# Multi-Armed Bandit Testing with Epsilon-Greedy Variant Selection

DevFeed: [Multi-Armed Bandit Testing with Epsilon-Greedy Variant Selection](<https://devfeed.tech/articles/big-wins-multi-armed-bandit-testing-32145.md>)

Original publisher: [Read original article](<https://adambard.com/blog/multi-armed-bandit-testing/>)

Published: 2013-05-22T00:00:00Z

Content type: opinion

Language: en

Sources: [Adam Bard](<https://devfeed.tech/sources/adam-bard.md>)

Topics: [A/B Testing](<https://devfeed.tech/topics/a-b-testing.md>), [Testing](<https://devfeed.tech/topics/testing.md>), [Algorithm](<https://devfeed.tech/topics/algorithm.md>), [Clojure](<https://devfeed.tech/topics/clojure.md>)

Tags: [a-b-testing](<https://devfeed.tech/tags/a-b-testing.md>), [algorithm](<https://devfeed.tech/tags/algorithm.md>), [clojure](<https://devfeed.tech/tags/clojure.md>), [conversion](<https://devfeed.tech/tags/conversion.md>), [experiment](<https://devfeed.tech/tags/experiment.md>), [library](<https://devfeed.tech/tags/library.md>), [statistical-significance](<https://devfeed.tech/tags/statistical-significance.md>), [testing](<https://devfeed.tech/tags/testing.md>)

## AI overview

The article explains multi-armed bandit testing as an adaptive alternative to standard A/B testing. It describes an epsilon-greedy strategy that shows a random variant part of the time and the best-performing variant the rest of the time, balancing conversion optimization with continued experimentation. It also mentions the author's Clojure library, bandito.

## Source excerpt

I'm all about big wins. I could care less about incremental optimizations if there's a major victory to be had instead. That's why I'm way into multi-armed bandit testing right now.