This metric measures the average number of other files changed when a file itself is changed. Files that cause a lot of other files to be changed may be too tightly bound to those other files.

It is important to note that this metric is an average for all commits of a file. When picking aggregation method to examine a directory of files in the source tree, this still applies. This means that if you use the 'max' aggregation method on a directory you will be presented with the metric for the file with the highest average number of co-commits across the files entire commit history.

Files with a high value for this metric may have a widely used public API. Special care should therefore be taken when updating it to ensure any other parts of the program that use the API are also updated.

It is also possible that the file is too tightly coupled with other files. Examine the afferent and efferent coupling metrics as well and consider the advice given.

A final possibility is that committers could be using single commits to introduce several unrelated changes. If this is the case then commit practices should be reviewed.