I consider something to be valid evidence if:
- It is not based on a logical fallacy
- It can be tested and has withstood that testing.
With regards to the Kalam cosmological argument, it is built on an assumption (that whatever begins to exist must have a cause) when this is just an argument from incredulity. "I can't comprehend how something could begin to exist without having a cause, therefore it doesn't happen." In any case, this claim really only can be said to apply in cases where our current laws of nature apply, and there's no evidence that this was the case at the Big Bang. In fact, what we understand of quantum level events indicates that our current laws of nature break down at that kind of scale. The argument also assumes that the universe began to exist, when the only thing we know is that it began to exist in its current form. The universe could have existed in some different state prior to the Big Bang (in whatever sense the word "prior" can be said to apply), and so the argument may not even need to apply. So, the Kalam cosmological argument fails on point one. It also fails on point two, since it is inherently untestable.