Maya, Chris Drayer and Gahlord Dewald in San Francisco

I love a bit of fun as much as the next person.  I have a twilk background on my Twitter page and it is fun to see it change over time as it looks at those I have interacted with the most in my Twitter stream.  I like that Twilk digs deep and goes back a bit or so it appears.  Then there is the new “Who are your Twitter BFF’s” craze going on.

I did it.  I won’t tweet it, but I did it because I was curious about how it calculated it.  I know who I interact with the most, it got two of those people right – but only by coincidence.  Based on the results I got when I did it I am fairly sure that the BFF program only looks at the last day, maybe two, of interactions and bases it on that.  That doesn’t work at all does it?  My list included three people I had only had one conversation with – ever.

I find it amusing and am glad so many people are finding the entertainment value – but maybe when you are doing that instead you could promote a cause like to “Help Bring Noor and Ramsay Home”  I think that might be a much better use of your time.  That link could do a lot of good.

If the “BFF” folks really want to do something that might even show a real result, or pretend to – they should consider looking at a one week sample of your tweet stream or better yet a percentage of your total Tweets.  Only looking at the very recent and only a small sampling doesn’t really prove anything or is accurate. Well, at least for me – but maybe it is right for your BFFs?  Do I have too many “friends”?

Were your results accurate?