It is now almost August, and we still know very little about Joe Biden. Nearly every other Presidential Candidate in the past two decades has made frequent appearances on Cable News.
What is Joe Biden afraid of?
I agree with others here who think he's not afraid it's a purposeful strategy. Coronavirus and the economic fall out of the response are making Trump far less popular in the general public. This means Trump will have a much harder time gaining voter than he would have last year. And he was....
The misassumption of most people, is they don't get the difference between an election model and a polling aggregate. Some take this even further and don't see the difference between a polling aggregate and a single poll. If you took "the polls" at face value you'd think Americans are voting different all the time. The truth is those who change their vote is quite small and elections are won and lost by far more specific shifts.
An example, in 2016, no matter how you cut it, Hillary's 65,000,000 votes were around the predictable baseline of Democratic voters for that year. It's very unlikely she netted any voters(gain more than she lost) during the campaign, this could be seen as multiple indicators showed she was never very popular and didn't add much beyond the work of who came before her. Polls meanwhile made it seem like there were dramatic shifts, an unlikely truth, she was solid, but that would only be true if Trump pulled in the predictable baseline of Republican voters for that year. Which would have been around 59-60 million votes. He netted only 3, but every indicator showed his gains were much higher having loss a sizeable chunk of the baseline.
Compare this to say Obama in 2008, who got 69.4 far above the predictable baseline of Democratic voters. By 2012, you already saw the new drop of those new voters back expanded base, both victories of course because the republicans never made gains. Again, this is just not what polls said even if they got it "right", it's a lot of noise to paint a big dramatic story of lots of swing-voters and such, but that's just not reality.
No, the truth is polls have a sample error. Lots of people don't respond, even more, lie about their likelihood to vote, so the best you can get is an approximation of popularity and how that falls by differnt groups. You can certianly use that to predict gains and losses, but only by applying an accurate model, which takes those data points and applies them to realistic predictors, not just by applying regions, but also by likelihood and at what rate different groups are likely to vote.
The trouble for Bidens is since 2016 Trump has already netted a good amount of voters. This gain may be down to trickle now, but if Dems are mistaken if relying on the higher democratic baseline. We're likely in for shocker come November. I'd predict Biden still needs to net a solid million to win. Doable, but unlikely. Of course, my model might be wrong -- I just hope, when trying to predict an election people are in fact comparing models, not simply raw polling aggregates which count on a lot of people who don't plan to vote. And a lot of especially working-class types (the rarely sampled, and oft called politically disenchanted) staying to their nature and not voting, a bad bet on a candidate like a Trump who base is the politically disenchanted.