3 ms·
To compare, I looked up the estimates[1][2] on Facebook and if we use 170 million out of 1.4 billion, that's ~11.7%. That is in the ballpark of Twitter's 15% (
by jasode 10y ago
To compare, I looked up the estimates[1][2] on Facebook and if we use 170 million out of 1.4 billion, that's ~11.7%. That is in the ballpark of Twitter's 15% (48m/319m).
I guess if you make a popular service that doesn't require a credit-card payment or phone SMS text as a test for human verification, you should expect >10% of fake accounts even with state-of-the-art heuristics detection.
[1] not sure why Huffington Post is top google link about it but the other news sites say the same 170M: http://www.huffingtonpost.com/james-parsons/facebooks-war-continues-against-fake-profiles-and-bots_b_6914282.html http://www.huffingtonpost.com/james-parsons/facebooks-war-co...
[2] 2015 1Q user count: https://www.statista.com/statistics/264810/number-of-monthly-active-facebook-users-worldwide/ https://www.statista.com/statistics/264810/number-of-monthly...
- blackice 10y agoYou can look for proxy / hosting IPs. These bots are most likely running on servers.
- ben_jones 10y agoSo wouldn't it be relatively simple to block AWS/Azure/etc ip blocks AND duplicate accounts from single addresses? Wouldn't that be a large drop in bots right there? Corollary: I once played a large MMO that intentionally inflated user numbers right before they were acquired via content updates that encouraged the creation of multiple bots per actual player. I'm sure at some level executives are aware that bots are good for business and allow them to persist.