Siena historically has a good record within the state of New York, right? Although I agree that their polls this year have not been very consistent.
I also agree with jaichind that this is probably not a good poll for Trump. Many national cross-tab patterns are suggestive of a situation where Republican gains are concentrated in the largest American megacities, which are mostly not found in competitive states. NYC is the most megacity of them all and it would make sense for patterns flowing from density or recent immigrant background to hit it harder than other places.
Can we trash this notion that “this poll is bad for Trump”? NY has not had an overwhelming influence on the popular vote since the 1940s. Unless Trump outright wins the state, the NY results wont swing the national vote by 1% (if even that). If we get polls with Trump within 10% in California and over 15% in Texas, then this argument can arise
If your argument is that, “this NY poll means there is a similar swing in Los Angeles, Chicago, Boston, and San Francisco too” … the issue is that NY is facing a more vicious anti-Democrat backlash because of the migrant crisis than those other cities are facing, and Zeldin’s performance gives Repubs a groundwork for a swing . It’s possible that NY swings right while those other cities stay as they are