Jump to content
270soft Forum

270soft Predicted the Election Better than Pollsters


vcczar

Recommended Posts

I was thinking about this during my hiatus. @admin_270's game was actually more accurate than Nate Silver, Larry Sabato, Sam Wang, etc. I don't know how often you simulated elections with the new "simulation" option, but the results were often strikingly familiar to the real life results. I would wager that it seemed as if half or more of the the simulated elections Hillary Clinton won the popular vote, but Donald Trump won the electoral vote. This actually applies to all the elections, however. I simulated 1788-1956, election by election, and about every 3rd election ended with the candidate that most people wanted not getting elected. Nevertheless, in 2016, 270Soft was the most accurate political machine in determining the election, in my opinion. 

I should also add that, contrary to my own belief, WI, MI, and PA, too easily switched from Blue to Red. However, this was again accurate to this election. However, OR and WA still too easily turn Red. 

Despite the poll numbers, Anthony never adjusted the states to stick stubbornly to how the polls were projecting the election, but rather made clear the possibility of WI, MI, and PA unpredictably going Red, when they are historically Blue-leaning battleground states. 

As such,

270Soft gets an A as far as predicting in this election. 

RealClearPolitics gets a C, they thought Clinton would win, but they had it as a nailbiter. She did win the popular vote (possibly by 2 million votes when they're all counted). 

538, Crystal Ball, Princeton Consortium and almost all the rest get an F. I, too, get an F. 

Link to post
Share on other sites
1 hour ago, Jonathan Kudelka said:

RCP had the final predicted electoral map at 201 EV's for Clinton and 164 EV's for Trump with 171 toss up. All the individual polls and averages were off having Clinton up 4 or 5 nationally. It wasn't RCP fault it was the polls they used that were off

RCP had a filled in map as well for their predictions. It had Clinton at 272, and Trump with the rest. 

Link to post
Share on other sites

That was just a no toss ups maps using the last five polls from each state and taking the average winner. there wasn't really a model or anything like all the other places 

Link to post
Share on other sites

"I would wager that it seemed as if half or more of the the simulated elections Hillary Clinton won the popular vote, but Donald Trump won the electoral vote."

That's interesting.

"I should also add that, contrary to my own belief, WI, MI, and PA, too easily switched from Blue to Red. However, this was again accurate to this election."

The problem with intuitive thinking about these sorts of things is that we're dealing with a small pool of examples (historical Presidential elections), and we have potentially constantly changing electorate-candidates-platforms. So something might seem unlikely ... until it happens.

I think pretty much all my statements from relatively soon before the election about who might win are in this thread that you posted in late October.

 

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...