An empirical investigation into merge conflicts and their effect on software quality

Merge conflicts are known to cause extra effort for developers, but little is known about their effect on software. While some research has been done, many questions remain. To better understand merge conflicts and their impact we performed an empirical study about the types, frequency, and impact of merge conflicts, where impact is measured in terms of bug fixing commits associated with conflicts. We analyzed 143 open source projects and found that almost 1 in 5 merges cause conflicts. In 75.23% of these cases, a developer needed to reflect on the program logic to resolve it. We also found that the code associated with a merge conflict is twice as likely to have a bug. When the code associated with merge conflicts require manual intervention, the code is 26× more likely to have a bug.