Tags

webcam vindaloo version vegan unix unicef trojan todo thinkpad textmate testing tagging syntax svn subversion sphinx spaces solaris sitemap sinatra sheet security search schema_info SchemaInfo ruby rinari relationships refresh rdiff-backup ramaze railsconf08 railsconf07 rails protools production power placeboeffect pink floyd PIC perl overheat outbreak osx os x NYHS NYC netbeans nanophotonics mysql music MPEG-4 model migration microvolunteer macbook mac log linux less leopard keynote JAX javascript java imunizator highlighting Handbrake haml hacks google geocoding genghistron gem gaming gabrielle's funny functional fun friends food fixes fixed firefox FF3 ferret fantasy' emacs DV donate datarecovery database D&D converter conference computing cheat capistrano business bribes

The Fixer said on Tue Aug 05 20:02:57 +0000 2008 | permalink
Tagged: placeboeffect

Microsoft's Mojave experiment

There’s been a bunch of press about Microsoft’s Mojave experiment, where they presented Vista to some naive users, telling them it was actually a new operating system, code-named ‘Mojave’. Unsurprisingly, it rated better than Vista, even though it was just a copy of Vista. This is actually a very valuable experiment, since it tells us exactly how much the placebo effect controls perceptions of a new operating system, and the bar which Microsoft must exceed in order to actually make improvements to their operating system. Microsoft claims a pre-demo rating of 4.4 for Vista and a post-demo rating of 8.5 for Mojave, so now I know that anything less than a doubling of user ratings for Microsoft’s next OS means it actually got worse ;-)