Presidential Plinko
presidential-plinko.com
presidential-plinko.com
anyway, I agree the cartoon fox seemed very condescending and I wonder whether there wasn't some real frustration with the public behind it.
Have you considered that you are quite possibly both more educated and more intelligent than the people whose experience the cartoon fox is designed to improve. Sure, I don't need it, but I haven't even noticed it since the first time I went to the forecast page this cycle.
I'd be kind of curious to see if they have any evidence that it actually helps. It seems like it... might? But it's always hard to tell with these things.
Moral of the story is, one should never assume that their candidate winning in polls is an excuse to not vote.
People are bad at dealing with uncertainty. And also don't read.
No, it is literally a set of maps of different outcomes of simulations representing (in a reduced way, because there's only room for 22 maps, at least on desktop; I think they have fewer on mobile) the range and frequency of simulation results.
The balls come after that.
It could be the counter-intuitive nature of probability, but I wonder if it might sometimes just be that, psychologically, people have trouble accepting the idea that nobody really knows.
You look at a site like FiveThirtyEight, you know the site is run by an expert, and you know that multiple sources of data have been fed into number-crunching computers. You've seen other situations where this type of process nails it. You have a strong desire to know the answer, and that biases you toward believing that you do. The idea that something so important is unknown is unpleasant, so you reject it.
About the difficult of grasping the basic concept, if you've ever seen the TV show Card Sharks (https://en.wikipedia.org/wiki/Card_Sharks), a key part of the game is that contestants see a card which is face up, and they have to make bets based on whether the next card drawn is higher or lower. In that context, they have no trouble understanding that if the face-up card is a 5, the next card is probably higher (6, 7, 8, 9, 10, J, Q, K, A) but it could also be lower (2, 3, 4).
Essentially, I think people can understand that unlikely things happen, but where they have difficulty is seeing/admitting that this is one of those situations.
I developed and taught an entire undergraduate math class on the mathematics of game shows.
https://people.math.sc.edu/thornef/schc212/gameshows-428.pdf
Much of the class I worked out probabilities like these, and compared what contestants "should" do to what they actually do.
Watching reruns of Card Sharks, The Price Is Right, and others, I've found that most people have a surprisingly good intuitive sense of probability. Not always, but more often than not.
I also seem to remember seeing a study that people made better probability-based decisions if given probabilities phrased as fractions than percentages; I wonder if something similar is going on with the cards example.
The Card Sharks example is a genuinely random event. If you draw a random card, it will probably be higher than 5, but it could be lower.
Whereas an election is not truly random. By election day the outcome is essentially foreordained. The "randomness" that 538 is trying to model are things like uncertainty and unreliability in the polls. What is the relationship between what voters said they will do, and what they actually will do? You can make inferences based on the past, but every election is different, and the mood of the electorate is definitely unique this year. Nate Silver knows from his experience to hedge his bets, as he did in 2016.
It's much more complicated and subtle than a genuinely random event. I find it understandable that people get confused.
What people often miss is that a statement like "X wins with 75% probability" means that X's supporters have a very solid chance of losing. The odds (and risks) in the long run aren't the same when there is no long run, just one experiment.
My recommendation would be to change the script so that "dropping all balls" stops after one run, and the button "drop one ball" gets highlighted after that, encouraging the user to click it.
Because it is then that you get to feel how real the chances of ball ending up on the right side are, even if most balls end up on the blue side.
We all only get one drop.
----------
Personal plug: my 3D-printed Galton/Plinko board[1].
If anyone's interested, I can share the STL file or OpenSCAD code.
The single ball drop, with slower motion, gives you a chance to feel the uncertainty.
Perhaps, whenever many balls shown at once, all but one should be faint/outlined or even X’d out?
They even have a mode that fills in the states in roughly the order of poll closing times, which helps to get your head around how election night is likely to go. (Except it'll probably be election week and poll closing times will matter less than how fast various states count mail-in votes.)
The probability is expressed by the number of times a ball ends up in each of the end slots. That's the whole point!
They are being dropped from the center of the probability distribution, not the center of the electoral vote threshold (which would be nonsense).
> skewing of the entry point
Skewing from what?
> bias in prior distribution
That's the point!
> color of the end slots should change based on probabilities
The height of the bar changes based on the probabilities.
A biased sampling with uniform slots at the bottom shows the same thing, yes.
1. We’re dropping the ball at the position the polls (and our models) indicate.
2. The balls may end up in a different place due to any number of unforeseen factors (the pegs).
That would be hitting on a 30% chance in 2016 and hitting on a 20% chance in 2020. That's a 6% chance overall.
6% likely outcomes happen all the time. It would be a textbook example of "resulting" to draw a conclusion based on this.
That's true only if you assume perfectly spherical pollsters :) After a real or perceived polling miss (in reality, most polls weren't off by all that much in 2016), pollsters tend to change assumptions, sometimes overcompensating. Notably, after a moderate polling miss in the 2015 elections, which predicted a hung parliament where in fact the Tories managed a small majority, UK pollsters went on to overcompensate in the 2017 election, showing a Tory blow-out win whereas in fact they ended up with a hung parliament.
(That one, incidentally, shows one danger of polls! The first poll allowed Cameron to safely promise a vote on the EU, because he assumed they'd be either out of government or in coalition with the libdems, who wouldn't allow it. In fact, they won and were forced to go through with it. The second one allowed Theresa May to call a snap election to gain seats so as to have enough spare MPs to be able to ignore the ERG and negotiate a semi-sensible Brexit. Instead, they lost seats, had to deal with the ERG and Unionists, and ultimately ended up with the current catastrophe. The current Brexit mess is at least in part due to polling misses.)
