Alexa is fine, just you should not use it as your major measure.
All you have to do is compare the Alexa traffic of gambling911.com and casinocity.com to realise that Alexa data could be quite flawed 
When you factor in that Alexa shows on average 2 page views per visitor for gambling911 and 4 for casinocity – then according to Alexa, gambling911 gets 4 times more visitors than casinocity, which is quite impossible.
And at the same time, porn.com has Alexa of 300,000+ :inlove: