Talk:3007: Probabilistic Uncertainty

Emotional spirals are useless. I've been coping by pretending we're in scenario 1, it keeps me sane. If I'm wrong, I'll jump off that bridge when we come to it. Barmar (talk) 20:23, 4 November 2024 (UTC)

And I have a friend whose strategy is baking. It's both therapeutic and delicious. Barmar (talk) 20:41, 4 November 2024 (UTC)

I see I don't know US geography well: which bridge you can jump from to leave it? -- Hkmaly (talk) 02:34, 5 November 2024 (UTC)

Most of them. Some of them may be 'caged in' for safety/anti-suicide/anti-DropThingsInThoseBelow purposes (or a covered bridge). Relatively few of the others will be ones that you would have no qualms about vaulting the railing, but (as well as it clearly being a witticism by Barmar) I think you could easily find a bridge that you could jump off. And the resulting falling part isn't at all the difficult bit. Landing safely (or, in extremis for those desperate enough, in a guaranteed immediately fatal manner) is more the challenge. 172.70.86.206 14:48, 5 November 2024 (UTC)

I can't help but think that at preparing for the negative outcome regardless of which outcome is more likely (unless that outcome is *very* unlikely) is a healthy thing to do. 172.71.147.141 20:30, 4 November 2024 (UTC)

"Hope for the best, prepare for the worst" is my usual approach to things. Elektrizikekswerk (talk) 07:45, 5 November 2024 (UTC)

This comic appeared the day before the 2024 United States Presidential Election. At publication time, polls were strongly suggesting about a 50/50 odds that either major candidate would win. Recent news items included advice from mental-health professionals on how to deal with election-related anxiety. 172.71.167.195 20:32, 4 November 2024 (UTC)

Definitely related. This should be in the text, not in the comments, frankly. The yanks are going nuts about the election right now. 172.71.124.243 20:57, 4 November 2024 (UTC)

Did the advice suggested narcotics? -- Hkmaly (talk) 02:34, 5 November 2024 (UTC)

My personal policy is to expect and prepare for the worst. That way I can be surprised when it doesn't happen, and not surprised when it does, rather than the other way around. I don't "do" emotions, so it's basically just planning and mumbling colloquialisms involving the digestive system... 172.71.134.64 21:31, 4 November 2024 (UTC)

As someone who used to think this way, this is obstructively cynical, and downright sad. I mean, in theory you should be pleasantly surprised by the good, and prepared for the bad, but in practice you just dismiss anything good and focus exclusively on the bad. As someone with experience in this type of thinking, it isn't healthy. 172.71.22.85 15:15, 5 November 2024 (UTC)

I can't help but feel that it's mostly Democrats that are anxious, where Trump winning is the bad case. Not being an American I don't have much perspective. Are many Republicans likely to also be anxious, and if so, why? 172.69.60.170 21:55, 4 November 2024 (UTC)

Not sure about "anxious", for Trump-supporting Rs (as opposed to Trump-opposing ones, who are both anxious and tremendously conflicted), but there's certainly a buzz of some emotion. That, if their expectations/hopes/desires are dashed, seem more likely to turn into more direct push-back than Ds would in their case. i.e. if Trump truly wins, there'll be turmoil as the legitimate government forcefully pushes against large subsets of the people, if Harris truly wins then small but determined fractions of the people will push back against the legitimate government. (If it's any way ambiguous, for long enough, which 'truth' indicates a win, it could easily be people vs. people for at least as long as the confusion lasts, with very little reason to believe that it'll be Harris supporters throwing the first stone, probably making Florida 2000 look like a "neat transition"). But this is just what it looks like at this moment. Within a day we might get to see whose words get eaten, or it could be at least a month of building tensions (due to the US system of elections, deliberately legislated to be so much more inefcicient than it needs to be, compared to various other Western nations). 172.68.186.106 15:28, 5 November 2024 (UTC)

From what I've seen the ones in public-facing forums seem pretty indifferent. They do talk a lot about election fraud though. 172.70.34.117 (talk) 22:42, 4 November 2024 (UTC) (please sign your comments with ~~~~)

I like that the comic leaves "good" and "bad" open to interpretation.172.70.211.83 22:29, 4 November 2024 (UTC)

He doesn't want to start fights in the comments/discussion pages/replies! Good to see him appealing to no specific demographic in this one. -P?sych??otic?pot??at???o (talk) 22:40, 4 November 2024 (UTC)

Considering that the "Harris for President" banner is still active, I'm not sure I agree with that. 172.68.22.4 22:53, 4 November 2024 (UTC)

yeah, for that reason i think it's more just so the comic can have further longevity, as this way it can be applied to any number of things with two outcomes, not just the current election 141.101.109.193 00:02, 5 November 2024 (UTC)

Well, so far so good ... -- Hkmaly (talk) 02:34, 5 November 2024 (UTC)

