Going Poll to Poll

Yesterday, a new poll from CNN was released, and I spent a good amount of the afternoon writing about it, talking deeply-alarmed Democrats down off the ledge.  The thing is, the poll not is not only not nearly as bad as the perceived appearance, but actually the one question that got most of the attention it was close to nonsense.

At its foundation, the poll showed Joe Biden leading Trump by five points, though that was closer than some other polls recently -- but what got the most attention was a result that showed Trump beating Biden by seven points in the 15 battleground states.  Worse, one friend said he'd heard a report that made the poll sound like it showed Trump winning all 15 battleground states.

First of all, I don’t care about any individual poll.  It's not just that a single poll is just a "snapshot in time," as is the phrase.  It's just a snapshot of that one polling service.  Another poll four days ago showed Biden up by nine points.  I don't know if he's actually up by that much nationally.  But I know that this shows how different one poll can be from another, so it's no use reacting much to any one poll result.  Further, there’s a margin of error built into every poll, and that means a single poll can be deceivingly flexible from what it was the week before -- actually giving the same results (with the margin-of-error added in) even though the cold numbers alone show a difference.  Yes, I certainly wish Biden's lead was bigger in this week's CNN poll, but it still had him up by five points.  That's not only about what most polls showed about a month ago, but it's also much-better than being behind by five points.  Moreover, with a margin of error of 3.7 points, that means the real difference between the two candidates could actually be 8-9 points.  Yes, it could be razor-thin tight in the other direction, but given results from all the other polls that seems unlikely -- and given the one previously-mentioned poll that had Biden up by 9, it would suggest that if the margin-of-error is a factor, then a higher result would be more likely.  But even if it really is a 5-point difference, okay, fine, not bad.

Much more to the point, though, is their poll result of 15 battleground states, showing Trump up in those by 7 points.  That’s what got the most attention and freaked out a lot of people.  But the thing is -- bunching battleground states together as One Big Number is totally meaningless.  Know this: 

If Joe Biden holds all the Blue states, which he certainly should, and of those 15 battleground states he only wins Arizona, Michigan and Pennsylvania – all of which he’s ahead in – Biden wins the election!  He could lose every other battleground state and still win the election.  But further, as it happens he’s also ahead in the battleground states of North Carolina, Wisconsin, Colorado and Florida.  If he wins all those, the election turns into a landslide.  So, the “15 battleground states” poll of them together really, truly doesn’t matter at all.

Again, the be clear, the poll did not show Trump winning all 15 battleground states -- it was an aggregate of all the voters together in those 15 states.  Which is why I say it's meaningless.  (You can look at the 49-page poll results here.  The “Who would you vote for?” question for people living in the 15 battleground states is Table 072.) 

And by the way, how they came up with there being 15 battleground states is totally beyond me – usually it’s that there are 6 or 7 battleground states.

There’s a wonderful website, www.270towin.com which has an interactive map you can manipulate and change each state to whatever you want – from safe to lean to toss-up for each party, and the electoral votes change for every category.  I was discussing all this with a friend, and he went to the website and changed every state to Red except the core Blue states and the aforementioned Arizona, Michigan and Pennsylvania only – and sent me a screenshot.  You’ll note in the upper-left corner the overall total of “Democrat – 270.”

When I explained all this to a friend, and he finally understood that the poll was not individual battleground states but rather them all crammed together, his deeply-relieved response was, "Oh, so in other words, the poll is garbage."  He had been especially freaked out, too, because he'd heard a commentator say that if if Trump won all the battleground states, as the poll seemed to suggest, then Trump would win the election.  As I explained, not only was it NOT true that the poll showed Trump winning them all, but the analysis was head-bangingly silly (which is the polite term)  – I mean, seriously now, of course if Trump won ALL the battleground states he’d win!  But that’s not much different from saying, “If someone gets the majority of the votes, they will win.”  

Believe me, I understand why people were distraught.  As I said, I spent a lot of time writing back to people and explaining as much as I could online.  They were SO relieved!!!

Here again is the reality -- at the moment, Biden is ahead in the battleground states of Colorado, Arizona, Michigan, Wisconsin, Pennsylvania, North Carolina and Florida.  That’s ahead in 7 of the 15 battleground states.  He won’t win them all, but again, if he just wins Arizona, Michigan and Pennsylvania – and loses every other battleground state (!!), even the ones he’s now ahead in -- and holds the Blue states, he wins.

To be clear, this doesn’t mean Biden will win, of course.  Nor that Trump will.  It just means that the perception of what the CNN poll said was totally, 100% misinterpreted by so many people.  Especially since it was a really stupid way to poll the battleground states.  And even more especially since Biden is ahead in about 7 of the 15.

But let's go even a step further, because there was some news in the poll which I thought was, in fact, significant.  Yet it got lost in the soul-wrenching concern.  It was a result that showed Trump going from “up 6 points” in 2016 against Hillary Clinton among voters in 65-and-older to now “down 15 points” in that same category.  That’s a massive shift of 21 points – and I suspect it could grow given how cavalier Trump is about coronavirus deaths, which impacts this group the most.  Making this even more significant is that older voters traditionally are the ones most likely to vote.

That's an actual story.

Polling 15 battleground states together, on the other hand, is garbage.

Total garbage, but headline-making that got CNN the attention they wanted.  But what an insanely meaningless way to handle battleground states.

And there are still six months to go.  Buckle up.