Learning Optimization Techniques

Hi,

Still using 2000; I'm trying to figure out where to even start with optimization in SQL queries. BOL is very confusing to me on this point. I ran the following query with and without and index as seen below. Is Trace a good place to start?

DROP INDEX OrdDetails.OrderID_ind
GO

CREATE INDEX OrderID_ind
ON OrdDetails (OrderID)
GO

select count(*) from OrdDetails = 6467155 rows
GO

-- query; without index Duration 156; CPU 95; Reads 54
-- query; with index Duration 66; CPU 15; Reads 97

select OrderID, Count(OrderID) AS OrderIDCnt from OrdDetails
group by OrderID
having count(*) > 1
GO

Why are the reads more with the index? This is just a simple non-clustered index, which I believe is the default. Should I be starting elsewhere to find out this info?

thx.

Kat