Starting Blocks tell the Data Miner where to start its analysis.  Every blank Data Miner opens with a single Starting Block in the top left corner.  

The Starting Block must be connected to the rest of the Data Miner flowchart to generate results.

You can add additional starting blocks as well.  Additional starting blocks allow for multiple streams of results to be produced within one Data Miner.  Right-click on a Starting Block to adjust its one parameter, its order.

There is no need to merge multiple starting blocks into the same stream.  Multiple starts need only be used to run separate streams in a certain order.  In the example below, Start #1 runs before Start #2.  Start #1's stream run successfully even though the left-most branch has no separate Starting Block. Once the Start #1 stream is done,  Start #2 runs its analysis and finishes the Data Miner.