MySQL - left join or nested select for filtering non-existing rows? -

table structure catalog_product_category_bindings:

`productid` integer unsigned not null `categoryid` integer unsigned not null 

table structure catalog_products:

`id` integer unsigned not null auto_increment misc unrelated columns 

task: data of products not bound category (no entry in catalog_product_category_bindings).

query #1 (using left join):

select cp.* catalog_products cp left join catalog_product_category_bindings cpcb on cp.id = cpcb.productid cpcb.categoryid null 

query #2 (using nested select):

select cp.* catalog_products cp id not in (select productid catalog_product_category_bindings) 

both queries seem quite similar in terms of speed on tables (i don't have in there), believe second 1 worse in performance since loops on every id in catalog_products table , compares every productid catalog_product_category_bindings. not mention might not return , break query altogether (though happen if table truncated).

which better? i, personally, prefer #1 since fits better query builder , seems better.

since columns defined not null mysql optimizes not in() query anyway not exists. there's not difference between two.

here's must read topic.