Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
You've gone off the reservation and into personal prediction land.
Not really. It's not all about the BS'd national polls that come out every other day. Every analyst agrees on the electoral vote total and the popular vote is projection from what can be inferred from state polling. Keep in mind that Trump is doing far worse than Romney in many red states like Texas, Arizona, and Georgia. Those votes will add up even if he does not win the state.
Then there's demographics. I think it was post Mr Inbetween who raised the question the other day how you can be losing badly in every demographic but poorly educated white males and still be close in the polls. The answer is, you can't. So either all the demographic polling is wrong or the sample being surveyed in the BS'd polls is off. I pick the latter.
I don't really see how that would be possible considering it looks like RCP had Romney down by an average of ~4 points nationally for the month of September. Regardless, all that matters is where Romney ended up...at 10%. As things stand now, Trump has much better odds - essentially 1 in 3 (with momentum) vs 1 in 10. I wouldn't expect any two elections to track exactly the same for months and months. I think it's safe to say Trump's odds on Election Day won't be crushed relative to Romney's....Trump's 538 odds weren't even that bad at his worst (the week following the DNC).
Again, you can't compare TODAY in one forecast to ELECTION DAY in the other. As election day nears, the uncertainty in the projections tends to go down because there are fewer variables that can crop in that much shorter period of time leading up to the counting of the votes. The nature of a statistical analysis is that the LESS the uncertainty (or the number of possible variables) the GREATER the probability.
In the case of Nates' 3 forecasts for example, as we get closer and closer to the election those 3 forecasts will align more and more and the probability of SOMEONE winning will increase higher and higher. It's just like the odds in a superbowl football pool. As we get closer and closer to the big game and more and more teams are eliminated the odds of a particular team winning it will generally go up and up as long as the team is still in the running. In this case, the factors affecting the odds are not the number of teams involved (since that number is already set) but factors such as "chance of major terrorist attack before the election", "chance that Trump will make another gaffe", "chance that a really incriminating e-mail comes out about Hillary", "chance the economy will tank", etc, etc, etc. All of those types of things are factored in to the polls plus calculations - and the less time there is between the projection and the actual election, the less likely something like that will crop up. In addition as we get closer and closer to election day, more and more "undecideds" will decide - which again reduces another area of uncertainty in the model(s).
The consequence of all that is that Nate's probabilities (for SOMEONE) tend to trend upward as election day gets closer and his projections become more and more lopsided. This tends to be so even when the election is close (remember, that projection is just the probability of a win, not the percent the winner will win by). So, Trump may have a 1/3 chance of winning TODAY, but IF the polls say on election day EXACTLY what they say TODAY, that 1/3 chance that Trump holds TODAY is likely to be somewhere under 1/10.
So, in order to be a valid comparison you have to compare where Trump is TODAY to where Romney was at this time - and THAT comparison shows Trump BEHIND where Romney was at this time. Does that mean Trump CAN'T pull it off? Of course not, but he can ONLY do that by doing a LOT better than Romney did between now and election day.
And yes, Trumps odd WERE under 10% (FAR lower than Romney's ever were - even just before election day). Trumps lowest Now-cast was a mere 3.6% on August 8th - when he was making gaffe after gaffe and Hillary was riding high after her nomination. Trump has of course shut his mouth in regards to "off the cuff" statement (at least to a degree) and so he's not hurting himself quite so much anymore. He likely will though. He can't seem to help himself.
^ Sorry not buying that candidate X polling at a certain percentage at a given time is an indicator of how another candidate will do. I don't think it's even a basis of comparison because no two elections are the same. Regardless, like I said, Romney was losing by ~4% according to RCP at this time - I would be highly skeptical of any forecast that gave a candidate losing by that margin a nearly 70% chance of winning. Do you have a direct link to the data?
Variables will go down as we get closer to Election Day, but that doesn't guarantee that we'll be anymore certain of who's going to win percentage wise. If that was the case, every election would end up with one candidate being a ~90% favorite. I'm pretty sure there have been elections with 20/80, 30/70, 40/60 splits heading into the day of the election. If today's data were to say Nov 5th, would Silver's now-cast not be a 30/70 split? If so, what value does the now-cast have?
^ Sorry not buying that candidate X polling at a certain percentage at a given time is an indicator of how another candidate will do. I don't think it's even a basis of comparison because no two elections are the same. Regardless, like I said, Romney was losing by ~4% according to RCP at this time - I would be highly skeptical of any forecast that gave a candidate losing by that margin a nearly 70% chance of winning. Do you have a direct link to the data?
Variables will go down as we get closer to Election Day, but that doesn't guarantee that we'll be anymore certain of who's going to win percentage wise. If that was the case, every election would end up with one candidate being a ~90% favorite. I'm pretty sure there have been elections with 20/80, 30/70, 40/60 splits heading into the day of the election. If today's data were to say Nov 5th, would Silver's now-cast not be a 30/70 split? If so, what value does the now-cast have?
Romney had a ~15-20% chance at this time in '08. Thus, your entire argument is pointless.
If you notice, the Now-Cast tends to always show a higher probability of one of the two winning. In other words the spread is greater - ALWAYS. Look at the "historical" charts and go back and forth between the 3. Now-cast ALWAYS shows more of lop-sided win for SOMEONE than the other 2 forecasts do. That's because there less uncertain the closer you draw to the election. It does NOT mean it will always be 90% for someone, but for whoever is ahead (assume one of the two IS ahead and that they are not tied) that probability WILL climb the closer the election is.
I do grant that your link was better than my "second hand" link to message board and that it does indeed show Trump doing better than Romney did at this time. That link is a great find so thanks! You'll get a +1 from me for that.
We can talk all day, but Romney's chance of winning at this time was 15-20%. Trump beats that by a good 10%. Wasn't the point you were making hinged on Romney having a nearly 70% chance of winning in September? That number looks to be proven false.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
Town Hall with the military and the veterans NBC poll
