People routinely attempt to predict the presidential election the wrong way. They look at popular polls over the whole country. Those are meaningless. It also doesn't matter who all your friends are voting for. Mainly because if all your friends are voting the same way, you live in a state that is already going 100% for one candidate, and all you have done is predicted how one single state will vote.
Nor is the economy the be all and end all. Among other things it can turn on a dime. And it isn't the best predictor. Most importantly, President Trump is more divisive than most - some people want him impeached while others think he is god's gift to mankind.
We use an electoral voting system. As such, you need to look at polls by each state, except for the two states that split their vote. For those two states, Maine and Nebraska, you need to look at polls by Congressional district.
Most states (and districts) are forgone conclusions. Even if the the Democrats pick a dead horse, they are going to win New York, California, etc. etc. Similarly, even if the Trump kills and eats a person on national TV, he will win Mississippi and every state that touches Mississippi.
There are only five states and one district that are "Swing States" - they are too hard to predict and could go either way. There are also another six or seven states and one district that are "Long Shots". The rest are already set in concrete and it will take a jackhammer to dislodge them.
The five real swing states are (in order of how many electoral votes they have):
The swing district is Omaha district 2 of Nebraska. The Republicans are going to take the state of Nebraska, but either party can win that district for a single extra electoral vote. Worth at least 1% of your total ad budget. Probably 2% if it is close, and if it is not close, your ad budget doesn't matter.
The long shots are (some people think Ohio is set in stone, but I disagree):
But the thing about the long shots is that usually they can not be won unless your candidate is strong enough to win the swing states. In which case, they do not matter. But not always. The reasons the Democrats lost in 2016 was in part because they thought they had Michigan and Wisconsin in the bag. They had them as Long shots, not swing states. Granted, Donald Trump did win the 2016 swing states as well, but still. The Democrats paid too much attention to national polls that correctly predicted Hillary winning the popular vote, and not enough to state polls.
For those reasons, if you want to predict the 2020 Presidential Election, you need to concentrate your data collection on the 5 swing states plus Omaha metropolitan area.
If you know what people in those areas are thinking, you know how the election will turn. Whoever wins three of the five swing states will almost certainly be the president.
As a a side note, if I were running the a presidential election, I would spend more than 1/2 of my money and time in those five swing states (plus Omaha). Probably less than a 1/4 of my money and time on the seven long shot states and rural Maine. That leaves less than a 1/4 to be spent on the remaining 33 states (and territories) of the US.
Why spend any time or money on the foregone conclusions? To help Congressman and Senators that have a chance of getting elected with your help. You may lose a state but gain a seat in Congress. Also, Puerto Ricans may not vote for the President, but they are citizens of the US (Citizens but no vote, is that not the vile?) and can move to the mainland US (Florida perhaps...), where they can vote. Quite a few of them did that after the last horrible storm when President Trump failed to respond in sufficient force. Those citizens will be voting, and I suspect few will vote Republican.
Nor is the economy the be all and end all. Among other things it can turn on a dime. And it isn't the best predictor. Most importantly, President Trump is more divisive than most - some people want him impeached while others think he is god's gift to mankind.
We use an electoral voting system. As such, you need to look at polls by each state, except for the two states that split their vote. For those two states, Maine and Nebraska, you need to look at polls by Congressional district.
Most states (and districts) are forgone conclusions. Even if the the Democrats pick a dead horse, they are going to win New York, California, etc. etc. Similarly, even if the Trump kills and eats a person on national TV, he will win Mississippi and every state that touches Mississippi.
There are only five states and one district that are "Swing States" - they are too hard to predict and could go either way. There are also another six or seven states and one district that are "Long Shots". The rest are already set in concrete and it will take a jackhammer to dislodge them.
The five real swing states are (in order of how many electoral votes they have):
- Florida (29)
- Pennsylvania (20)
- Michigan (16)
- Arizona (11)
- Wisconsin (10)
The swing district is Omaha district 2 of Nebraska. The Republicans are going to take the state of Nebraska, but either party can win that district for a single extra electoral vote. Worth at least 1% of your total ad budget. Probably 2% if it is close, and if it is not close, your ad budget doesn't matter.
The long shots are (some people think Ohio is set in stone, but I disagree):
- Ohio (18) - most likely Republican
- Georgia (16) - most likely Republican
- North Carolina (15) - most likely Republican
- Minnesota (10) - most likely Democrat
- Iowa (6) - most likely Republican
- Arizona (6) - most likely Democrat
- New Hampshire (4) most likely Democrat
But the thing about the long shots is that usually they can not be won unless your candidate is strong enough to win the swing states. In which case, they do not matter. But not always. The reasons the Democrats lost in 2016 was in part because they thought they had Michigan and Wisconsin in the bag. They had them as Long shots, not swing states. Granted, Donald Trump did win the 2016 swing states as well, but still. The Democrats paid too much attention to national polls that correctly predicted Hillary winning the popular vote, and not enough to state polls.
For those reasons, if you want to predict the 2020 Presidential Election, you need to concentrate your data collection on the 5 swing states plus Omaha metropolitan area.
If you know what people in those areas are thinking, you know how the election will turn. Whoever wins three of the five swing states will almost certainly be the president.
As a a side note, if I were running the a presidential election, I would spend more than 1/2 of my money and time in those five swing states (plus Omaha). Probably less than a 1/4 of my money and time on the seven long shot states and rural Maine. That leaves less than a 1/4 to be spent on the remaining 33 states (and territories) of the US.
Why spend any time or money on the foregone conclusions? To help Congressman and Senators that have a chance of getting elected with your help. You may lose a state but gain a seat in Congress. Also, Puerto Ricans may not vote for the President, but they are citizens of the US (Citizens but no vote, is that not the vile?) and can move to the mainland US (Florida perhaps...), where they can vote. Quite a few of them did that after the last horrible storm when President Trump failed to respond in sufficient force. Those citizens will be voting, and I suspect few will vote Republican.
No comments:
Post a Comment