Hillary canceled and event today. I wonder if it has something to do with her wacky eye movements yesterday which Drudge featured prominently today... hiding a major illness definitely causes scheduling conflicts. Drudge had a big headline UNFOCUSED...
almost all the methods of the pollsters and nate silver were totally discredited by their failures in predictions in 214 and their predictions about Trump. saying that presidential election polls work better is unfounded and probably specious. especially since we know the herding makes all the pre herding polls ridiculous.
Hillary Clinton cancelled an appearance in Chapel Hill, NC today. Clinton was supposed to hold a "Conversation with Hillary" event at the home of Better Craven and Michael Warner.
many polls had romney ahead a week out. you complained about that yourself. nate silver admitted he blew it with trump. the 2014 polls were ridiculously bad. your beloved polls pretty much had trump being destroyed a few weeks ago. now most of them are within the margin for errors except for the Nothing But Crap polls. so were they wrong a few weeks ago?
1.Most polls did not have Romney ahead a week out,only Rasmussen and Gallup. 2.Nate silver admitted he blew it with trump but A.Silver said he ignored polls when he made that call. B.As previously stated,I don't relay only on Silver. C.The polls did not blow it with Trump.The majority of them had him ahead from July 2015. 3.To discuss the 2014 polls we would have to examine every race and every poll.Sorry I will have to pass on that. 4.The way I look at polls they aren't wrong now or a few weeks ago.The good polls still have Hillary ahead.I don't look at how many points they have a candidate ahead,all I look at is who they have ahead in the large majority of their polls be it +1 or + 10.The good polls lowered Obama's numbers as the race got closer to the end,but the majority of them still had Obama ahead.Hillary being ahead in 7 out of 10 credible polls at +1 or + 10 makes no difference in The TS algo,all that matters is that she is ahead in 7 out of 10.If NBC does 20 polls for the 2016 election and 15 had Hillary ahead and she wins they were accurate for 2016,doesn't matter if they had her ahead +1 or +10.
Some differences in The TS algo and the jem algo as I see it. The final 2012 polls.The jem algo says most polls are tied or in the margin of error.The TS algo says Obama is ahead in 6 and Romney is ahead in 2.The two polls that have Romney ahead were consistent outliars for most of the race having Romney ahead while most had Obama ahead. Rasmussen is the worst poll on RCP so The TS algo takes that into consideration as well.The jem algo on the other hand gives more weight to Rasmussen and gallup because it believes they are honest and unbiased in skewing/herding/R vs D stuff that The TS algo doesn't consider.The TS algo considers that the majority of polls had Obama ahead the last 3-6 months and how many points he was ahead didn't matter.