In the US, pollsters have started to pay a lot more attention to education, which was a major predictor in the 2016 election.
So while you say 6% "happens all the time" - no it actually happens 6% of the time. But 40% for both terms would indicate more like he had a 70% last time and maybe another 60% this time. The pollsters can be correct for fucking California, but what good is that if they're super wrong because racist people in Alabama don't click or pick up the phone (or have no phone to begin with) or talk to pollsters.
To clarify, my intention was to state, "outcomes with 6% of odds happening will occur very frequently" - not relatively. A good hitter in baseball hits homers about 6% of the time. If they hit homers in 2 at bats, we would not have enough information to say much about their true talent.
The real problem was a lack of high-quality state-level polls, and in one or two cases major misses on what state polls there were; there wasn't all that much visibility on many states.
And ultimately, the final result was so close that it was below the margin of error of any reasonable poll to detect. Even assuming the highest quality polls conducted in each state, every day, the best they'd have been able to say would have been, pretty much "It's 50/50", in retrospect.
That wasn't even really a problem; the error in state polling was pretty consistent with the error in national polling; the problem with predictors other than 538 is they treated state variations from polling averages as independent, 538 correctly assessed them (based on past evidence) as highly correlated, which is why Trump had a nearly 1 in 3 chance in 538s forecast.
The problem is people taking (honestly or just for the purpose of after-the-fact criticism) "1 in 3" to mean "absolutely won't happen".
I'm starting to believe that sites like 538 feed off of the general public not understanding that a 20% chance to win is 1 in 5. Endless ink has been spilled about how the public does not understand statistics but very little effort has been made to communicate effectively.
The whole thing seems like a navelgazing sideshow where the pollsters want to have it both ways. They want to claim their predictions are infallible but then when the public says they failed they want to backpedal with holier than thou "well, actually..." excuses.
How, then, should statistics be appropriately communicated?
> They want to claim their predictions are infallible but then when the public says they failed they want to backpedal with holier than thou "well, actually..." excuses.
Do pollsters generally try to claim their numbers are infallible? If anything, the fact that margins of error are included in the results would seem to imply the opposite.
*https://pbs.twimg.com/media/Cwc-j6YXUAEIMaM.jpg
The pollsters take too much credit when they're right and deflect too much blame when they're wrong. They need to do a better job of being humble and stop trying to pass themselves off as apolitical number crunchers just giving us the facts.
What would "good faith" communication look like, then?
> I remember very clearly sources like this* in 2016.
Can you elaborate on what exactly is wrong with that source?
(That isn't a pollster, too; that's a (meta?)analysis/aggregation, but that's a relatively minor nitpick)
> They need to do a better job of being humble and stop trying to pass themselves off as apolitical number crunchers just giving us the facts.
Are they themselves claiming that they are "giving us the facts", or is that how they are being represented by others?
And are you talking about actual polls, or about predictions based on said polls?
In particular look at this (which is almost exactly what happened):
But if there’s a 3-point error against Clinton? That would still leave her with a narrow lead over Trump in the popular vote — by about the margin by which Gore beat Bush in 2000. But New Hampshire, which is currently the tipping-point state, would be exactly tied. Meanwhile, Clinton’s projected margin in Michigan, Pennsylvania and Colorado would shrink to about 1 percentage point, while Trump would be about 2 points ahead in Florida and North Carolina. It’s certainly not impossible that Clinton could win under those circumstances — her turnout operation might come in really handy — but she doesn’t have the Electoral College advantage that Obama did in 2012, when he led in states such as Ohio and Iowa and had larger leads than Clinton does in Michigan and Pennsylvania. In particular, Clinton could be vulnerable to a slump in African-American turnout.
You are confusing pollsters with "forecasters using data from pollsters". These are not the same people.
Many of the forecasters in 2016 were very bad because of naive assumptions about how polls related to election results, particularly many making the assumption that polling errors were independent between states (also, lots did a really poor job of poll aggregation before those naive models.) The reason is Nate Silver and 538 had made a name for themselves in the preceding couple of cycles, and lots of people who didn't understand the process but saw the outcome decided they could do the same thing, because, hey, how hard could it be to compile a bunch of polls and model outcomes based on them?
It does a pretty good job - it gave Trump nearly 30% chance of winning in 2016 and given his small margin that seems reasonable.
Sites like Huffington Post which gave Clinton 99%+ chance are the ones which should be criticized.
To the author: please turn on TLS for your site. It's free.
https://www.zdnet.com/article/china-resurrects-great-cannon-...
I expected better statistical understanding from this audience, but maybe that only underscores how difficult it is to get people thinking correctly about probabilities.
And, of course, a few tens of thousands of votes different on the day, below the margin of error of any poll, and no-one would be bothered about them. In a way, 538 can't win here; they said "there's a 30% chance of this thing happening", it happened, based on differences below the resolution that they could even see at, and everyone's annoyed with them.
A model that gave Clinton a 70% chance of winning is not proven wrong if she loses.
We can't know if the model is right or wrong unless we run the election many times.
Nope. To fully demonstrate how uncertain predictors’ forecasts are, you need to animate a few meteorites crashing into the Plinko board, one said meteorite with Hillary's emails on them. And then go read The Black Swan, especially the chapter where one of the reference figures is a Plinko board.