By splitting them into multiple deploys it makes it easier to identify bad commits when an error starts. We can usually correlate the start of an error with a s

By splitting them into multiple deploys it makes it easier to identify bad commits when an error starts. We can usually correlate the start of an error with a specific deploy, and instead of digging through 50 commits to find the cause we only have to look at a few.