How is "Most Concurrent Users" generated?

[djross]

x

[PeaceAnarchy]

I don't have these features, but my guess would be the percentage of films in common(relative to your total I would assume).

[djross]

x

[mpowell]

PeaceAnarchy is mostly right, but it's not just the percentage of films in common relative to YOUR total, but relative to the other user's total as well.

For Example:
UserX has seen 78% of the films you've seen.
UserY has seen 78% of the films you've seen.
You've seen 10% of the films UserX has seen (because he's seen many, many more than you).
You've seen 80% of the films UserY has seen.

The "concurrency" is basically calculated as an average... UserX = 44%, UserY = 79%.

We plan on doing a review of how accurate this is later, once we get a few more people using the special new features. But does that make sense?

[djross]

x

[mpowell]

It's definitely different than the TCIs, because it doesn't take the ranking itself into account. So we can't really use this statistic to predict a score.

But as a way to identify the users who tend to watch the same movies as you, it's pretty accurate. And it helps identify those films, good or bad, which people with highly concurrent lists have seen, and which you might be missing.