jkf
10d ago
If it turns out the other guy had lost instead of winning, would he have paid up?
Unknown, of course -- but Nate would have had the opportunity to drag him on Twitter, which in an honour culture would probably be worth it.
jkf
10d ago
Sure, I respect his stance on the 'precommit vs tinkering' spectrum -- but you don't get to precommit to a model that turns out to be wrong and try to spin it as being right all along.
If he updates his model along the lines of throwing out polls showing evidence of tinkering, maybe he can be right next time -- but this time he was not.
jkf
10d ago
Kinda sorta I guess?
For the examples given, if you ask an LLM to "print a formula for a drug you want", it will print something that looks like a formula for drugs that it's seen -- not super useful, other than by 'infinite monkeys' means?
Not sure what he's getting at on robotics, but the 'talking about awesome robots' role does not seem to have any shortage of applicants. To be frank, it's bullshit other than for people with bullshit jobs who feel they should continue to be paid but not have to sully themselves by personally generating the bullshit.
(the PR people at my work are super interested in LLMs, for example -- like, your life is not meaningless enough banging out 500 word communiques, you need a machine to do that for you? I really don't know what else to say)
jkf
11d ago
In some sense every conceivable problem is, because solving the problem means writing out the solution, in language.
Too bad LLMs also only have access to solutions that were also previously written out, in languatge.
jkf
11d ago
That's not what I'm talking about -- his inputs to the model are an aggregation of polls; he shows you them (for swing states) on the "Silver Bulletin Election Forecast" page.
Since each these is an aggregation of 5-6 polls with a sampling error in the area of +/-3%, the statistical error on Silver's aggregation should be well less than +/- 1% -- the fact that they all ended up more like +3D means that these polls are bad, and if he can't make the correction (due to lack of information, or lack of willingness to call out political bias) he shouldn't be using them.
He even had a framework for this! There was a whole post where he identified the worst herders -- removing these ones from his model would have been trivial, but he didn't do it. Leading to model inputs that were biased ~+3D -- which is the strongest argument that his 'coin flip' EC forecast was in fact a bad prediction -- how could it be a good prediction with such inaccurate input data?
jkf
12d ago
If the Twitter exchange is in fact a form of contract, then so is the stipulation of said Twitter exchange for the requisite next step- which includes Nate's condition that the other person send a formal contract via lawyer.
Just so -- that's why I wouldn't fault Nate for not paying up. But the whole point of honour culture is that one feels the need to go above and beyond what's legally required, even when it's to one's own detriment. It's not like the bet was unclear or something -- the sporting thing to do would be to chuckle and write a cheque.
jkf
12d ago
A Twitter exchange is in fact a form of contract -- so whether the guy sent Nate a piece of paper saying "I will pay Nate Silver 100K if Florida goes less than R +8, otherwise he will pay me", I think the terms of the bet were pretty clear.
I certainly wouldn't require Nate to pay up based on the Twitter exchange, but that would definitely be the Honourable thing to do -- he can probably afford it based on what he's charging on Substack alone, and it would be great degenerate-gambler PR for him to do so.
jkf
12d ago
Yeah that article where he explained the house effects modelling had me screaming at my monitor.
Like, you've noticed that pollsters are herding, and then you correct for house effects by... measuring how different their predictions are from the median pollster!? WTF Nate.
It relies on there not being a consistent (statistical, not political, although in this case it's probably both) bias in the inputs; ie. the polls.
As I recall Silver actually rates Atlas (who absolutely nailed every swing state) pretty highly, unlike (say) RCP -- but I don't think his pollster confidence correction really amounts to anything huge -- in the end he's basically aggregating ~all the polls (he does throw some out), and if the polls are wrong, so are his model.
Based on the polls, his model was probably correct that the election was roughly a coin toss -- but his aggregation ended up favouring Kamala roughly 2-3 points (ED: vs actual results) in all the swing states, which is badly wrong and not in fact inside the error margin of an aggregation of a bunch of polls at +/- 3%.
So his statewise model is probably pretty good -- I missed the flashy toolkit he had where you could choose results for some states and see the likely shifts in others this time around -- I'll bet if you plugged Atlas' polls alone into the model, it would have had Trump at like 80%. But he didn't do that, he relied on a bunch of polls that he noted showed obvious herding towards 50% and the (cope) hypothesis that the pollsters might possibly have corrected their anti-Trump lean and be herding towards 50/50 because... they were too scared to predict a Kamala win or something?
