Basic because it is not even queued. It is synchronous I/O, read/write is performed in sequence with user CPU insturctions, no read-ahead, no overlap queuing for write later. It is not even buffered, reading one record at a time.
Direct addressing for speed, using track and record count. It needs to be allocated via ISPF using a special parameter. Unless you have inherited a BDAM file and legacy code, no point using BDAM.