4 ms·

So the fate of the world rests in the hands of Floridians...

by splatcollision 6y ago

Damogran6 6y ago
See: Hanging Chads
dumbfoundded 6y ago
Or PA.
bryanlarsen 6y ago
According to https://projects.fivethirtyeight.com/trump-biden-election-map/ https://projects.fivethirtyeight.com/trump-biden-election-ma... If Biden wins Florida, then it's a >99% chance of a Biden victory. But if Trump wins Florida, it's a 58% chance of a Biden victory. So Florida is essential for Trump, but not for Biden. Biden has a 70% chance in Florida. So I wouldn't say the fate of the world rests in the hands of Floridians. However, an early Biden win in Florida could allow networks to call the election on election night rather than having it drag out for weeks like it otherwise might.
- rayiner 6y ago
  If Trump wins Florida and Pennsylvania, there is a 92% chance of Trump winning.
  - erichocean 6y ago
    I get the sense you are expecting a Trump win, correct?
    - rayiner 6y ago
      I really have no idea. It just seems hard for me to believe Biden will win either Florida or Pennsylvania, even though I see the polls are predicting that.
  - CameronNemo 6y ago
    PA results will not be known on election night, Florida results will.
- robrenaud 6y ago
  You are giving a little bit too much weight on Florida in your analysis and perhaps confusing the casuality a little bit. It's not just directly that the Floridians have that much influence, it's that knowing the outcome in Florida tells one quite a bit about, the direction/magnitude of polling errors across the nation. The 538 model factors in the outcome of Flordida's election with its own belief about the other states results. This is why if you select Biden wins Texas, the model believes the national result likely in landslide territory, the Biden expected electoral college outcome goes up +70, even though Texas only has 38 electors. If Biden wins Texas, he also probably wins in Georgia or maybe even Arizona.
  - bryanlarsen 6y ago
    Where did I mention any causality?
lmilcin 6y ago
Actually, if you take a look at the graph, you will notice that paths that direct to red leaves are almost all almost entirely red whereas there is a lot of paths to blue leaves that are mostly or even almost entirely red. Or think another way. The largest number of states that can be blue and red still wins is 4. There isn't any combination of any combinations where 5 of selected states are blue and red can still win. To calculate actual probability don't look at just the number of read leaves to number of blue leaves. You need to take into account probability of each of the state being blue or red. I am too lazy to calculate it but at first sight red outcome seems extremely unlikely.
- 205guy 6y ago
  I believe that the "direct" path of one color and "alternating" path are just a choice of the graphing and not dictated by the data. If you make different node selections, you could make the reverse. Same for the layout of the nicely ordered trees and those that seem more chaotic--it's just a matter of adjusting the layout of the nodes to determine the angles of the edges.
  - lmilcin 6y ago
    I don't mean the order, but the number of paths with given number of nodes of given color. This is independent on the chosen order or layout of the graph.
advisedwang 6y ago
The visualization presents Florida as the first decision point, but that is a choice of the visualization - you could make it the last choice and the visualization wouldn't look like this. That said Florida does have a large number of electoral college votes and can swing, so it makes sense as a a choice of the designer.
dehrmann 6y ago
Fivethirtyeight puts Binden's change at winning Texas(!) at 36%.
dillondoyle 6y ago
And Pennsylvania. But yeah if Trump loses Florida it's really hard.

So the fate of the world rests in the hands of Floridians...

Damogran6 6y ago

See: Hanging Chads

dumbfoundded 6y ago

Or PA.

bryanlarsen 6y ago

According to https://projects.fivethirtyeight.com/trump-biden-election-map/ https://projects.fivethirtyeight.com/trump-biden-election-ma... If Biden wins Florida, then it's a >99% chance of a Biden victory. But if Trump wins Florida, it's a 58% chance of a Biden victory. So Florida is essential for Trump, but not for Biden. Biden has a 70% chance in Florida. So I wouldn't say the fate of the world rests in the hands of Floridians. However, an early Biden win in Florida could allow networks to call the election on election night rather than having it drag out for weeks like it otherwise might.

rayiner 6y ago
If Trump wins Florida and Pennsylvania, there is a 92% chance of Trump winning.
- erichocean 6y ago
  I get the sense you are expecting a Trump win, correct?
  - rayiner 6y ago
    I really have no idea. It just seems hard for me to believe Biden will win either Florida or Pennsylvania, even though I see the polls are predicting that.
- CameronNemo 6y ago
  PA results will not be known on election night, Florida results will.
robrenaud 6y ago
You are giving a little bit too much weight on Florida in your analysis and perhaps confusing the casuality a little bit. It's not just directly that the Floridians have that much influence, it's that knowing the outcome in Florida tells one quite a bit about, the direction/magnitude of polling errors across the nation. The 538 model factors in the outcome of Flordida's election with its own belief about the other states results. This is why if you select Biden wins Texas, the model believes the national result likely in landslide territory, the Biden expected electoral college outcome goes up +70, even though Texas only has 38 electors. If Biden wins Texas, he also probably wins in Georgia or maybe even Arizona.
- bryanlarsen 6y ago
  Where did I mention any causality?

lmilcin 6y ago

Actually, if you take a look at the graph, you will notice that paths that direct to red leaves are almost all almost entirely red whereas there is a lot of paths to blue leaves that are mostly or even almost entirely red. Or think another way. The largest number of states that can be blue and red still wins is 4. There isn't any combination of any combinations where 5 of selected states are blue and red can still win. To calculate actual probability don't look at just the number of read leaves to number of blue leaves. You need to take into account probability of each of the state being blue or red. I am too lazy to calculate it but at first sight red outcome seems extremely unlikely.

205guy 6y ago
I believe that the "direct" path of one color and "alternating" path are just a choice of the graphing and not dictated by the data. If you make different node selections, you could make the reverse. Same for the layout of the nicely ordered trees and those that seem more chaotic--it's just a matter of adjusting the layout of the nodes to determine the angles of the edges.
- lmilcin 6y ago
  I don't mean the order, but the number of paths with given number of nodes of given color. This is independent on the chosen order or layout of the graph.

advisedwang 6y ago

The visualization presents Florida as the first decision point, but that is a choice of the visualization - you could make it the last choice and the visualization wouldn't look like this. That said Florida does have a large number of electoral college votes and can swing, so it makes sense as a a choice of the designer.

dehrmann 6y ago

Fivethirtyeight puts Binden's change at winning Texas(!) at 36%.

dillondoyle 6y ago

And Pennsylvania. But yeah if Trump loses Florida it's really hard.