Pollster.com

Take this horserace result with a couple of grains of salt: the presidential preference question is #42, after a battery of energy and environment questions, which undoubtedly primed responses. There is no indication as to what constitutes a likely voter. Only a registration question proceeds the presidential preference question. Although an interest question is #47, the survey instrument says that questions #42-44 are for likely voters only, so there is no good indication which questions were used to screen likely voters (note this screening method is different from Gallup's method of constructing a voting propensity index).

Further note that on p.28 is reported the full (voting-age population), registered and likely voter samples sizes. There were 2,504 people in the full sample and 1,404 in the likely voter sample, for an expected turnout rate of 56.1% (the survey is also weighted for phone usage and demographics). The 2004 California turnout rate for the voting age population was 46.6% (it was 60.1% for those eligible, which excludes non-citizens and ineligible felons). The likely voter percentage thus looks unreasonably high since it would imply a turnout rate among eligibles of 70+%. It's going to be a high turnout election, but not that high.

Michael,

No grains of salt needed -- Obama will win California.