You have a good spam filter installed, it is advertised as removing 95% of all spam, and yet your inbox is still full of the stuff. The software company must be lying when they say that it is 95% accurate, right? Wrong…
Let’s assume you get 10,000 emails in a year (sounds a lot but it’s less than 30 a day, you probably get a lot more).
About 95% of all email is spam (figures vary but it is somewhere around this number)
So of the 10,000 emails heading your way, 500 are genuine and 9,500 are spam.
Of the 500 genuine ones your spam filter will get it wrong 5% of the time which means 25 genuine emails will be marked as spam. The other 475 will end up in your inbox.
Of the 9,500 spam emails, 5% will be missed by the filter and these will end up in your inbox. 5% of 9,500 is 475.
So your inbox will have 475 genuine emails and 475 spam emails. 50% of each, even though your filter is 95% accurate.
The software company didn’t lie, there is just so much spam that a lot of it still gets through.
If spam rises to be 99% of all email then your spam filter will have to be 99% accurate just to keep the level of spam in your inbox at 50%.
Today’s takeaway: Be careful when you think about statistics, you can sometimes get some funny results.