Axios has a new piece out today with the headline: Biden’s Polling Denial. It’s not spin, the article says. The President and his top advisors actually don’t believe his bad poll numbers. “That bedrock belief has informed Biden’s largely steady-as-she-goes campaign,” says Axios. The article notes yesterday’s NYT-Siena poll of swing states and another recent Bloomberg set of swing state polls as examples of bad polling numbers the White House refuses to believe, before then shifting gears to note that other polls actually show him doing significantly better.

The factual questions here aren’t terribly complicated and they’re not really the reason I note this article or write this post. Most polls currently show Biden just behind Trump in a tight race. Others show him either tied or just ahead. And there is a theory of the election that those polls, with a greater emphasis on high propensity voters and the concentrating effect of the final months of the campaign, will put Biden on top in November. I’ve tried to air these different arguments here in the Editors’ Blog. You can believe one or the other.

I note the article because of what it says about the group psychology of each party and the related and intertwined factor of how the political press treats those parties.

Let me expand a bit on today’s Backchannel about that Times-Siena poll and polling issues more generally. I mentioned keeping an eye out for the delta between Likely Voter screens and Registered Voter ones. As we’ve noted a few times, this has become a consistent pattern in this presidential race: Joe Biden does substantially better in likely voter polls. And this isn’t an arbitrary difference. It’s not like saying Trump does better in the NYT-Siena poll than he does in the ABC-Ipsos one, so I prefer the latter. A Likely Voter screen is the pollster’s attempt to poll the actual voting electorate as opposed to the population of registered voters. So that distinction is very key. And if they diverge you really want to be doing better with Likely Voters, as Biden is.

If you’re continuing to rise and fall with the latest polls, you know that the NYT/Siena poll came out today with pretty unwelcome news for the Biden campaign. Biden is close to tied or just behind in the key midwestern states of Wisconsin, Michigan and Pennsylvania, but he is far back in the sunbelt states of Arizona, Nevada and Georgia. Possible preview: if Biden won the first three and lost the second three he would win the Electoral College with 270 electoral votes to Trump’s 268. These are bad numbers. There’s no arguing that. The NYT/Siena poll hasn’t been friendly to Biden, but it’s also a quality poll. I would recommend focusing on the averages which suggest a slightly, though not dramatically, different picture. Also, pay close attention to Likely Voter screens as opposed to Registered Voter screens.

I won’t lie to you. I don’t like the look of this poll. But the overall picture continues to be one in which the polls have continued to tighten, albeit not as quickly as I’d like, since around the beginning of March.

But I want to flag a different point: Congress.

Remember a week ago I was trying to find out more information or really just any information about who the assailants were when a group of “pro-Israel” rioters raided the pro-Palestine encampment at UCLA on April 30th. I’ve mentioned a few times hearing pervasive rumors (“common knowledge,” the origin of which is unclear) which I wasn’t publishing because I couldn’t verify them. But on Friday The Guardian published a story confirming that a far-right Armenian-American man named Narek Palyan had been photographed both at one of the daytime counterprotests and then on the evening of April 30th, but still hours before the overnight assault.