nate virtually 538.com says hillary has an 84.7% chance of winning. and yet are they to be believed? trump accuses that the polls are crooked. he says they are rigged. the usc- l.a. times daybreak tracking poll consistently shows trump doing better than the others. as of this morning, 44.2% for trump, hillary at 44.1. what explains the difference and how well do pre-election polls match the final results? i think you re about to be schocked. david lauder is l.a. times washington bureau chief. the methods of the poll don t meet the cnn standards and yet in 2012 that poll said obama would win by 3.38% and he won by 3.85 and andrew, a professor of statisticics at columbia
university he proved that the error in polls is double. david how does your methodology differ from the conventional polls? well, michael, there are a number of things that are different. probably the most important thing is we ask a different question and because we ask a different question we get a somewhat different answer. rather than forcing voters to say i m for this person or i m for that person we ask them on a scale of zero to 100 what s the chance that you re going to support hillary clinton, what s the chance that you re going to support donald trump, what s the chance you ll vote for somebody else and on a same scale what s the chance you ll vote at all? so that is designed to capture the ambivalence that a lot of voters feel about the election, like said it work very well four years. that s no guarantee it will work again the same way but it has a track record. and by doing it that way, when you have one candidate who has very intense support and another