state / output in every cell.
Why bother when you have the diagram? A table makes it impossible to forget an arrow, because every row has a cell for every input. And it's the bridge to hardware: once the states are given codes (state encoding), the table becomes a truth table whose inputs are the present-state bits and the machine inputs, and whose outputs are the next-state bits. From there you write next-state equations with K-maps.
Counters use a simpler version, the next-state table: with no inputs except the clock, each present state has just one next state.
Tables also make it easy to spot equivalent states: rows with the same outputs and the same next states can be merged.