Error guessing
Error guessing is an experience-based test technique in which the tester anticipates where defects are likely to be, drawing on past failures, similar systems and common mistakes, and designs tests to provoke them. Typical guesses are empty input, division by zero, a name with an apostrophe, a date on 29 February or an upload of a zero-byte file.
The longer version
ISTQB classes error guessing as an experience-based technique, next to exploratory testing and checklist-based testing. No formal rule derives its tests, as equivalence partitioning does from input ranges. The tester asks where this kind of software usually breaks and goes there, so two testers produce different lists, and an experienced one produces the better of the two.
Guesses draw on several sources. Defect history is the richest, since a product that broke on time zones once will break on them again, and failures remembered from similar products are another. Some teams write the recurring ones down as a list of failure types, which turns private knowledge into something a newer tester can use. Others keep a standard set of awkward inputs: very long strings, emoji, leading spaces, negative quantities, a file with the wrong extension.
Error guessing is most useful after the formal techniques have done their work, as a quick pass for the awkward cases they do not generate, and under time pressure when there is no room to design a full set. Because nothing measures how complete a set of guesses is, it cannot show coverage on its own, and recording the guesses that found bugs keeps the technique from living only in one person's head.
Common questions
What is error guessing in software testing?
Error guessing is a technique where the tester uses experience and intuition to predict the defects most likely to exist and writes tests that target them.
Is error guessing the same as exploratory testing?
The two are separate techniques that ISTQB groups together as experience-based. Error guessing focuses on anticipating specific likely defects, while exploratory testing is the broader practice of learning the product and designing tests during the session.
What are examples of error guessing?
Typical guesses include submitting a form with every field empty, entering zero or a negative number where a quantity is expected, pasting a very long string, using a name with an apostrophe or accent, picking 29 February as a date, and uploading a file with no content.
Where do good error guesses come from?
Good guesses come from the product's own defect history, from failures seen in similar products, and from lists of common faults that a team has collected. Experience with a type of system teaches where it tends to break.
Can error guessing be measured?
Error guessing has no coverage measure, since nothing defines the full set of guesses. Teams track it by what it finds, counting the defects each guess uncovered and keeping the productive ones as a reusable list.
Bug reports that arrive ready to triage.
Collect replay, console, network and browser context with the Free plan.