Well, we'd need to see whether the percentage of voters this year that only cast ballots for president differs greatly from other years. Google that and try to find an answer. Where the district by district polls vary as greatly from the result as do the presidential polls v thus far reported results? Try to find that on Bing. Anomalies might be indicated given 50 results with only a small number of deviant results that aren't as apparent among 460 different races, each with differing sets of candidates rather than just the two (realistic options) under consideration with the presidency.