Notatypewriter's Blog

Umm… what?

Posts Tagged ‘statistics

Twitter Spam Statistics #fb

leave a comment »

Twitter published a blog post about how spam has been reduced in the system. While I personally haven’t seen much spam tweets so I can guess that their anti-spam measures are working, I think their blog post is slightly misleading in how it uses the power of numbers.

Their main piece of evidence is that the percentage of spam tweets per day has gone down from a high of 11% in August 2009 to about 1.5% in February 2010. This piece of data would allow us to conclude that the chances of seeing spam are way lower.

However, percentages do a great job of obfuscating information. By their very nature, percentages are ratios of one number to another. When a percentage changes, one has no way of determining which number in the ratio actually changes.

In this case, we have no way of knowing whether legitimate tweets went up or spam tweets went down. Given the increasing popularity of Twitter, I’d bet on legitimate tweets increasing, at least at a faster rate than spam. I guess that’s a good thing for one of my favorite social platforms.

(Not to mention, we don’t know their methodology of distinguishing a spam tweet from a legitimate tweet.)

Written by notatypewriter

2010 March 23 at 7:58 pm

Posted in Nerding out

Tagged with ,

%d bloggers like this: