Penrose's calculations show that our Universe is very special, so special that it could not have come into existence as a product of chance. The problem this causes for atheistic scientists, doesn't go away by saying that the calculations are useless, wrong or nonsense; in fact they are willing to invent a "multiverse", consisting of an infinity of universes, in order to make our Universe probable. That they are violating Ockham's razor in the most extreme way ever witnessed, doesn't bother them, apparently, as long as their invention can explain away the anthropic principle.
They do not explain away the anthropic principle. They
invented the anthropic principle. As for Ockham's razor, some will deem it a simpler theory that each and every universe has an equal and finite chance of existing and indeed do, than to imagine that a particular universe is favourable because it can develop life (which raises theoretical questions about why not
any of the other universes that can develop life?). Not that I am in favour of the anthropic principle, but I hope you'll be aware that there
is a reason people believe it.
It's like looking at winning the lottery. Let's say there are a million numbers for the lottery. You wake up one day and
shock! AWE! your neighbor's gardener has just bought the winning number and earned US$3.3 trillion! What are his chances? Literally one in a million. So which conclusion do you derive?
a) The lottery company knows him and chose his number. Otherwise, how could someone with a one in a million chance possibly win it?
b) Summing up the probabilities over all the other 999,999 people who bought the lottery, there is a total and exact probability of 1 that
someone would get it. It is only when you indiscriminately shrink the sample space that you think it's impossible.
You might say that one in a million is still big compared to the chances for our universe forming and supporting life. Well, firstly, the principle above is qualitative and the actual number doesn't matter. But supposing you still aren't convinced ... there's always the sample set of all possible chess games. I haven't worked on it but I'm sure there are a lot more than a million possible chess games. And out of those, how many have back-rank mates? I'm sure a negligible portion - yet among amateur players back-rank mates happen all the time.