We really need a further correlation between people who pass the interviews and job performance a year later. I do a lot of interviewing at my current job and w

We really need a further correlation between people who pass the interviews and job performance a year later. I do a lot of interviewing at my current job and we have found no strong correlation at all between CS skills and actual ability to "get things done". We toned down the CS type questions since they tend to take too long. We still ask a few basic tree and string manipulation questions to weed out the people who have no idea how to program and get insight into how the person thinks. I still feel at the end of the day we could flip a coin on accepting an interview candidate once they have shown basic competency and have the same results. I have been telling candidates that a public github repo with a nice commit history carries much more weight with me then a CS degree since we have been burned so many times before.