Kevin Purdy ran benchmark tests on Chrome, Opera, Firefox, and Internet Explorer. He scored them university-style against the total number of points they could have received and ranked them.
Purdy should have included Apple's Safari. It is popular enough to deserve inclusion.
He also should have rated the browsers on security or borrowed ratings from elsewhere. However, I speculate that he might not have done so because he was concerned only with performance. Security also depends on the use of anti-virus and anti-spyware applications such as my favorites, Microsoft Security Essentials, Spybot Search & Destroy, and Lavasoft Ad-Aware. In fact, security depends so much on security tools and on security awareness that the security of browsers may be a wash.
Not satisfied with Purdy's scoring method, I created a table and assigned my own scores (1 to 10, with 10 being best) to the performance of each browser.
|Click on the table to see a larger view.|
Purdy did not list results for Internet Explorer in the Memory Use with Extensions test. I assume this is because you can't use extensions with IE, or at least not the same ones available with other browsers. That reflects very poorly on IE's flexibility, so I don't feel bad giving it a score of zero.
Purdy did not list results for Opera 10.63, either. I assume Opera 10.63 had issues similar to IE's limitations. However, since Opera 11 has a score, some functionality must be there for 10.63, so I gave it a 1. Give me a reason to change it and I will.
Next, I ranked the tests in order of importance and assigned weights to them.
I made a decision matrix spreadsheet, normalized the weighted scores, and found the following results.
Or, if you prefer a bar chart....
|Click on the figure for a larger view.|
Surprisingly, my scores did not greatly differ from Purdy's.
Chrome and Opera have the best performance, with Google Chrome 10 winning by both Purdy's scoring and by my weighted scoring. If you use Firefox, you can do better (although you may prefer the add-ons that are available for it). If you use MS Internet Explorer, you can do a lot better.