Okay...
To dumb it down even further:
"In
statistics,
dependence or
association is any statistical relationship, whether
causal or not, between two
random variables or
bivariate data. In the broadest sense
correlation is any statistical association, though in common usage it most often refers to how close two variables are to having a
linear relationship with each other." - Wikipedia
What correlation refers to in this case is a relationship between the two variables of points in junior and points in the NHL.
The correlation is obviously not 1:1, since some players put up big numbers in the CHL but never match their expected production in the NHL (ie: Daigle), and some put up modest numbers in the CHL and far exceed it in the NHL (ie: Lucic). As a generalization though, point production in lower leagues is one of the better tools that we have to predict NHL point production.
What this means is we cannot say with absolute certainty:
"because a player put up 100 points in junior and is successful at the nhl level another player putting up those same points in junior will be successful as well"
This is not a helpful statement to make because we are not attempting to trade in absolute certainties.
But we can say: (because correlation is a thing)
"because a player put up 100 points in junior and is successful at the nhl level another player putting up those same points in junior will
likely be successful as well"
And that is exactly what I am saying.