View Full Version : Polls and pundits

Norman Bernstein
09-21-2016, 08:39 AM
Hillary supporters might argue that they have a real reason to be depressed and discouraged today: www.fivethirtyeight.com (http://www.fivethirtyeight.com) now lists the 'chances' of the election as 57% Hillary, 43% Trump. That's the narrowest margin since August 14th. Yesterday, it was even closer than that.

However, the problem with fivethirtyeight is that these numbers are 'predictions' or 'forecasts' based on the odds.... as seen by Nate Silver. They are NOT a count of electoral votes. While Silver has a reputation for being canny about these things, he's also gotten it wrong, occasionally, so his predictions are a great deal like Las Vegas oddsmakers: they tend to be 'grossly' accurate, but deviate a great deal.

Not that he isn't generally 'right', in the sense that the race is tightening... something I've said over and over for months now. ALL Presidential races tighten as they approach election day, and it's rare to ever see a Presidential 'blowout' in the popular vote (and only slightly less rare to see an electoral vote blowout).

I think it makes much more sense to watch the electoral vote... and www.270towin.com (http://www.270towin.com) is the best site for this that I've seen.

270towin offers up two electoral maps. One is based purely on state polls, in which only states where the popular vote margin is in excess of 10% (certainly, 'safe' states for a candidate) are pictured in blue or red.... states with popular vote margins of from 5-10% are pictured in light red or light blue (the 'leaning' states), and 'tossup' states, pictured in tan, are states where the popular vote margin is less than 5%, which is not far above the typical margin of error, in a poll.

Today, the 'polls only' map shows 224 electoral votes for Hillary that are either 'strong' or 'leaning', 120 votes for Trump that are either 'strong' or 'leaning', and 194 'tossup' states. This doesn't mean that the 'tossup' states don't have a 'lean' one way or the other... just that the 'lean' is less than 5%, when the margin of error in the poll may be not much less than 5%. (In this map, PA is considered leaning blue, and Florida is a tossup).

The OTHER electoral map, however, MIGHT be more instructive. It takes into account not only the polls, but some judgment: judgment about how races have gone in the states, in the past, the political sentiment in these states, and professional observer's judgments about what the polls might mean. NONE of this is 'certain' in any sense... but surely, there is more information to be had than just the pure poll numbers.

On THIS map, Hillary has 273 electoral votes (enough to win), Trump has 191 electoral votes, and 74 electoral votes are still considered 'tossups'.

I am NOT saying that this version of the map is accurate.

Another metric commonly used in election forecasts is the 'paths to win' count... this is a numerical analysis of the various ways that state electoral votes could combine, to win the election. It describes the number of combinations of states that might deliver a win, to each candidate. www.fivethirtyeight.com (http://www.fivethirtyeight.com) posts a graphic to show this, and as of today, Clinton has 751 possible winning paths, Trump has 259 paths, and 14 paths result in a tie.

Finally, it may come down to a small number of critical states... and even more strongly, who wins PA and FL, with 20 and 29 electoral votes, respectively. Fivethirtyeight offers this graphic:


Once again, this is yet another 'oddsmaker' graphic, which doesn't reflect the pure polling data. If PA and FL are examined separately, then the average polling data for those states is as follows: Florida is dead even, while PA is 47% Hillary, 41% Trump.

What does this all mean? Leaving partisan bias aside, there's no question that Hillary has a better chance than Trump, whether you include the pundit judgments, or not. It's NOT a big margin... but it's still there.

09-21-2016, 08:42 AM
What it means, in part, is that a shocking percentage of American voters slept through civics classes, or have very little actual respect for the Founders and their efforts to enable a "more perfect union."

Norman Bernstein
09-21-2016, 08:48 AM
What it means, in part, is that a shocking percentage of American voters slept through civics classes, or have very little actual respect for the Founders and their efforts to enable a "more perfect union."

There still would be a 'saving grace' to a Hillary win: it would force the GOP to undergo a radical transformation, and will probably split the party apart. What's the true percentage of Trump supporters among the entire GOP? My guess is that it's actually less than half.. which will decrease somewhat if Trump loses... and the rest of the party will have to face the fact that their message has lost them three consecutive elections, and a change is needed.

What is likely to happen is the emergence of a reasonably sizeable 'radical right party', distinct from the mainstream Republican party, which will likely insure Democratic presidential wins for some time to come. The downside is that the radical elements will continue to elect legislators that will preclude GOP majorities in congress from moderating, thereby continuing the legislative logjam.

The collapse of the Republican party is NOT something I see as good for the nation.

Jim Mahan
09-21-2016, 09:27 AM
The GOP needs to gnaw off it's own leg to get free from the tea-stupids.

If the party, led by the usuals, could agree to let go of one plank in the platform, just stop being completely intransigent for one worthyt issue, and so begin a general process of rehabilitating the party, and so also the process of government, and freeing up the logjam. Offer a concession that isn't just lip service and chin music, and cohere to fight again another cycle. How could it be any other way? What is the alternative, if not stultifying political chaos?

Norman Bernstein
09-21-2016, 09:36 AM
The GOP needs to gnaw off it's own leg to get free from the tea-stupids.

That leg will continue to live, and jerk it's knee. The 'Tea Party' mentality is not going away. The REAL question is whether the GOP will 'double down' on it's new ultra-right faction... or split into two parties, which will damage their ability to be a genuine opposition party. It's pretty clear that if the 'leg' consumes the rest of the GOP 'body', the 'leg' sin't going to be a contender for the Presidency for perhaps decades.

But this is thread drift... the point of the thread was to indicate that, despite the 'tightening' of the race, Hillary still has a better chance of winning.

09-21-2016, 10:03 AM
Cathargo delenda est.​ You know the drill.

Norman Bernstein
09-21-2016, 12:53 PM
Cathargo delenda est.​ You know the drill.

Well, nothing is forever... as both you, as well as history, illustrates.

On the other hand, we all tend to focus on the nearer term, than on events with 'historical' scale.

Nearer term, this election, more than any other in our lifetimes, will reshape the Republican Party....

...we just don't know what that new shape will be.