Do we trust a pollster who had Democrats ahead in both the Florida governor’s race and the Florida Senate race by a beefy seven points in its final polls of 2018 — only to have Republicans win both races by less than a point?

We shouldn’t trust them a lot.

But just because we don’t trust the topline numbers doesn’t mean we should ignore the trendline. Trump’s not going to lose Florida by anything close to 13 points, assuming he loses it at all, but the evidence here is strong that the state’s coronavirus surge is taking a bite out of him and Ron DeSantis. We see that in the RCP poll average too. On June 10, before the epidemic there began to rage, Biden’s lead was 3.9 points. Today it’s 7.0. Every survey of the state taken over the past three weeks has him ahead by at least six points and two of the last three have had him up by double digits, an almost unthinkable margin in Florida.

The good news is that there’s evidence that Florida’s COVID surge has peaked. As the number of daily cases comes down, Biden’s lead may — may — come down with it. One of the biggest mysteries of the election right now is whether the damage Trump has taken due to the summer “second wave” that the country has experienced is reversible or not. If cases are clearly receding by late October, do Floridians say, “We beat it, now it’s time to worry about the economy,” and vote Trump? Or do they say, “I’ll never forgive him and DeSantis for not taking this seriously enough,” and vote Biden?

It’s also worth bearing in mind that Trump routinely polls terribly in Quinnipiac surveys. There may be a “house effect” here that helps explain Biden’s implausibly large lead.

In some ways the story of the election right now is simple. Trump leads solidly with men but Biden is blowing out Trump with women. Case closed. Note too that he leads in every age demographic here, including senior citizens. In 2016 Trump won seniors in Florida by 17 points; they were decisive in his upset victory in the state over Hillary. If he’s not winning them comfortably there this year, his chances at reelection are hanging by a thread.

Again, though, it’s the trends that matter more than the raw numbers:

Voters give Governor DeSantis a negative job approval for his handling of the response to the coronavirus, as 38 percent approve, while 57 percent disapprove. In April, 50 percent approved, while 41 percent disapproved.

Voters give President Trump a negative job approval for his handling of the response to the coronavirus; 37 percent approve, while 59 percent disapprove. In April, 46 percent approved, while 51 percent disapproved.

Governor DeSantis receives a negative 41 – 52 percent job approval, a 31-point swing in the net approval from April when he received a positive 53 – 33 percent job approval rating. Today’s numbers are the lowest for DeSantis since taking office in 2019.

President Trump receives a negative 40 – 58 percent job approval in Florida, down from a 45 – 51 percent approval rating in April.

It’s hard to digest but nonetheless true that Andrew Cuomo has a higher — much higher — job approval in the COVID mass grave known as New York than DeSantis does in Florida. That’s mostly a function of the two states’ partisan demographics, I assume. Cuomo’s state is deep blue, DeSantis’s is purple, so naturally DeSantis has less of a margin for error than Cuomo does. But Cuomo made some big errors and still he soars along. It may be that he’s being graded on a curve by New Yorkers because he was the first governor to have to manage coronavirus when it arrived in the U.S. whereas DeSantis had a few extra months to try to figure out how to avert a major epidemic in Florida.

Or it may be that Cuomo’s being given a certain benefit of the doubt because he’s been more cautious about reopening than DeSantis has. When Quinnipiac asked Floridians if they thought the state reopened too quickly, too slowly, or right on time, 61 percent said “too quickly.” The fact that Cuomo is moving so slowly on reopening gives him political cover to say, “I’m doing everything I can to prioritize safety over the economy,” such that if there’s a new outbreak New Yorkers may be less inclined to blame him for it. DeSantis, by being one of the first governors to urge reopening, doesn’t have that same cover.

Other numbers related to COVID in this poll are terrible for him and Trump. Eighty-three percent say the virus is a “serious problem”; 70 percent say it’s “out of control.” And the dimming views of DeSantis and Trump seem to be influencing how the public feels about their latest push to reopen schools as fully as possible too. Both are underwater by 20 points or so on how they’re handling school reopenings, with 62 percent saying it’s unsafe to reopen elementary, middle, and high schools this fall.

In case you need further evidence that COVID-19 is hurting governors in hot-spot states, here’s what Quinnipiac found yesterday when it polled Greg Abbott’s approval in Texas:

More than half of voters, 52 percent, say looking back, Governor Abbott reopened the economy “too quickly.” Thirty-three percent say he reopened the economy “at about the right pace,” and 13 percent say he did it “too slowly.”…

Voters are split on the way Governor Abbott is handling the response to the coronavirus with 47 percent approving and 48 percent disapproving. It’s a 21-point swing in the net approval from early June when 56 percent of voters approved and 36 percent disapproved…

Governor Abbott: Voters approve with a split 48 – 44 percent of the job Governor Abbott is doing, a 20- point swing in the net approval from June when voters approved 56 – 32 percent.

Interestingly, there’s been no slide in Trump’s job approval in Texas similar to the slide we’re seeing in Florida. Maybe, again, that’s because Florida is more of a true purple state than Texas is. But I wonder if the fact that DeSantis is a Trump protege of sorts has bound his fortunes up more tightly with Trump’s in Florida than Abbott’s are in Texas. DeSantis won the GOP gubernatorial primary by positioning himself as the most ardent Trumper in the race. It paid off, and Trump was expecting him to help deliver Florida in the presidential contest this year. Now it appears that either DeSantis is dragging down Trump or Trump is dragging down DeSantis.

Things may get … awkward between them before this is all over.

If you’re looking for more reasons to discount the Quinnipiac data, here’s an interesting one. We don’t pay much attention to live-interview polls versus automated/online polls, but the pros do. And they’ve noticed a trend:

If we only had live interview polling, we'd probably say these last ten days of results among RVs–Q+15, ABC/Post+15, NBC/WSJ+11, Monmouth+13 PA, Q+13 FL–we might say that's Biden's best period yet

— Nate Cohn (@Nate_Cohn) July 23, 2020

Maybe there really are a bunch of secret Trump voters out there who are embarrassed to tell live interviewers that they’re supporting him but are more willing to reveal their true preferences when talking to a machine. Quinnipiac’s poll is a live-interview survey — but interestingly, a second poll of Florida out today that uses automated phone calls found Biden’s lead to be more modest, at just six points. Could there be six or seven percentage points’ worth of “shy Trumpers” out floating around in the electorate?

On the other hand, live-interview polls are generally considered to be more reliable than automated polls. And they’re the ones picking up blowout margins for Sleepy Joe.

I’ll leave you with this clip, which made the rounds on social media last night. I don’t know why Trump continues to insist on not just talking about the cognitive test he took but boasting about it, with increasingly greater detail about the questions he was asked. It does him no favors in prosecuting the case that Biden has “lost a step”; if anything, him rambling about it here might leave the average viewer wondering if he’s lost a step himself. Watching it was one of the few times I’ve felt genuinely sorry for him.

Hoo boy. This is rough. pic.twitter.com/B7l98sO8Pf

— Matt Rogers 🗳 (@Politidope) July 23, 2020