Unless the changes are a critical bug fix, deploying on Friday is an asymmetric bet with high risk and low reward. That's why people prefer not to do it. Even

Unless the changes are a critical bug fix, deploying on Friday is an asymmetric bet with high risk and low reward. That's why people prefer not to do it. Even if the chances something will go wrong are very slim and you have lots of really great tests, they are still non-zero. Lots of unexpected things can go wrong in modern interconnected systems with lots of network dependencies. Not deploying on Fridays is not a matter of confidence. It's recognition that the downside is huge (losing your weekend) and the payoff is small for the vast majority of changes you could deploy. That's it.