"Projected" Voting results?

[Catholic_Dude]

I keep seeing the term “projected” outcome in regards to many election results…what does it mean? If all the votes arent counted then how do we REALLY know?

I understand that in certain circumstances after a certain percentage have been counted and there is a certain majority then that majority will win…but when say ONLY 40% of the votes have been counted, you cant know who wins though…can you?

I keep seeing FINAL “results” with 40% or so counted.

I see these “results” as a joke in my opinion because they are too early to tell.

 

[chicago]

Often the projections are made through exit polling. The other thing which is done is to look at what has been counted and what hasn’t, taking into consideration a candidate’s base of support and where votes are likely to come from which will boost their totals or make margins closer. As the saying goes “What’s out?” Often reasonable assessments can be made via these methods.

 

[didymus]

I keep seeing the term “projected” outcome in regards to many election results…what does it mean? If all the votes arent counted then how do we REALLY know?

I understand that in certain circumstances after a certain percentage have been counted and there is a certain majority then that majority will win…but when say ONLY 40% of the votes have been counted, you cant know who wins though…can you?

I keep seeing FINAL “results” with 40% or so counted.

I see these “results” as a joke in my opinion because they are too early to tell.

I think the 40% (or whatever) refers to percentage of precincts reporting, not percentage of total vote in. So 40% of precincts, depending on the district, might represent 90% of the votes and leave no mathematical possibility of the remaining precints changing the result.

 

[BioCatholic]

I think the 40% (or whatever) refers to percentage of precincts reporting, not percentage of total vote in. So 40% of precincts, depending on the district, might represent 90% of the votes and leave no mathematical possibility of the remaining precints changing the result.

You are correct. There is an algorithm that i cant remember the name of that calculates the outcome of a binary state (on/off, win/loss) based on the probablity that all parameters and numbers that are (name removed by moderator)utted AFTER the current calculation will be AGAINST the hypothesis, and FOR the null hypothesis.

x= 30,000 votes for Jones
y=15,000 votes for Smith
z=%precincts voting (weighted by population)
D=estimated % of total registered voters

hypothesis=Jones wins
Null hypothesis=Smith wins

For a= Precinct 1 To PrecintCurrent
if nullhypothesis<>0
Smithvotes+ votesFromprecinct
else
break
return a
endif
next

this is a small example of how it might work, where A would be the # precincts at which Jones wins hands down.

at some value, there will be a point where even if smith gets every single vote thereafter, he will not win, and thus the hypothesis is true indefinitely.

 

[Catholic_Dude]

I think the 40% (or whatever) refers to percentage of precincts reporting, not percentage of total vote in. So 40% of precincts, depending on the district, might represent 90% of the votes and leave no mathematical possibility of the remaining precints changing the result.

That makes perfect sense.