I’ve been critical of Kevin Rose and Digg in the past, and as Muhammad Saleem points out — its been almost half a year since many promised features would be rolled out. Well, as you probably heard from all your favourite bloggers, many of them are being rolled out as we speak, and damnation, its about frackin’ time.

There are quite a few of them that are worth noticing including how Digg is getting all socialized up, giving it the appearance of a social network — and it kind of is. There’s a more robust identity amongst Diggers, and you have the opportunity to message people within the Digging community as well as send out private “shout” outs, which is pretty much like sending out links.

Although most of the hoopla seems to be around the whole “Digg is becoming Facebook” meme (which to a tech blogger is deliciously irresistible — Facebook, check, Digg, check, someone copying someone else? check), I think the best and most important feature is one that’s kind of hidden in blog/press release.

And that’s actually the “recommendation” feature, which allows Digg to recommend *both* friends *and* stories that you might like based on your submission and voting patterns in the past. I think this feature more than any other will create a bigger and better social community as Digg is using the technology to help its community in the most meaningful way — to help people find other people they didn’t already know *or* other news they didn’t already read (hopefully) that will be helpful to them.

One other thing.

I think that the recommendation engine, as Kevin blithely alludes to by describing that it requires “dozens of servers to crunch the math”, gives you an idea of the kind of stuff that goes behind the scenes at Digg — and what they’re doing with the data. There are probably a bajillion ways to measure what’s actually going on at Digg, and I suspect that KRo et al are mining the hell out of it *all* to do all kinds of things. We know they’re already crunching numbers to find associations based on existing behavioural patterns. Who knows what else they’ll be able to do — or are doing already.

Let’s get concrete for a second. Does that make it invulnerable to gaming? I don’t know if “invulnerable” is the right word, but clearly no matter what we think we know about Digg’s algorithm, I think there’s probably a whole lot we *don’t* know about what’s being measured, and how those variables interact. This probably affects everything from deciding when or how something is buried to measuring the trajectory of how popular stories are getting.

Its something to think about the next time another Digg controversy comes up, I think.

Update: Mashable points out that some/most/a vocal minority don’t like the social features at all.