Watch live streaming: https://www.youtube.com/watch?v=y60wDzZt8yg
Viewing is banned in Russia by uploader :-(
Live Presidential Forecast
31–40 of 318 posts
Re: Live Presidential Forecast
#32Earlier quoted context omitted.
I just checked. They gauges moving around is definitely random. The updates are done with a "president.json" file, fetched every 15 seconds, but the timestamp in the json data only changes every 30-180 seconds. So all of the movement is just random jitter to make it engaging. Maybe a bit disingenuous. (Note: I first thought they might be using websockets, but no WS connection is made on the page. Only a ajax fetch of…
I think the idea is to indicate these numbers have error bands: The estimates below include an estimate of uncertainty. We expect the uncertainty around these estimates to narrow, especially after races are called. I don't see any reason why we'd expect the uncertainty to be updated more than every 30-180 seconds or why any of this is the least bit "disingenuous".
That would make sense, if they didn't have _actual_ error bands as well.
Re: Live Presidential Forecast
#33Earlier quoted context omitted.
I just checked. They gauges moving around is definitely random. The updates are done with a "president.json" file, fetched every 15 seconds, but the timestamp in the json data only changes every 30-180 seconds. So all of the movement is just random jitter to make it engaging. Maybe a bit disingenuous. (Note: I first thought they might be using websockets, but no WS connection is made on the page. Only a ajax fetch of…
I think the idea is to indicate these numbers have error bands: The estimates below include an estimate of uncertainty. We expect the uncertainty around these estimates to narrow, especially after races are called. I don't see any reason why we'd expect the uncertainty to be updated more than every 30-180 seconds or why any of this is the least bit "disingenuous".
I'm pretty sure it's so people don't just leave right away.
No one want's to look at a static page fore more than a few seconds, once you have taken in the information.
And I guarantee you that there are plenty of visitors on that page right now, staring at the gauges with fascination.
Re: Live Presidential Forecast
#34I need a drink. Or three :-/
Re: Live Presidential Forecast
#35Nice but the three top gauges are moving suspiciously fast. Do they add a little random the the real number to give a sort of error margin visualisation ?
I think it's an interesting way to represent uncertainty; it seems pretty intuitive to me.
It is a intuitive visualization, but also makes it hard to tell when things actually change.
Re: Live Presidential Forecast
#36Nice but the three top gauges are moving suspiciously fast. Do they add a little random the the real number to give a sort of error margin visualisation ?
Disable jitter by running the following in console: require('forecast/jitter').get = function(){return 0;}
Haha, awesome find.
Did you just dig through the code or are you a NYT dev?
Re: Live Presidential Forecast
#37Earlier quoted context omitted.
I just checked. They gauges moving around is definitely random. The updates are done with a "president.json" file, fetched every 15 seconds, but the timestamp in the json data only changes every 30-180 seconds. So all of the movement is just random jitter to make it engaging. Maybe a bit disingenuous. (Note: I first thought they might be using websockets, but no WS connection is made on the page. Only a ajax fetch of…
I'd venture further to say it's a perversion of "news" for a news site to be injecting garbage into the data for entertainment purposes.
Yes, it is an abuse of our attention and a poor choice in information presentation. Who can we contact at NYT to complain?
Re: Live Presidential Forecast
#38This is very cool. I'd love to see a tech write up later this month... and some stats!
It sure looks good. I hope there's a solid statistical model behind it too though.
Yeah... the jitter being fake (and fast?) seems unnecessary per your other comment.
If the jitter was an actual uncertainty measure or something... or less fast...
Re: Live Presidential Forecast
#39Why is voting actually hard to do? Even though we're so encouraged to vote, there is a lot of red tape. For example, depending on your street address, you may have to vote at a garden variety pick of a building. At my university, I couldn't vote in building A because my apartment was on street X. I ultimately was not able to vote today because it was actually _hard_ to vote.
Freedom isn't easy.
Jokes aside, you're overstating the difficulty to vote. You couldn't vote today because you couldn't pick the building to vote at?
Re: Live Presidential Forecast
#40Earlier quoted context omitted.
I just checked. They gauges moving around is definitely random. The updates are done with a "president.json" file, fetched every 15 seconds, but the timestamp in the json data only changes every 30-180 seconds. So all of the movement is just random jitter to make it engaging. Maybe a bit disingenuous. (Note: I first thought they might be using websockets, but no WS connection is made on the page. Only a ajax fetch of…
I think the idea is to indicate these numbers have error bands: The estimates below include an estimate of uncertainty. We expect the uncertainty around these estimates to narrow, especially after races are called. I don't see any reason why we'd expect the uncertainty to be updated more than every 30-180 seconds or why any of this is the least bit "disingenuous".
> I think the idea is to indicate these numbers have error bands:
It's also a hint that the page isn't static.