Mysql performance sending data -

i've got table looks this:

id (int, 9, unique key) --- item (int, 3) --- userid (int, 9) --- position (varchar, 12)

now want run folling query:

select count( * )  table position  'bag%' , userid = ".userid." 

the possible positions cases "bag0", "bag1", [...], "bag6"
, unique - means there no userid has 2 items in bag2, example.
but it's not unique - example 2 rows having "userid: 82378, positon: 'box1'" possible

if try in()-clause doesn't change result.

the "sending data"-time query 50ms.

do have ideas how modify table query works faster?

this query:

select count( * )  table position 'bag%' ,       userid = ".userid."; 

this pretty simple query. way can speed using index. suggest composite index of table(userid, position). mysql should optimize like because initial prefix constant. but, if want really, sure index gets used, can do:

select count( * )  table position in ('bag0', 'bag1', 'bag2', 'bag3', 'bag4', 'bag5', 'bag6') ,       userid = ".userid.";