I think it may look a lot like luck because they make so many speculative 'bets' and so many don't pan out, but that's more a function of the vast sums the successful ones end up being given to invest combined with poor deal flow. But they still aren't random odds.
Simple example: the VC that invested into WhatsApp and got a massive payout when Facebook bought them. Nobody in the Valley VC community knew about WhatsApp because it had no penetration in the USA. But the firm had hired a developer intern to write a script that downloaded the app store listings for every country in the world individually and calculate fast risers. So this VC was the first to realise that WhatsApp was getting huge everywhere except where they actually lived. But then he visited the WhatsApp website and discovered there was no address or contact information anywhere. He did some research and discovered they were based in Mountain View, but literally couldn't find an actual address anywhere.
So he did the obvious thing - he walked the streets of Mountain View systematically looking at every single office building until he found them, tailgated his way in and then sweet talked the receptionist into giving him a meeting with the CEO.
Well the CEO didn't want to take VC investment. So then the guy started on the final stage: convincing him it was a good idea. Eventually the WhatsApp founders concurred and took the money.
Those sorts of actions aren't the actions of gamblers. The guy had a thesis, he worked to put it into action, he even did a lot of physical work and ended up getting a distinctly non-random reward.