I have a general query regarding table statistics in SQL Server.
Context:
In a project that I am working on, I have several databases that are being used - each contains tables with their respective indexes and constraints. So I was carrying out several test cases and I turned the SQL Server Profiler to create traces so that I can use them with the Tuning Advisor to generate recommendations. The only recommendations that the Advisor has generated for me were a bunch of statistics.
After creating these statistics, I tested out the stored procedures with statistics io set to on, and I noticed a small improvement on the logical reads.
Question:
My question is this: having inserted many statistics that the Advisor generated for me (about 3-4 per table), how do these statistics affect performance on INSERT, UPDATE or DELETE?