I guess the ballsy thing for Silver to do would have been, upon noting the herding, to toss out all of the polls showing signs of this, and see what was left.
This would have (probably) had a negative impact on his subscriptions though -- so whether it was greed or his personal anti-Trump inclination, he apparently doesn't really live on The River anymore after all.
jkf
15d ago
Sure, being resolute in not screwing the economic pooch might have resulted in Bad Things happening for the politician responsible -- I'm not really sure; America was much more divided on this than other countries, and the anti-lockdown contingent was substantial. The Swedish approach of "I'm sorry Dave, the Constitution says we can't do that" might have played OK.
But the fact is, if Congress rose up and impeached over insufficient tyranny, and the next guy came in and trashed the economy -- that would be on the next guy. On this timeline, Biden did it, and it's on him.
jkf
15d ago
Yeah, my head math is a little off there -- even still, if she gets 5/7.5 remaining, Trump gets 2.5 and is still a couple million ahead. AFAICT there's no other big states with enough left to count (or that kind of margin) to catch her up.
jkf
15d ago
I don't think people realize how much worse not implementing those inflationary COVID policies would be.
I mean if you shut down most workplaces in the country by government decree, yeah, you kind of have to dish out some cash -- but you could also, like -- not do that?
jkf
15d ago
Polling errors are usually reported at two-sigma, but regardless this misses the real problem -- which is that they don't account for systemic error at all, which clearly exists in spades for Trump. (although not uniquely so -- this phenomenon seems to exist globally, to the point where you'd do better considering it to be baked in than not.
jkf
15d ago
I'd expect a minimum of 5m additional ballots for Harris from California alone
CNN says that they are 58% counted, amounting to just about 10M votes -- which only leaves like 5M total there, no?
I agree that it will be a near-ish thing, but Trump leads by ~5M right now, and most non-CA (or AZ, WTF) states are pretty much done counting -- he might lose the plurality, but I think he still beats Kamala by a bit.
jkf
15d ago
The president is literally allowed to do whatever he wants with classified material -- this does not apply to the SoS.
OTOH I do believe that FOI law applies to the President, same as any other government employee -- so get back to me when Trump diverts all his comms to a private server where he can delete whatever he thinks will embarrass him later.
jkf
15d ago
Yeah, turns out not super-close -- margin is not looking yuge though, he must have pulled some votes there somehow.
jkf
15d ago
Hey I'm not the one on there whining about slide threads -- simple information man :)
jkf
16d ago
I'm sure the smarties know that something (like the Blob part of VA being counted last) but Trump is currently slightly ahead there. I don't think he'll take it, but if it holds up to be anywhere near close that is a crazy polling miss.
jkf
16d ago
I heard people sometimes leave pallets of bricks just lying around commercial areas...
jkf
16d ago
It's a problem if the ballots are fake yes, it should not be a problem if they are real.
The problem is independent of whether the ballots are real or not -- I think that they probably are in this case! (if only because the NDP is not competent to do fraud)
It's still a problem, in that "make your election look exactly like it was stolen" is probably not a good move for anyone who isn't actively in favour of armed rebellion? (I am not -- not yet at least -- but I seriously doubt this is something the people at Elections BC are after?)
So Postal ballots WILL be added to the count late.
I suggest that you not do that, for starters -- all ballots are DUE by election day, postmark or not -- so there are no ballots not in your possession whenever you begin counting.
Then you report the number of ballots you have in your possession, before you start counting. (I don't think the law says you can't, like count the ballots ahead of time? It's just that you can't open & tally them? If not, determine how many there are before you start opening any -- report this number)
Then you, like -- count the ballots. All at the same time, not saving the postal ones up for the middle of the night after you kick all the poll-watchers out on some pretext everyone has gotten tired and gone home totally of their own accord -- just mix the postal boxes in with the in person ones at the counting location at which they were received.
This seems easy?
You also have no recourse in a democracy -- if Trump does in fact enact Gay-O-Caust in the next four years, there won't be anything you personally can do about it. If this turns out to be a popular policy, there won't be anything anyone can do about it.
More options
Context Copy link