Using EXISTS (or NOT EXISTS) instead of COUNT() is both a well-know technique among trained DBAs and people with relational database experience, and something I rarely see developers use. I often rewrite existence or duplicate checking queries to eliminate the COUNT() and replace with EXISTS.
The author uses SELECT 1 in the example. The usual idiom is SELECT *, and query planners recognize EXISTS (SELECT * ...) and don't actually select any columns.
The author uses SELECT 1 in the example. The usual idiom is SELECT *, and query planners recognize EXISTS (SELECT * ...) and don't actually select any columns.