Google Chrome came in first, with 97% of people using the most current version after 21 days. Mozilla Firefox placed second, at 85% after 21 days. Apple Safari came in third, with 53% after 21 days. And Opera, with its manual reinstallation requirement, came in last, with 24% after 21 days.
Microsoft Internet Explorer wasn't included in the test because it only reports the major version number in the user string sent to Web servers, which prevents the tracking of patches over time.
"The often-stated reason for this omission is to reduce information leakage and make it harder for an attacker to select a working exploit for the actual browser version in use," the researchers observed. "As we have seen drive-by download Web sites trying many different exploits at once, it's unclear how much additional protection this omission really gives."
While some might argue that Duebendorfer's goal in participating in this study is to make a case for using Google Chrome, the researchers also have more altruistic motives. They conclude by encouraging browser makers to adopt Google's silent update mechanism, which has been made available in the form of an open source project called Omaha.
Coincidentally, Google Chrome was updated on Tuesday to address two security vulnerabilities.
InformationWeek Analytics has published an independent analysis on the current state of security. Download the report here (registration required).