If we have stats for all of the files being read, we can respond to: COUNT(*) MIN(col) MAX(col) from just the stats