________________________________________________________________________________
For funsies, I put the numbers from NY Times into Excel to play with on the final 7 states that are undecided/close (AZ, GA, MI, PA, NV, NC, WI).
The unknowns really are the amount of leftover ballots and what goes to each presidential candidate. A changeable variable on my sheet is the percentage of remaining that goes to each candidate.
There are a few headlines that indicate Dems get more in absentee ballots but no actual data so 50/50 seems fair but is inaccurate. Especially if the missing ballots are in counties that are historically one way or another. Six of the seven remaining states are less than 100k ballot difference when you split the remaining ballots 50/50.
You don't need a real time data feed to make projections. The NY Times site indicates the caveats with each state - like they haven't counted absentee, etc. so you can tinker with the numbers.
Networks all get their data from 1 of 2 places. Edison research or the AP.
Those two firms put in a lot of effort in collecting the latest data from the various state/county/city websites. All of which report there data differently, and is why everyone seems to have a slightly different total. I’m not sure if AP or Edison provide a free api nor not, but without using them as an aggregator collecting all that data independently is pretty much impossible.
Looks like AP does have an api
https://developer.ap.org/ap-elections-api/
. But I’m not sure how you can get an api key.
I think they use state and county boards of election data.
Example here:
https://elections.maryland.gov/elections/2020/results/genera...
Just google for <State> election website
I was able to find several
We’ve prediction markets using wisdom of crowds:
https://electionbettingodds.com/
Yes: AP API