SQL NULL Handling Practice — 16 Challenges (15 Free)

Filtering on absence — why = NULL never matches: 9 challenges (8 free)

Comparing to NULL with = does not return false, it returns unknown, and WHERE keeps only rows that are true. That is why WHERE manager_id = NULL returns nothing at all rather than the rows you meant — you need IS NULL. The same rule is what makes NOT IN against a nullable column return zero rows, the trap the NOT IN trap walks through.

Counted from the challenge bank, September 2026. Easy first, then Medium, then Hard.

Supplying a default — COALESCE, IFNULL and NULLIF: 8 challenges (8 free)

COALESCE(col, 0) returns the first argument that is not NULL, which is how a missing number becomes a zero in a report rather than a hole in a chart. NULLIF(a, b) does the reverse and is the standard guard against divide-by-zero. The judgement is not the syntax, it is whether a default belongs there at all: a missing rating is not a rating of zero, and averaging it as one is a wrong answer that looks right.

Counted from the challenge bank, September 2026. Easy first, then Medium, then Hard.

Counting what is there — the aggregates that skip NULLs: 7 challenges (6 free)

COUNT(*) counts rows; COUNT(column) counts rows where that column is not NULL, and the gap between the two numbers is the missing data. AVG, SUM and MIN ignore NULLs entirely, so an average over a half-empty column is an average of the half that exists — usually right, occasionally the bug. These are the challenges where the answer depends on knowing which.

Counted from the challenge bank, September 2026. Easy first, then Medium, then Hard.

Where this shows up next

NULL rarely arrives on its own. It arrives through a LEFT JOIN, which is the machine that manufactures it; it survives into an aggregate that quietly drops it; and it breaks anti-joins written with NOT IN instead of NOT EXISTS. If you want the reading rather than the practice, IS NULL vs = NULL is the short version and five NULL mistakes the long one.