Re Further, with regards to N/A - the odds of "precisely" 50/50 are probabilistically zero: Bear in mind that with the Electoral College system and the fact that only 7 US states are "likely in play," we are talking only hundreds or thousands of realistic possibilities. The odds of a 269-269 tie in the Electoral College are far more than 0. One possibility of a tie that is "on the radar" is if the Republicans take Georgia, North Carolina, Wisconsin, and the 2nd Congressional District of Nebraska (which is very likely to go Democratic) and the Democrats take Pennsylvania, Michigan, and Wisconsin. If you consider just the 7 "in play" states but Arizona "flips" from Republican to Democratic, there are 3 combinations that yield a 269-269 tie. 172.70.210.249 01:29, 5 November 2024 (UTC)

If there's a 269-269 tie, that's basically going to be a Trump win due to how the contingent election process works. (For that matter the far more plausible 270-268 to Harris, which happens if she wins Nevada but not Pennsylvania, is likely going to result in Trump getting the presidency as well, but let's ignore that.) However, many analysts, when faced with numbers like Nate Silver's 50.015%, are going to round it to 50% or 50.0% in the public-facing reports, resulting in apparent exact 50/50 odds even if mathematically they actually favor one side slightly. 172.71.130.3 10:07, 5 November 2024 (UTC)

There's little point in being so precise, since the fraction is far less than the margin of error in the polling. Anything between 49% and 51% is essentially a toss-up. If the 51% is in your favor you can feel hopeful, but hardly confident. Barmar (talk) 15:11, 5 November 2024 (UTC)

Re We contacted several researchers who are experts in emotional spirals to ask them, but none of them were in a state to speak with us: Is it a stretch to think that the emotional-spiral experts were all "in Puerto Rico" (which is not a state), emotionally speaking? In the last week a supporter of one of the candidates insulted Puerto Rico and by extension, people of Puerto Rico and Puerto Rican descent, causing an emotional uproar all over the inter-tubes. 162.158.90.210 01:37, 5 November 2024 (UTC)

Created an account just to say this; don't get mad at me but in my opinion, both candidates are equally bad, which has led to a weird sense of calmness in me due to my belief that we'll be equally screwed no matter what, just in different ways. Tbh in my opinion both candidates are in between what their supporters think of them and what their opponents think of them. Please be civil if you reply, no ad hominem please. BurnV06 (talk) 05:24, 5 November 2024 (UTC)

No, one of them is clearly worse than the other. How do you feel about LGBTQ+ rights? Abortion? Medicare? Teaching kids that racism and homophobia in schools is bad? Well, if Project 2025 is anything to go by, one side clearly is the unpreferable unless you're a white, Christian, rich, and male. This is not a "both sides" issue. One is clearly the worse option. And frankly, I wish centrists knew this. I can agree to disagree on some issues but I just cannot elect someone who wants to punish people for the egregious crime of, gasp, not conforming to societal standard of gender and romance.172.71.22.85 15:15, 5 November 2024 (UTC)

... And Project 2025 is absolutely nothing to go by. It's what a (private) conservative think tank (privately) wants to see implemented. Trump had no involvement in its contents or publication. The Heritage Foundation has been publishing things like it since 1981; it only attracted attention THIS year because politically-motivated people are trying to scare you, and were running out of ideas. It should not surprise you to learn that people who you already disagree with, have ideas that you also disagree with, and might publish compilations of those ideas you disagree with on a regular basis. Freaking out over Project 2025 is like if conservatives started freaking out over a set of published policy recommendations by the Center for American Progress. 172.68.3.96 16:56, 5 November 2024 (UTC) MeZimm

Fair enough, but the point still stands that this is explicitly nota both sides issue. Even taking Project 2025 out of account, one side is clearly worse.172.68.71.44 17:02, 5 November 2024 (UTC)

"Equally bad" is highly subjective, Burn. More people would consider "a total disaster" vs "at least they're not a total disaster!" as a closer truth (whether their own personally-configured disastermeter comes in a Red or Blue casing), and consider balancing dead in the center of the fence to be the most inexplicable position to take. (Not to mention those like above, and also their antithesis opinions, who have a very definite good/bad opinion 9n the pair.)

Not that I'd support being mad at you, as the problem with politics today is too much extreme polarization (we need more moderate voices, rather than wedging open an ever wider void between both limits of opinion). But there's just no realistic middle-ground to gather support around, and what middle-ground there is might also be moving one way or another (depending upon who you ask), so I'm afraid that the strictly neutral "as bad as each other" types are just guaranteed to be setting themselves up to be disappointed. In the 'best' case scenario, disappointed that things aint turning out to be as bad as feared, but I'm not sure that's reassuringly likely enough to comfort you. 172.68.186.106 15:28, 5 November 2024 (UTC)

Well said, and I think it's important to mention that the reason there isn't any moderates is that the moderates just don't care anymore. At least online, complete political apathy is a position I've seen a lot of people take ("Why are they constantly slamming politics into my face, I just don't care"). Unfortunately, these kinds of people are also the moderates, people who aren't particularly one side or the other. This leads to a political landscape where you have 2 extremes, and a bunch of people in the middle who couldn't care less because of said extremes. 172.71.31.24 15:39, 5 November 2024 (UTC)

Not to be a “Discord mod” here, but the entire idea of the 50-50 portion of the comic alluding to the election today is just a theory. Y’all are reading in wayyyyyyy too deep. The comic isn’t even directly saying if one candidate is better (although the Header text is supporting Harris). The discussion is supposed to be for discussing the comic and how to improve it, not clash over ideological differences. Maybe instead of arguing about who’s the better candidate, we can finish up the comic explanation, which is extremely bare bones? TL;DR: break it up, people. 42.book.addict^{Talk to me!} 18:19, 5 November 2024 (UTC)

I think it's funny that so many Democrats are genuinely terrified of the results and spend their days anxiously refreshing 538, whereas Republicans are filled with optimism and already know that the democrats have run the weakest candidate since Dukakis. Ah well, maybe in four years you'll actually get to vote for who leads your ticket instead of having them be appointed by the party elites directly without a vote. ;) 172.71.22.120 (talk) 07:35, 5 November 2024 (please sign your comments with ~~~~)

Given the indirect democracy system the US has, there's a number of problems with who gets to be President. And if Harris is weaker than H. Clinton, but it's still on a knife-edge of popular/EC voting, does that mean that Trump's win was therefore less legitimate? 172.68.186.106 15:28, 5 November 2024 (UTC)

Do I really have to remind you that election results are not the same thing as poll results? In 2016, FiveThirtyEight gave Trump a 28.6% chance vs Clinton's 71.4% chance. Most polls were even more egregiously in favor of Clinton. Yet Trump won. Now in 2024, you say "it's still on a knife-edge of popular/EC voting" - somehow pretending the PREDICTIONS of right now are in any way comparable to the ACTUAL RESULTS of 2016. Yet polls get "shy Tories" and pranksters and all kinds of complicating factors (even assuming the pollsters are being honest - which is not something you should EVER "simply assume"). Polls are a little bit better than astrology in terms of actual predictive power. So comparing "polls now" to "actual election then" is completely wrongheaded. You have to compare predictions to predictions. And the predictions of 2016 were "the odds are MASSIVELY in Clinton's favor" - yet now they are running a WEAKER candidate and rate her has having even LESS probability of winning than Clinton did. Don't worry, though, I'm sure they figured out some way to solve all the problems with their 2016 process, and are now 100% trustworthy again! /s 172.68.3.127 19:45, 5 November 2024 (UTC) MeZimm

Whatever direction we're going with the rest of the arguments, don't lead us down the route of misanalysing (say) 28.6% vs. 71.4% as meaning anything other than that's the predicted chance (by a necessarily incomplete process) of the process coming out one way or another (even by just one vote that swings just one EC contribution). It doesn't mean that the popular vote will split by that proportion or the EC votes will split that way, it is just an assessment of how much the (each slightly biased) coins will fall either majority heads or majority tails. But we only see the one end result (itself a fudge of a fudge of many possibly imperfect opinions) and try to read the entrails all while hearing "but the predictions were 29/71, and it was much closer than that, so obviously those stats guys were wrong".

Do try not to fall for such statistical fallacies. The polling will be refined for all the things that it can be refined for (accounting for the kinds of people who do vote but don't answer to pollsters, or don't vote even though they say they will, etc) and should come with error bars which can be very telling but rarely get mentioned in 'executive summaries' that get selectively quoted by the headlines of organisations with less integrity and more of their own message to try to promulgate.

But looking across many polls, you can see even the 'headline figure' end predictions, shorn of the most obviously optomistic/pesimistic extremes, smeared from several percentage points one way to a similar the other. If the result is within one, two or three swingstates'-worth of ECs, it'll still vindicate most of the polling opinions. Though doesn't mean you can guarantee the reverse. Anyway, not long now until the process stops being fed by votes and starts being fully chewed on by those who produce the 'answer' to this year's big question. 172.70.162.185 20:56, 5 November 2024 (UTC)

I'm reminded of some of my coding theory class, where the absolute worst bit error rate is 50%. Less than 50% and you can repeat the data to detect and correct the errors to some vastly low probability of an incorrect result, and more than 50% and you can invert the signal which flips it to less than 50%, then do the same. At exactly 50% you're essentially getting random noise, and there's nothing you can do about that (but allow allows a one-time pad encryption to be unbreakable if done correctly). --172.71.214.13 18:03, 5 November 2024 (UTC)

There's been talk about pollsters potentially herding because of just how tight the race allegedly is across all of the swing states (which should be more inclined D/R relative to each other, not all exactly even). I think Nate Silver made a tweet about the odds that the odds are so close. Could that be related to this comic, indirectly? 108.162.238.61 (talk) 20:02, 5 November 2024 (UTC) (please sign your comments with ~~~~)

What's with the section talking about strategies to manage expectations? It reads like it came straight out of ChatGPT. 172.71.102.155 04:29, 6 November 2024 (UTC)